i

ROUTLEDGE HANDBOOK OF SPORT

AND EXERCISE SYSTEMS GENETICS

Technological advances over the last two decades have placed genetic research at the forefront
of sport and exercise science. It provides potential answers to some of contemporary sport and
exercise’s defining issues and throws up some of the area’s most challenging ethical questions,
but to date, it has rested on a fragmented and disparate literature base. The Routledge Handbook
of Sport and Exercise Systems Genetics constitutes the most authoritative and comprehensive ref-
erence in this critical area of study, consolidating knowledge and providing a framework for
interpreting future research findings.
Taking an approach which covers single gene variations, through genomics, epigenetics,
and proteomics, to environmental and dietary influences on genetic mechanisms, the book is
divided into seven sections. It examines state-​of-​the-​art genetic methods, applies its approach to
physical activity, exercise endurance, muscle strength, and sports performance, and discusses the
ethical considerations associated with genetic research in sport and exercise.
Made up of contributions from some of the world’s leading sport and exercise scientists and
including chapters on important topical issues such as gene doping, gender testing, predicting
sport performance and injury risk, and using genetic information to inform physical activity
and health debates, the handbook is a vital addition to the sport and exercise literature. It is
an important reference for any upper-​level student, researcher, or practitioner working in the
genetics of sport and exercise or exercise physiology, and crucial reading for any social scientist
interested in the ethics of sport.

J. Timothy Lightfoot is the Omar Smith Endowed Chair in Kinesiology and the Director
of the Sydney and JL Huffines Institute for Sports Medicine and Human Performance at Texas
A&M University in College Station, Texas (USA) and holds the rank of Professor of joint
appointments on the Texas A&M Genetics Faculty and the Texas A&M Institute for Genome
Sciences and Society. His research focus has been the genetic factors that regulate daily physical
activity. He is an Associate Editor of Medicine and Science in Sports and Exercise. He was a founding
member of the National Exercise Clinical Trials Network and is a Fellow of the American
College of Sports Medicine.

Monica J. Hubal is an Associate Professor of Kinesiology at Indiana University–​Purdue

University Indianapolis, USA. She is also an Adjunct Associate Professor of Cellular and
ii

Integrative Physiology, a research scientist in the Diabetes Translational Research Center and
is affiliated with the Indiana Center for Musculoskeletal Health at the Indiana University
School of Medicine. Her research focuses on understanding the systems biology of emergent
cardiometabolic disease and the mechanisms driving response to various interventions. She was
a principal investigator in the Research Center for Genetic Medicine at the Children’s National
Medical Center. She is an Associate Editor for Medicine and Science in Sports and Exercise and
Exercise and Sport Sciences Reviews and is a Fellow of the American College of Sports Medicine.

Stephen M. Roth is a Professor of Kinesiology and Associate Dean at the University of

Maryland in College Park, USA. He has researched the genetic aspects of exercise and sport for
over 15 years. He is an author or co-​author of over 90 peer-​reviewed articles, book chapters,
and books. He served as an Associate Editor for Medicine and Science in Sports and Exercise and
Exercise and Sport Sciences Reviews. He is a Fellow of the National Academy of Kinesiology and
the American College of Sports Medicine.
iii

ROUTLEDGE HANDBOOK
OF SPORT AND EXERCISE
SYSTEMS GENETICS

Edited by J. Timothy Lightfoot, Monica J. Hubal,

and Stephen M. Roth
iv

First published 2019

by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2019 selection and editorial matter, J. Timothy Lightfoot, Monica J. Hubal and
Stephen M. Roth; individual chapters, the contributors
The right of J. Timothy Lightfoot, Monica J. Hubal and Stephen M. Roth to be identified as
the authors of the editorial material, and of the authors for their individual chapters, has been asserted
in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised
in any form or by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
British Library Cataloguing-​in-​Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-​in-​Publication Data
Names: Lightfoot, J. Timothy, editor. | Hubal, Monica J., editor. |
Roth, Stephen M., 1973– editor.
Title: Routledge handbook of sport and exercise systems genetics /
Edited by J. Timothy Lightfoot, Monica J. Hubal, and Stephen M. Roth.
Other titles: Handbook of sport and exercise systems genetics
Description: Abingdon, Oxon ; New York, NY : Routledge, 2019. |
Includes bibliographical references and index.
Identifiers: LCCN 2018051636 | ISBN 9781138504851 (hardback) |
ISBN 9781315146287 (ebook)
Subjects: LCSH: Sports–Physiological aspects–Research. |
Exercise–Physiological aspects–Research. | Human genetics–Research. |
Athletic performance–Physiological aspects–Research.
Classification: LCC RC1235 .R683 2019 | DDC 612/.044–dc23
LC record available at https://lccn.loc.gov/2018051636
ISBN: 978-​1-​138-​50485-​1 (hbk)
ISBN: 978-​1-​315-​14628-​7 (ebk)
Typeset in Bembo
by Newgen Publishing UK
v

CONTENTS

List of figures  ix
List of tables  xiii
List of contributors  xv
Foreword  xxi
Preface  xxv

Introduction  1
J.Timothy Lightfoot, Monica J. Hubal, and Stephen M. Roth

SECTION 1
General systems genetics  5

1 Why study the systems genetics of sport and exercise?  7

Frank W. Booth,Taylor J. Kelty, Kolter B. Grigsby, and Gregory N. Ruegsegger

2 Expansion of knowledge and advances in genetics for quantitative analyses  16

Penny K. Riggs

3 Human systems genetic modeling used in exercise  28

Jaakko Kaprio

4 The translation of systems genetics of exercise to everyday life  41

Molly Bray

v
vi

Contents

SECTION 2
Systems genetics of physical activity  53

5 Is physical activity regulated by genetics? Evidence from animal models  55

Scott A. Kelly

6 Is physical activity regulated by genetics? Evidence from studies in humans  67

Matthijs D. van der Zee and Eco de Geus

7 The evolution of genetic mechanisms controlling physical activity  80

J.Timothy Lightfoot, Ayland C. Letsinger, and Jorge Z. Granados

8 Neurogenetics of motivation for physical activity  94

Justin S. Rhodes

9 Peripheral mechanisms arising from genetics that regulate activity  107

David P. Ferguson

10 Toxicant and dietary exposures as unique environmental factors that

affect the genetic regulation of activity  117
Emily E. Schmitt and Heather L. Vellers

SECTION 3
Systems genetics of exercise endurance and trainability  133

11 The evolution of the human endurance phenotype  135

David A. Raichlen, James T. Webber, and Herman Pontzer

12 Endurance phenotype primer  148

John C. Quindry and Michael D. Roberts

13 Heritability of endurance traits from animal research models  164

Joshua J. Avila, Sean M. Courtney, and Michael P. Massett

14 Heritability of endurance traits from human research models  178

Jacob L. Barber and Mark A. Sarzynski

15 Genetic contributions to cardiorespiratory fitness  187

Reuben Howden, Benjamin D.H. Gordon, and Ebony C. Gaillard

16 Genetic contributions to mitochondrial traits  200

Mark Tarnopolsky

vi
vii

Contents

17 Angiotensin-​converting enzyme and the genomics of

endurance performance  216
Linda S. Pescatello, Lauren M.L. Corso, Lucas P. Santos, Jill Livingston,
and Beth A. Taylor

SECTION 4
Systems genetics of muscle mass, strength, and trainability  251

18 Heritability of muscle size and strength traits  253

Martine Thomis

19 Genetic contributions to muscle strength  264

Matthew D. Barberio, Emidio E. Pistilli, and Monica J. Hubal

20 Genetic contributions to skeletal muscle size  277

Philip J. Atherton, Jessica Cegielski, and Daniel J. Wilkinson

21 Genetic contributions to neuroendocrine responses to resistance training  290

William J. Kraemer, Nicholas A. Ratamess, and Jakob L. Vingren

22 Myostatin’s role in genetic control of muscle size and strength  310

Dustin S. Hittel

23 Alpha-​actinin-​3’s role in the genetic control of muscle strength

and performance  323
Jane T. Seto, Fleur C. Garton, Kathryn N. North, and Peter J. Houweling

SECTION 5
Systems genetics of sports performance  345

24 Summary findings on genetics and sport performance  347

Macsue Jacques, Shanie Landen, Sarah Voisin, and Nir Eynon

25 Using elite athletes as a model for genetic research  357

Colin N. Moran, Alun G. Williams, and Guan Wang

26 Twin and family studies in sport performance  372

José Maia and Peter T. Katzmarzyk

27 Sport concussion genetics  388

Ryan T. Tierney and Jane K. McDevitt

vii
viii

Contents

28 Systems genetic factors underlying soft tissue injury  402

Masouda Rahim, Alison V. September, and Malcolm Collins

29 Sex and performance: Nature versus nurture  416

Mindy Millard-​Stafford and Matthew T. Wittbrodt

SECTION 6
The ethics of systems genetics in exercise and sport  431

30 Race and sports performance  433

John Nauright and David K. Wiggins

31 The scientific and ethical challenges of using genetic information

to predict sport performance  442
Andrew C. Venezia and Stephen M. Roth

32 Gene doping: Ethical perspectives  453

Verner Møller and Rasmus Bysted Møller

33 Enhancing evolution: The transhuman case for gene doping  463

Andy Miah

34 The ethics of sex testing in sport  475

Jaime Schultz

SECTION 7
Conclusions  485

35 Exercise genomics, epigenomics, and transcriptomics: A reality check!  487

Claude Bouchard

36 Afterword –​closing the loop: Observations and conclusions  494

J.Timothy Lightfoot, Monica J. Hubal, and Stephen M. Roth

Index  503

viii
ix

LIST OF FIGURES

2.1 An example of a hand-​held minION sequencer in use in a

laboratory setting  17
2.2 The Medline database was queried with the keyword search terms
“genome wide association” or “genome-​wide association studies” and
publication numbers per year are plotted in this graph  19
3.1 Univariate twin model path diagram for decomposition of variance
into additive genetic factors (A), environmental factors shared by the
twins (i.e., are in common C), and environmental factors unique
(E) to each twin  33
3.2 Schematic diagram of the essential feature of the Cholesky
decomposition multivariate model  35
4.1 Genetic information can be used to: 1) improve exercise adherence;
2) optimize exercise prescriptions; and 3) identify individuals at risk
for dropout  43
5.1 Voluntary physical activity is simultaneously influenced by genetic
architecture, the environment, and gene-​by-​environment interactions  63
6.1 Forest plot containing study-​specific and meta-​analytic estimates of
heritability (A) of total physical activity (TPA), moderate to vigorous
activity (MVPA), leisure-​time physical activity (LTPA), and voluntary
exercise behavior (VEB), for males and females  71
6.2 Forest plot containing study-​specific and meta-​analytic estimates of
shared environmental contribution (C) to the variance in total physical
activity (TPA), moderate to vigorous activity (MVPA), leisure-​time
physical activity (LTPA), and voluntary exercise behavior (VEB), for
males and females  72
7.1 Hours per day spent in moderate or vigorous activity per day  85
7.2 Estimated mutation age of SNPs correlated to physical activity  87

ix
x

List of figures

7.3 Summary of potential evolutionary processes that developed genetic

control of physical activity  90
8.1 A simplified illustration of interactions between the hippocampus and
the natural reward circuit in the rodent brain  96
9.1 Schematic indicating the key processes involved with the peripheral
regulation of physical activity along with associated journal citations  109
9.2 Voluntary wheel running in mice that had skeletal muscle annexin
VI or calsequestrin 1 transiently knocked down by Vivo-​Morpholinos  113
10.1 Schematic outlining the unique environmental factor of endocrine-​
disrupting chemicals (EDCs) and the effect on physical activity
regulation  120
10.2 The effect of chronic overfeeding on physical activity regulation and
potential mechanisms  128
11.1 Phylogeny of Hominidae  136
11.2 Range of daily walking distances for great apes  137
12.1 Figure demonstrating that PGC-​1α mRNA increases in human
vastus lateralis (VL) muscle up to 2 hours following exhaustive
cycling at 65% VO2max  153
12.2 Recent data demonstrate in global Pgc-​1α knockout (KO)
mice that hindlimb muscle mitochondrial content is lower
than wild-​type controls  154
12.3 Endurance exercise rapidly activates multiple pathways (e.g., p38
MAPK pathway, CaMK pathway, and AMPK pathway) that
subsequently activate the PGC-​1α promoter and increases mRNA
levels, and subsequent translation rates  154
12.4 Left: the unstimulated AMPK enzyme; right: increased AMPK activity
(during exercise)  156
12.5 Calcium (Ca2+) flux increases in the cytosol during endurance exercise  158
14.1 Mean baseline VO2max across white families from the HERITAGE
Family Study, with families ranked by family mean  179
14.2 Meta-​analysis of heritability estimates of intrinsic VO2max adjusted
by body weight  180
14.3 Meta-​analysis of heritability estimates of submaximal endurance
phenotypes in the untrained state  181
14.4 Heterogeneity of VO2max response to endurance exercise training
across 1724 participants from 8 studies and 14 different interventions  182
16.1 Mitochondrial ultrastructure  202
16.2 Mitochondrial respiratory chain  204
17.1 Angiotensin-​converting enzyme gene (ACE) and physical
performance number of citations per year in PubMed  217
17.2 Flow chart detailing systematic search of potentially qualifying primary
level trials and systematic reviews and meta-​analysis that examined
ACE and endurance and/​or endurance performance related outcomes  219

x
xi

List of figures

17.3 Angiotensin-​converting enzyme (ACE) and cardiovascular

control: classical and nonclassical axes  221
19.1 Molecular pathways involved in the regulation and balance of protein
synthesis and protein degradation  269
21.1 Factors influencing the acute hormonal responses to resistance exercise  291
21.2 Testosterone biosynthesis  292
21.3 A simplified model of intracellular androgen signaling in skeletal muscle 
294
21.4 The androgen receptor gene and protein  295
21.5 Growth hormone somatotrophs in the anterior pituitary  298
21.6 IGF-​1 isoforms  301
21.7 Intramuscular IGF-​1 signaling  303
22.1 The activation (→) and inhibition (-​-​|) of myostatin signaling by
exercise, disease, and its rescue by pharmacological targeting  312
23.1 Localization and structure of the sarcomeric α-​actinins in
skeletal muscle  324
23.2 Frequency of the ACTN3 genotype in controls and elite sprint and
endurance athletes  327
23.3 α-​Actinin-​3 deficiency in skeletal muscle  328
23.4 Z-​disk proteins are upregulated in Actn3 knockout mice  329
23.5 Postnatal replacement of α-​actinin-​3 in Actn3 knockout muscles
restored muscle metabolism and fatigue response to wild-​type levels  335
23.6 WT and Actn3 knockout [Ca2+]i transients at 0, 10, 40, and 60 seconds
during contraction-​induced muscle fatigue  336
23.7 Mechanistic insights from Actn3 knockout mouse model highlighting
the structural, metabolic, signaling, and calcium handling pathways  336
23.8 Actn3 knockout/​mdx muscles show slower disease progression
compared to mdx  339
24.1 Timeline of genetic studies in sports science  348
25.1 Elite distance runner and elite weight lifter: Almaz Ayana (left) and
Mi-​ran Jang (right)  358
25.2 Shinty skill from the Kinlochshiel versus Strathglass Balliemore Cup
Final (2009)  359
25.3 Photograph of the Finnish skier Eero Mäntyranta (1937–​2013) during
the 1964 Winter Olympics in Innsbruck  361
26.1 A three-​generation pedigree of a successful Czech sports family  377
26.2 Pedigrees of Garcés Lopez (above) and Trujillo (below) sports families  379
26.3 Pedigrees of elite and/​or subelite Colombian siblings’ sports families  380
26.4 Family pedigree of Cecília Margarita Baena  381
26.5 Family pedigree of Sandra Viviana Roa  381
26.6 Pedigree of the Tovar Ante family  382
26.7 Pedigree of the Urrutia family  382
27.1 Traumatic brain injury pathomechanics spectrum  389
27.2 Cellular structures involved in concussion pathomechanics and
pathophysiology  390

xi
xii

List of figures

29.1 A theoretical model for contributing factors impacting sport

performance  417
29.2 Mean (± standard deviation) age for US Olympic team members in
swimming (top) and running events in track (bottom) over time  420
29.3 Performance gap in elite top-​16 all-​time US age group, senior national
and world record swimmers  421
29.4 Performance gap in US Olympic trials (1972–​2016) for top eight
finalists by event  424

xii
xiii

LIST OF TABLES

3.1 Definitions of selected terms  29

5.1 Heritability estimates of voluntary wheel-​running traits in three
different mouse models  58
7.1 Human, Neanderthal, and chimpanzee allele comparisons of physical
activity phenotypes  88
10.1 Summary of studies demonstrating effects of endocrine-​disrupting
chemicals (EDCs) on physical activity in rodents  121
10.2 Summary of studies demonstrating direct and indirect links between
overfeeding and reduced activity in humans and rodents  125
13.1 Significant quantitative trait loci (QTL) for intrinsic endurance
exercise capacity  168
13.2 Significant and suggestive quantitative trait loci (QTL) for responses
to endurance exercise training  172
14.1 Heritability estimates for endurance-​related traits in the intrinsic/​
untrained state  180
17.1 A summary of the qualifying trials that examined the influence of
ACE rs4340 on endurance performance  224
17.2 A summary of the qualifying trials that examined the influence of
ACE rs4340 on endurance exercise health-​related outcomes  238
21.1 Comparison of assay signal concentrations from immunoassays and
bioassays during resistance exercise  300
23.1 Sprint/​power elite athlete association studies  326
23.2 ACTN3 genotype and training association studies (endurance and
resistance training)  331
23.3 ACTN3 genotype and inherited muscle diseases  338
25.1 Major study cohorts in the genetics of human elite athletic performance 
364
25.2 Major study consortia in the genetics of human elite athletic
performance (excluding recently established consortia)  366

xiii
xiv

List of tables

26.1 Parameter estimates from the Colombian sports families  378

26.2 Information summary of the seven most outstanding Colombian
sports families  378
27.1 Concussion pathomechanical genetic associations  391
27.2 Concussion pathophysiological genetic associations  392
27.3 Potential concussion genetic associations  397
28.1 Genetic variants within the collagen genes and the associated
musculoskeletal soft tissue injuries in which they were implicated  404
28.2 Genetic variants within genes encoding regulatory components of
the extracellular matrix and the associated musculoskeletal soft
tissue injuries in which they were implicated  406
28.3 Genetic variants within genes encoding cell signaling molecules
and the associated musculoskeletal soft tissue injuries in which they
were implicated  409

xiv
xv

LIST OF CONTRIBUTORS

Philip J. Atherton, PhD Claude Bouchard, PhD

MRC-​ARUK Centre for Musculoskeletal Boyd Professor, John W. Barton Sr. Chair in
Ageing Research & NIHR Nottingham Genetics and Nutrition
BRC University of Nottingham, School of Human Genomics Laboratory
Medicine Pennington Biomedical Research Center,
Royal Derby Hospital Baton Rouge, LA, USA
Derby, UK
Molly Bray, PhD
Joshua J. Avila, PhD Professor and Susan T. Jastrow Chair for
Department of Health and Kinesiology Excellence in Nutritional Sciences
Texas A&M University Chair, Department of Nutritional Sciences
College Station, TX, USA College of Natural Sciences
The University of Texas at Austin
Jacob L. Barber, MS Austin, TX, USA
Department of Exercise Science
Arnold School of Public Health Jessica Cegielski, MSc
University of South Carolina MRC-​ARUK Centre for Musculoskeletal
Columbia, SC, USA Ageing Research & NIHR Nottingham
BRC University of Nottingham, School of
Matthew D. Barberio, PhD Medicine
Center for Genetic Medicine Research Royal Derby Hospital
Children’s National Medical Center Derby, UK
Washington, DC, USA
Malcolm Collins, PhD
Frank W. Booth, PhD Division of Exercise Science and Sports
Department of Biomedical Sciences Medicine
College of Veterinary Medicine Department of Human Biology
University of Missouri Faculty of Health Sciences
Columbia, MO, USA University of Cape Town
Cape Town, South Africa

xv
xvi

List of contributors

Lauren M.L. Corso, MS Fleur C. Garton, PhD

Department of Kinesiology Murdoch Children’s Research Institute
College of Agriculture, Health and Natural Royal Children’s Hospital
Resources Victoria, Australia
University of Connecticut
Storrs, CT, USA Benjamin D.H. Gordon, PhD, CEP
Department of Exercise & Rehabilitative
Sean M. Courtney, PhD Sciences
Department of Health and Kinesiology Slippery Rock University
Texas A&M University Slippery Rock, PA, USA
College Station, TX, USA
Jorge Z. Granados, MS
Eco de Geus, PhD Department of Health and Kinesiology
Head of the Department of Biological College of Education and Human
Psychology Development
Human Behavioral and Movement Sciences Texas A&M University
Vrije Universiteit Amsterdam College Station, TX, USA
Amsterdam, Netherlands
Kolter B. Grigsby, PhD
David Epstein Department of Biomedical Sciences
Former reporter for ProPublica and Sports College of Veterinary Medicine
Illustrated, University of Missouri
author of The Sports Gene and Range Columbia, MO, USA

Nir Eynon Dustin S. Hittel, PhD

Institute for Health and Sport Cumming School of Medicine
Victoria University, and University of Calgary
Murdoch Children’s Research Institute Calgary, Alberta, Canada
The Royal Children Hospital
Melbourne, Australia Reuben Howden, PhD
Laboratory of Systems Physiology
David P. Ferguson, PhD, RCEP Department of Kinesiology
Assistant Professor UNC Charlotte
Director of the Neonatal Nutrition and Charlotte, NC, USA
Exercise Research Lab
Department of Kinesiology Peter J. Houweling, PhD
College of Education Murdoch Children’s Research Institute
Michigan State University Royal Children’s Hospital
East Lansing, MI, USA Victoria, Australia

Ebony C. Gaillard, BS
Laboratory of Systems Physiology
Department of Kinesiology
UNC Charlotte
Charlotte, NC, USA

xvi
xvi

List of contributors

Monica J. Hubal, PhD, FACSM Shanie Landen, BSc

Department of Kinesiology Institute for Health and Sport
School of Health and Health Sciences Victoria University
Affiliate –​Indiana Center for Melbourne, Australia
Musculoskeletal Health
Affiliate –​ Diabetes Translational Ayland C. Letsinger, BS
Research Center Department of Health and Kinesiology
Indiana University –​Purdue University at College of Education and Human
Indianapolis Development
Indianapolis, IN, USA Texas A&M University
College Station, TX, USA
Macsue Jacques, BSc
Institute for Health and Sport J. Timothy Lightfoot, PhD, FACSM,
Victoria University RCEP, CES
Melbourne, Australia Omar Smith Endowed Chair in Kinesiology
Director, Sydney and JL Huffines Institute for
Jaakko Kaprio, MD, PhD Sports Medicine and Human Performance
Professor of Genetic Epidemiology Department of Health and Kinesiology
Department of Public Health, Faculty of College of Education and Human
Medicine Development
& Institute for Molecular Medicine FIMM Texas A&M University
University of Helsinki College Station, TX, USA
Helsinki, Finland
Jill Livingston, MLS
Peter T. Katzmarzyk, PhD Homer Babbidge Library
Pennington Biomedical Research Center Health Sciences
Louisiana State University University of Connecticut
Baton Rouge, LA, USA Storrs, CT, USA

Scott A. Kelly, PhD José Maia, PhD

Associate Professor Faculty of Sport
Department of Zoology University of Porto
Ohio Wesleyan University Porto, Portugal
Delaware, OH, USA
Michael P. Massett, PhD
Taylor J. Kelty, PhD Department of Health and Kinesiology
Department of Biomedical Sciences Texas A&M University
College of Veterinary Medicine College Station, TX, USA
University of Missouri
Columbia, MO, USA Kathryn N. North
Murdoch Children’s Research Institute
William J. Kraemer, PhD The Royal Children’s Hospital
Department of Human Sciences Melbourne, Australia
The Ohio State University
Columbus, OH, USA

xvii
xvi

List of contributors

Jane K. McDevitt, PhD Emidio E. Pistilli, PhD

Temple University Department of Human Performance –​
Department of Kinesiology Exercise Physiology
Philadelphia, PA, USA West Virginia University
Morgantown, WV, USA
Andy Miah, PhD
University of Salford Herman Pontzer, PhD
Salford, UK Department of Evolutionary Anthropology
Duke University
Mindy Millard-​Stafford, PhD Durham, NC, USA
Exercise Physiology Laboratory
School of Biological Sciences John C. Quindry, PhD, FACSM
Georgia Institute of Technology Department Chair, Associate Professor, IHI
Atlanta, GA, USA Cardiovascular Research Fellow
Department of Health and Human
Verner Møller, PhD Performance
Department of Public Health University of Montana
Aarhus University Missoula, MT, USA
Aarhus, Denmark
Masouda Rahim, PhD
Rasmus Bysted Møller, PhD Division of Exercise Science and Sports
Department of Public Health Medicine
Aarhus University Department of Human Biology
Aarhus, Denmark Faculty of Health Sciences
University of Cape Town
Colin N. Moran, PhD Cape Town, South Africa
Physiology, Exercise and Nutrition
Research Group David A. Raichlen, PhD
University of Stirling School of Anthropology
Stirling, UK University of Arizona
Tucson, AZ, USA
John Nauright, PhD
University of North Texas Nicholas A. Ratamess, PhD
Denton, TX, USA Department of Health and Exercise Science
The College of New Jersey
Linda S. Pescatello, PhD, FACSM, FAHA Ewing, NJ, USA
Board of Trustees Distinguished Professor of
Kinesiology Justin S. Rhodes, PhD
Department of Kinesiology Associate Professor
College of Agriculture, Health and Natural Department of Psychology
Resources Beckman Institute for Advanced Science and
The Institute for Systems Genomics Technology
University of Connecticut University of Illinois
Storrs, CT, USA Urbana-​Champaign, IL, USA

xviii
xix

List of contributors

Penny K. Riggs, PhD Alison V. September, PhD

Associate Professor of Functional Genomics Division of Exercise Science and Sports
Department of Animal Science Medicine
College of Agriculture and Life Sciences Department of Human Biology
Texas A&M University Faculty of Health Sciences
College Station, TX, USA University of Cape Town
Cape Town, South Africa
Michael D. Roberts, PhD
School of Kinesiology Jane T. Seto, PhD
Auburn University Murdoch Children’s Research Institute, and
Auburn, AL, USA Department of Paediatrics
University of Melbourne
Stephen M. Roth, PhD The Royal Children’s Hospital
Department of Kinesiology Melbourne, Australia
School of Public Health
University of Maryland Mark Tarnopolsky, MD, PhD
College Park, MD, USA Department of Pediatrics and Medicine
McMaster University
Gregory N. Ruegsegger, PhD Hamilton, Ontario, Canada
Division of Endocrinology, Diabetes, and
Nutrition
Department of Medicine Beth A. Taylor, PhD
Mayo Clinic Department of Kinesiology
Rochester, MN, USA College of Agriculture, Health and Natural
Resources
Lucas P. Santos, MSc University of Connecticut
Cardiology and Cardiovascular Sciences Storrs, CT, and
Faculty of Medical Sciences Department of Preventive Cardiology
Federal University of Rio Grande do Sul Hartford Hospital
Porto Alegre, Brazil Hartford, CT, USA

Mark A. Sarzynski, PhD Martine Thomis, PhD

Department of Exercise Science KU Leuven
Arnold School of Public Health Faculty of Movement and Rehabilitation
University of South Carolina Sciences
Columbia, SC, USA Department of Movement Sciences Physical
Activity
Emily E. Schmitt, PhD Sports & Health Research Group
Assistant Professor Tervuursevest Leuven, Belgium
Division of Kinesiology and Health
College of Health Sciences Ryan T. Tierney, PhD
University of Wyoming Temple University
Laramie, WY, USA Department of Kinesiology
Philadelphia, PA, USA
Jaime Schultz, PhD
Department of Kinesiology
Pennsylvania State University
State College, PA, USA

xix
xx

List of contributors

Matthijs D. van der Zee, MSc Guan Wang, PhD

PhD candidate, Department of Biological Collaborating Centre of Sports Medicine
Psychology University of Brighton
Human Behavioral and Movement Sciences Eastbourne, UK
Vrije Universiteit Amsterdam
Amsterdam, Netherlands James T. Webber, AS, BS, MA
School of Anthropology
Heather L. Vellers, PhD, RCEP University of Arizona
Assistant Professor of Exercise Physiology Tucson, AZ, USA
Department of Kinesiology and Sport
Management David K. Wiggins, PhD
Texas Tech University George Mason University
Lubbock, TX, USA Fairfax,VA, USA

Andrew C. Venezia, PhD Daniel J. Wilkinson, PhD

Department of Exercise Science and Sport MRC-​ARUK Centre for Musculoskeletal
The University of Scranton Ageing Research & NIHR Nottingham
Scranton, PA, USA BRC University of Nottingham, School of
Medicine
Royal Derby Hospital
Jakob L. Vingren, PhD
Derby, UK
Department of Kinesiology
Health Promotion and Recreation
Alun G. Williams, PhD
University of North Texas
Sports Genomics Laboratory
Denton, TX, USA
Manchester Metropolitan University
Manchester, UK
Sarah Voisin, PhD
Institute for Health and Sport
Matthew T. Wittbrodt, PhD
Victoria University
Exercise Physiology Laboratory
Melbourne, Australia
School of Biological Sciences
Georgia Institute of Technology
Atlanta, GA, USA

xx
xxi

FOREWORD
David Epstein

Since writing The Sports Gene, a book about genetics and athleticism for a lay audience, I’ve
had the pleasure of discussing genes and exercise with a broader swath of humanity than I ever
expected. A theme emerged in those conversations: every important aspect of an individual’s
personality and all of their proud accomplishments, they are certain, are purely functions of their
unalloyed free will and expert decision-​making; whenever they get really sick, that is purely
genetic.
Interestingly, human intuition might more closely approximate the opposite of the truth.
Evidence for important genetic contributions to personality is so voluminous that it is encoded
in the so-​called first law of behavioral genetics: “All human behavioral traits are heritable.” That
is not strictly true –​if a Japanese couple moves to England and raises a daughter there, she will
speak English but may inherit no Japanese –​but it isn’t far off. Meanwhile, we are facing an
epidemic of common diseases –​metabolic syndrome, obesity, diabetes –​that were hardly even
on the radar a few generations ago.
Sixty years ago, the number of Americans with diabetes was about equal to the population
of Detroit. Today, that number is equal to the population of the 15 largest cities in the country
combined. Throw in the number of Americans with prediabetes, and you’d have the 14th lar-
gest country in the world. To belabor the point, 1.3% of the global population is made up of
Americans who are diabetic or likely to be diabetic within 5 years. And the US is merely leading
the charge. A peloton of other countries is giving game chase. Needless to say, nothing changed
in humanity’s gene pool overnight to make so many people sick. Rather, our environment
changed. Evolution did not prepare us for such rapid social change. The combination of super-
market culture and the knowledge economy have made for the perfect maladaptive storm: a
practically limitless supply of food for a population of desk jockeys.
I was mulling over exactly this issue in my Brooklyn apartment in 2012 when Hurricane
Sandy swept through. I decided to pass the time (until the electricity went out) watching the
Disney film WALL-​E, which one of the editors of this book (JTL) had recommended to me.
(All in the name of science, of course.) It’s a feel-​good movie, to be sure, but I found part of
it more harrowing than the hurricane. Smuggled inside a lovable Disney cartoon is a Philip
K. Dick-​esque sci-​fi premonition. In the film, humans live aboard a spaceship, which they move
through in what look like hovering La-​Z-​Boys. They need not even unglue their faces from

xxi
xxi

Foreword

hovering screens to summon robots with jumbo beverages. The bulbous space citizens have let
their lower limbs atrophy, so when one falls out of his chair, an emergency message alerts him
that he should not attempt to move. The message of modern society hasn’t been quite so frank,
but the results are trending in the same direction.
WALL-​E crystallized an important question for me: Does the fact that the most pressing
health concerns facing society are due to environmental change mean that there is no sense in
bringing our ingenuity to bear to investigate the genetics of physical activity, and inactivity? In
my opinion, absolutely not. The public is inundated with “nature versus nurture” headlines, but
scientists know that without both genes and environments, there are no outcomes. Neither one
can be understood in isolation.
Consider height, as easily measurable and intuitively heritable a trait as exists. In developed-​
world populations, variance in height between adults is typically 80–​90% heritable. Nobody
would argue that height does not have an important genetic component. And yet, visit a
developing nation and you can see that height need not run in families. Ultimately, the work of
scientists studying exercise genetics will hopefully help society learn how to ensure that physical
inactivity need not run in families. Perhaps that will occur through a deeper understanding of
the genetic mechanisms that regulate physical activity and how that knowledge can be used to
create the most effective environmental interventions. That is, after all, the primary reason for
studying genetics related to health and performance: to figure out how best to change the envir-
onment. Writing about genetics, I have frequently been asked (usually by fellow journalists and
social scientists) why anyone should bother studying genetics since an individual cannot change
their genes. I think the answer is simple: we need to understand what differences between
people are real –​and not intuition or folklore; which of those are important for outcomes we
care about; and, finally, how to use that knowledge to get optimal outcomes for every individual.
Someday, perhaps, we can even go beyond tailoring environmental interventions. With
developments such as the gene-​editing technology CRISPR, there is, at least theoretically, the pos-
sibility of tailoring the genome.Alas, as anyone who has been paying attention to the last 15 years of
research has noticed, the field might as well adopt a motto that Laura Hercher, director of research
in human genetics at Sarah Lawrence College, NY, uses as her Twitter bio: Genetics. It turned out to be
more complicated than we thought. In 2015, a team of psychologists and economists proposed a fourth
law of behavioral genetics. (There were three already.) Their proposed law stated: “A typical human
behavioral trait is associated with very many genetic variants, each of which accounts for a very
small percentage of the behavioral variability.” A corollary, it would seem, is that it will be extremely
hard ever to use gene editing technology to shape a complex trait, because there are so many genetic
variants contributing (not to mention nongene components of DNA, and other factors).
Except, several case studies in sports genetics have demonstrated how a single, powerful
mutation can completely alter a phenotype. While reporting The Sports Gene, I tracked down
Eero Mäntyranta, then an elderly but vigorous reindeer farmer in the Arctic region of Finland.
In the 1960s, he was one of the best endurance athletes in the world –​a seven-​time Olympic
medalist in cross-​country skiing who won the 15 km in 1964 by a margin unequaled before or
since. Twenty years after he retired, Finnish geneticists identified a mutation in his erythropoi-
etin receptor gene (EPOR) that caused a massive overproduction of red blood cells. Another
member of the family with the mutation was also an Olympic gold medalist, and a third was a
world junior champion. Nobody in the family without the mutation was an accomplished ski
racer. More recently, I was contacted by a self-​taught amateur geneticist with two rare genetic
diseases who thought she spotted one in an Olympic medalist sprinter. Together, we convinced
the sprinter, Priscilla Lopes-​Schliep, to take a genetic test, and it turned out she had an LMNA
gene mutation that caused her to have lipodystrophy coupled with extreme muscle growth.

xxii
xxi

Foreword

Finally, there is the so-​called German Superbaby, with two mutated myostatin genes. Myostatin
is a topic covered later in this book. As a teaser, Superbaby’s mother is the only adult I’m aware
of in the scientific literature with a single myostatin mutation, and she was a professional sprinter.
Any time a single mutation can have such a large effect, it is a potential gene-​editing target.
That also means that it is a potential gene-​doping target. It might seem silly to contemplate
gene doping when traditional methods of doping are so effective, but if sports history has
proven anything, it is that whatever it is and whether it works or not, someone is going to try
it. (And that someone is going to make big news and influence public perception if they get
caught.) I’m glad to see that gene doping is one of the topics in this book, because it is one
segment of an important discussion about gene editing and gene therapy in general. One of
my favorite things about sports is that they provide a unique entry point into thorny ethical
discussions that extend well beyond the field, to all of society. And this is an ethical discussion
we need to have, sooner rather than later. For the most part, though, hopefully society-​wide
genetic alteration is something to be concerned about only in the longer term.
Right now, I hope, we can focus on accruing knowledge that will help make for impactful
environmental interventions. If the health benefits of physical activity came in pill form, I think
it’s safe to say there is not a single healthcare provider who would not prescribe it. It is hard
to imagine a body of evidence for anything that is more compelling than the evidence for the
virtues of exercise. And yet, that evidence isn’t having the widespread impact that, say, informa-
tion about the harms of smoking has had. One need only look at cancer statistics to see that in
the latter half of the 20th century, deaths related to smoking swamped all other forms of cancer.
There is a 20-​year lag between smoking rates and cancer deaths, and the current decline in
deaths is not due to a med-​tech victory in the “War on Cancer,” but rather it is in large part the
result of successful public health messaging about tobacco. Today, we are reaping the rewards of
a decline in smoking rates that began decades ago. Unfortunately, messaging about the benefits
of exercise appears to be less effective than warnings about the dangers of smoking. It suggests
to me that addiction to inactivity in our society is more complex and stubborn than cigarette
addiction.
That is an important reason for a systems approach to this research, and to a volume like
this one. As scientists are increasingly forced to hyperspecialize, they necessarily tend to look at
smaller pieces of a large puzzle. A cardiologist used to be the height of medical specialization,
and now there are cardiologists who study only cardiac valves, with the muscle and electricity
of the heart outside their purview.That is necessary for advancing the frontier of knowledge, but
entails the risk of seeing the entire world through cardiac valve-​colored glasses. Scientists run
the risk of ending up like the ancient people Tolstoy wrote of who see only the figurehead on
the prow of an approaching vessel and assume that it must be the motive force behind the ship.
A reference book like this, which puts an incredible diversity of background information (and
the associated citations) in one place, is a gift to the field. (Not to mention to curious layfolk,
like the author of this foreword.)
It also ensures that we won’t mistake the figurehead on the front of the ship, or any other
single component, for the only important piece driving a complex phenomenon. That is
important, because the work discussed in this book is ultimately about the most difficult and,
I think, most important public health challenge in the developed world. Let this book be a call
to interdisciplinary-​research arms for an absolutely critical area of science.

xxiii
xxvi
xxv

PREFACE

As Monica, Steve, and I worked on this volume over the past 2 years, we all feel privileged to be
a part of the growing foundation of sport and exercise systems genetics. All three of us, through
luck, good timing, and/​or inspiration, were fortunate to get into this field during its early days,
when we were all primarily doing applied exercise physiology. The optimism that surrounded
the general field of genetics at that time, as well as the belief that genetics would help us solve
intractable health problems, led all of us to paddle hard enough to catch the front of the exercise
genetics wave. Since that time, we have been fortunate to have many publications regarding the
role of genetics and genomics in exercise as well as fortunate to have held funding from national
granting agencies to investigate the role of genetics and genomics in determining exercise or
activity responses. All three of us have also been called upon many times to be both scientific
reviewers as well as associate editors dealing specifically with genetics articles submitted to
national scientific journals. Through these experiences, we have had the opportunity to witness
and be a part of the explosion over the past 25 years in knowledge regarding the genetics and
genomics of sport and exercise.
While one of us (SR) had previously published books in this area, all three of us had separ-
ately realized that the amount of people wanting to work in the field was outstripping the avail-
able reference material.There are fine volumes available, but these are all several years old and, in
this field, knowledge is moving quickly which outdates books rapidly. Additionally, all three of
us have given numerous talks about specific areas within the field and during these talks we have
found much curiosity, interest, and desire from colleagues to bring their efforts and expertise
to the communal collaborative space that we are calling sport and exercise systems genetics. As
such, there has always been a question about what to read to acquire a good background in the
field and other than the few dated volumes available, there has been little to suggest.Thus, when
we were approached by Routledge about developing this book, we jumped at the chance to
help establish a reference manual in this fast-​moving field; one that we acknowledge will at best
serve as a marker of the available knowledge at this time and one that will require updates and
revisions in the future to stay current.
Certainly, another aspect of this project which attracted us was the opportunity to bring
together the thoughts of the finest minds in the world in these areas –​scientists and writers
who have contributed significantly to scientific thought in these areas, as well as new and

xxv
xxvi

Preface

upcoming scientists who are already making an impact. A perusal of the author list will show
you that we are welcoming some bright newcomers to the field –​e.g., Dr. David Ferguson at
Michigan State University who has provided the only global proteomic scans available of the
genetic mechanisms associated with physical activity as well as a novel method of transiently
silencing genes in healthy animals –​as well as hearing from established authors in the field,
such as Dr. Claude Bouchard from Pennington Biomedical Center, who with over 1100 peer-​
reviewed articles to his name, really blazed the path for this field. These two examples are just
representatives of the bright and curious minds that agreed to populate this volume with their
knowledge and we profusely thank them and all of those that have helped write the individual
chapters.
As editors, we want to thank our original Routledge development editor William Bailey
who initiated and believed in this project, as well as Rebecca Connor who got us through the
contract stages and all the Routledge internal processes. On our side, we want to thank Katie
Feltman from the American College of Sports Medicine who gave us invaluable advice in the
early stages of the proposal and contract development. As we drew closer to the end, as happens
at times, both William Bailey and Rebecca Connor moved on to other projects and the book
was ably picked up and carried over the goal line by David Varley and Megan Smith whom we
thank both for their guidance, patience, and multiple conversations regarding “discoverability”
of book chapters in this age of online digital literature discovery.
Lastly, as an editorial group, we want to thank you the reader for taking the time to read this
book. We hope that you’ll share it with your colleagues and make suggestions to us on how to
improve future editions. What topics to include in this type of volume is a question that only
those in the field –​or those coming into the field –​can help us with. In the end, this volume
and future editions will only be stronger if the community helps us develop it.
As individuals, we each have many people to thank for their support as we developed this
volume and it is only correct that we recognize those folks at the beginning of this volume:

• From Tim: I must thank my two co-​editors Steve and Monica for keeping us on
track with this book as well as all my genetics colleagues over the years who have
patiently “tutored” me in the fine and sometimes not-​so-​fine points of genetics.Those
colleagues include people like Drs. Steve Kleeberger, Larry Leamey, Daniel Pomp, and
Ted Garland, Jr. Further, a huge thanks goes out to all my colleagues and students that
have journeyed with me on all our “genetics adventures” including Drs. Mike Turner,
Reuben Howden, Robin Fuchs-​Young, Weston Porter, and Mike Massett. Lastly, the
most thanks go to my wonderful and brilliant wife Faith who has had to put up with
me babbling about SNPs and proteins at the dinner table for years; her support and
encouragement is one of the pillars of my life (and my writing!).
• From Monica: It seems like just yesterday that my PhD mentor, Dr. Priscilla Clarkson,
wandered into my tiny graduate student office and asked me if I wanted to play with
some new research “toys” called muscle GeneChips that were being used in Dr. Eric
Hoffman’s laboratory. Never one to resist new toys, that day led me down a path
(including a postdoctoral stint in the Hoffman laboratory and time as a junior faculty
in his department) that balanced my background in applied muscle physiology with
systems biology and genetics. As Tim mentioned earlier, the timing of getting to work
within this emerging field of molecular exercise physiology brought me so many
exciting opportunities, for which I am eternally thankful. For those, I have to thank
Drs. Clarkson and Hoffman, along with other colleagues and collaborators, including
Drs. Bill Kraus, Paul Thompson, and Joe Houmard, among many others, along with

xxvi
 xxvi
newgenprepdf

Preface

my fellow trainees and, later, my mentees. Finally, it has been a great pleasure working
with Tim and Steve on this project and I appreciate being a part of such a great col-
laborative project.
• From Steve: I am indebted to the many colleagues I’ve had over the years, including
several in this volume, who have helped push my work to be better. A few names in
particular stand out, including Drs. Jim Hagberg, Robert Ferrell, and Ben Hurley,
without whom I doubt very much genetics would have been a landing place for me.
And years before that, Drs. Brian Sharkey and Brent Ruby sparked my interest in the
field of exercise physiology as an undergraduate at the University of Montana. My
students and technicians were the folks who pushed the science day-​to-​day and I’m
grateful to be able to watch their successes and hope I played a part in the good work
they are now doing. I’ve enjoyed my time working on this volume with Tim and
Monica and thank them very much for their efforts.

xxvii
xxvi
1

INTRODUCTION
J. Timothy Lightfoot, Monica J. Hubal, and Stephen M. Roth

There is little doubt that humans have entered the “genetic” age, where the effort to under-
stand and alter humanity’s basic underlying code has become foremost in the minds of many
scientists while many media outlets have declared the “21st century will be the age of genetics”
(e.g., 2). The advances made in sequencing the human genome in the early 2000s –​as well as
other species’ genomes –​promised “to help convert this growing knowledge into treatments
that can lengthen and enrich lives” (1). Since the early 2000s, information regarding pieces of
the “genetic story” have continued to be revealed with continued hope that precision medi-
cine –​or treatments tailored to each of us based on our underlying genetic code –​will become
commonplace.
The idea that our underlying genetic code plays a role in sport and exercise performance is
not a new one but, with access to a growing number of molecular tools associated with genetics,
has become of greater interest and the topic of a burgeoning new discipline within exercise
science. This text is meant to serve as a handbook –​a reference marker if you like –​for where
this burgeoning discipline is today. As such, this reference marker is a foundation that includes
most of what is known about the genetics of exercise and sports, written by the scientists who
have discovered and developed much of that foundation over the past 40 years.
It may be logically asked, why use the term “systems genetics” to describe the various gen-
etically related mechanisms involved in exercise characteristics? Our use of the term “systems
genetics” is one indication of the tremendous changes that have occurred over the past 20 years
in the field traditionally called “genetics.” Formerly, “genetics” was used to refer to any factor/​
mechanism that was related to the heritability and function/​structure of inheritance mechanisms.
With the tremendous expansion of knowledge in this area, the term “genetics” is now more
often used to refer to the action and mechanisms of single genes, while a newer term –​“gen-
omics” –​is used in relation to the action and mechanisms of several genes in combination (3).
In discussing whether to use “genetics” or “genomics” to describe the general content of this
book, we believed that we needed to use a broader term that encompassed all mechanisms that
arose from genetic and genomic pathways, as well as those mechanisms that were influenced
by potential interactions between genetic, biological, and environmental factors. When we
consulted with three geneticists (yes, they still call themselves that), each with extensive publi-
cation records regarding this issue, all three basically advised that we could use whatever term

1
2

J.T. Lightfoot et al.

we thought was most appropriate. Thus, to encompass the diversity that is present in the field
today, we choose to use the term “systems genetics” to indicate all mechanisms, processes, and
interactions that influence the heritable characteristics of an organism. Applying this description
to exercise and sport resulted in “sport and exercise systems genetics” which we define as “all
heritable mechanisms, processes, and interactions that impact the capability and performance of
sport and exercise.” We believe that after reading this book, you will understand why it is well
accepted that most exercise and sport characteristics (i.e., phenotypes) are influenced by not
only classical genetic mechanisms, but also interactions among these mechanisms, other bio-
logical systems, and what we know today as epigenetic influences.
As science in systems genetics has advanced, the media has warmed to the topic, resulting
in more awareness by the public of news with genetic findings. Most of us are familiar with
the increasing number of “Scientists Find Gene that Causes [Insert Disease/​Characteristics Here]”
types of articles that have been written since 2000, many of which may not be valid. Indeed,
most scientists who work in the genetics world would immediately spot at least two problems
with the hypothetical headline just mentioned. However, mistaken concepts aside, the public
awareness of advances in genetics has caused an increase in awareness of the possibility of gen-
etic involvement in sport and exercise performance among many exercise professionals. Indeed,
there is a whole sub-​field concerned with using genetic results to predict sport capability in
athletes, especially child athletes (Drs.Venezia and Roth review this issue in Chapter 31). Thus,
as research in genetics has grown, so has the interest in applying research findings to exercise
and sports, as well as the interest from commercial companies in profiting from this knowledge.
The growing research interest in sport and exercise systems genetics, from our observa-
tion, has led two types of science professionals to come into the field. One is the classically
trained geneticist who views sport and exercise as interesting physiological and psychological
perturbations to which s/​he can apply their systems genetics knowledge. The other type of
science professional, of which the three editors are all examples of, are traditionally trained
exercise physiologists who see systems genetics as a way to understand the variation in response
observed in humans engaged in acute or chronic exercise. This book targets both populations
of science professionals to provide a review of the current knowledge on what is known about
systems genetics across various areas of exercise and sport. Due to the dynamic nature of both
systems genetics and exercise science, we anticipate that this volume, rather than being a static
foundational reference manual, will also grow and change through future editions.
The structure of the book revolves around separate topical sections, each of which comprises
multiple chapters. Four of these topical sections form the core of the book by focusing on gen-
etic findings in the common forms of sport and exercise. These four core sections are prefaced
by a section on general systems genetics concepts and capped off by a section on the ethics
of employing systems genetics in exercise and sport settings. The four core sport and exer-
cise sections address physical activity, exercise endurance and trainability, muscle strength and
trainability, and sport performance. Within these sections, the reader will find a diversity of
approaches and methodologies that have been used to explore not only whether there are gen-
etic influences on these core areas, but also what genetic mechanisms regulate these core areas.
While these four core sections comprise the crux of the sport-​and exercise-​related findings, the
beginning and ending sections are no less important, especially in the context of today’s systems
genetics climate.The opening section deals with general systems genetics concepts and serves as
a conceptual foundation for the rest of the book, as well as providing a succinct overview of the
tremendous strides in systems genetics since 2000. Further, the concluding section deals with
the ethics of applying these genetic findings to sport and performance. These types of ethical

2
3

Introduction

considerations are critical as we face –​as practitioners –​questions regarding talent selection,
gene doping, and even gene editing.
To begin the book, we are honored to have a Foreword which sets the stage for why systems
genetics in sports and exercise is important, written by Mr. David Epstein who –​to our know-
ledge –​is the only author to have written a sports genetics book that appeared on the New York
Times best-​seller list. After the six primary sections of the book, we have a closing section that
includes a “last words chapter” as well as a conclusion chapter. With this last section, we have
attempted to set the stage for future research in the field.The “last words chapter” was graciously
written by Dr. Claude Bouchard, who is easily the most published and cited scientist in our
field and whom many of us consider the “godfather” of exercise genetics. All in all, we believe
that the structure of the book provides the most comprehensive coverage of this field to date
and we are so thankful for all the authors who agreed to participate.
One of the challenges in writing about a relatively new area of scientific investigation –​and
especially a field that is at the intersection of two dynamic and rapidly changing disciplines –​is
that there are many more questions than there are good data and studies available to answer
those questions. As such, one of the limitations of this book is that there are several areas where
some existing science is available, but by no means a substantial enough base to provide solid
conclusions at this time. We believe the general topics falling into this category currently are:

• the role of epigenetics in determining exercise responses;

• the role of miRNA –​both circulating and tissue based –​in regulating exercise
responses;
• the role of genetics in determining the limits of performance; and
• the role of genetics in modulating performance psychology.

This list is by no means all inclusive, but represents a sampling of the topics about which
we kept saying, “Wow, it’d be great if there was enough information on this topic for a
chapter!” We do believe that there will be enough information in these areas in the future to
address them in revised editions and would welcome scholars willing to join the author team
and accept the challenge of describing those topics!
One area that we wanted to approach in this volume, but could not find any authors willing
to tackle the topic (even after almost 9 months of the three of us searching), regarded the tech-
nical aspects of gene therapy in relation to the potential for gene doping. Given the tremendous
interest and research with CRISPR-​Cas9 systems, as well as existing work with in vivo gene
silencing techniques, there have been suggestions that the use of these technologies in an exer-
cise/​sports environment may be imminent and even appropriate (see Chapters 32 and 33 in
this volume). Some of these techniques are applied in Chapter 9 in this volume, but the uses
described are purely from an animal hypothesis-​testing approach. We hope that future volumes
of this handbook will address this topic more fully because shining light on the topic is one of
the few ways to prevent use and abuse of these technologies in the exercise and sports arena.
As we have worked on this volume, the new knowledge that is becoming apparent through
the genetic approaches to exercise is quite exciting. Sport and exercise systems genetics
encompass two disciplines that are individually rapidly growing and adding new theories and
perspectives to their canon, while at the same time, intersecting to add substantially to each
other. As such, we are proud to present this volume that summarizes where the field is at this
moment. Perhaps more importantly, we are most excited to have the following chapters give us
a benchmark so we can look forward to the discoveries that lie ahead.

3
4

J.T. Lightfoot et al.

References
1. Collins FS. Remarks at the Press Conference Announcing Sequencing and Analysis of the Human
Genome. February 12, 2001. www.genome.gov/​10001379/​february-​2001-​working-​draft-​of-​
human-​genome-​director-​collins/​ [Accessed December 18, 2018.]
2. Editorial Board. March of science: the 21st century will be the age of genetics. Independent
November 6, 2013.
3. World Health Organization. WHO definitions of genetics and genomics. www.who.int/​gen-
omics/​geneticsVSgenomics/​en/​. [Accessed January 1, 2018.]

4
5

SECTION 1

General systems genetics

Introduction
With the completion of the first human genome sequence in 2000, scientific and public interest
increased in regard to the impact of genetics on a wide variety of physiological traits. Sport and
exercise traits were no exception, with many in the public and in science questioning the role
that “genetics” played in both sport and exercise performance. However, because most sport and
exercise scientists had/​have little in the way of genetics background and training, catching up
and keeping up with the rapid changes occurring in genetics continues to be a major challenge
for the application of genetics to sport and exercise questions. Thus, the following four chapters
aim to provide a foundation for the systems genetics concepts and techniques that are described
and discussed in the remainder of the book.
As it should be with a general foundational section, the set of chapters that make up this
section is wide ranging with each written by an acknowledged international leader in the field.
The first chapter, written by Dr. Frank Booth and co-​workers, encompasses a broad overview
of why considering systems genetics is important, with the consistent context of the chapter
focused on the personal and health cost of physical inactivity.
Maybe more than any other scientific discipline, the conceptual and theoretical framework
of genetics has undergone massive changes in the past 20 years. Much of the genetics many of
us learned before 2000 is obsolete and so the second chapter, by Dr. Penny Riggs, is a review
of the major changes in genetics since the sequencing of the human genome, with an emphasis
on the new concepts such as miRNA and epigenetics that can and should be applied to exer-
cise and sport.
Physiologists do not often consider how the individual’s genetic makeup could alter the
responses that are observed during experiments, when in fact the genetic framework of an
individual (or animal) can lead to marked differences during an experiment. Therefore, appro-
priate modeling systems are critical in genetic studies, so Chapter 3 provides an overview of the
common human modeling systems used in genetic studies by Dr. Jaakko Kaprio.
Lastly, understanding the mechanisms and modeling are important, but most scientists agree
that if we cannot apply the results to a population in general, then our science has difficulty
making a difference. That is why the section finishes with a critical chapter by Dr. Molly Bray

5
6

General systems genetics

dedicated to how we can apply systems genetics to translational models and everyday life, with
Dr. Bray using physical activity as the exercise phenotype described.
We believe that these four chapters will provide a great foundation for the rest of the book.
You will find that the basic concepts covered in these chapters will be returned to repeatedly
throughout the rest of the book as specific exercise and sport situations are dealt with. Due to
the foundational nature of these chapters, we are grateful that these scientific thought leaders
all agreed to participate. Enjoy!

6
7

1
WHY STUDY THE SYSTEMS
GENETICS OF SPORT AND
EXERCISE?
Frank W. Booth, Taylor J. Kelty, Kolter B. Grigsby,
and Gregory N. Ruegsegger

Introduction
This chapter is tilted toward the authors’ perspective of the future of exercise science in contrast
to our recent reviews (8, 9, 44) which were more focused on the retrospective side of exercise
mechanisms. Both types of reviews can be controversial. However, perspective reviews distin-
guish themselves in that they are based more upon an opinion of the future, and less on data
from the literature. As such, the future being unpredictable, the authors’ comments could be
considered to provide backgrounds for future research considerations as well as the rest of the
chapters of this volume.

Exercise systems genomics

This section bookends the final contributed chapter of the book, which is written by the
founder of the field of exercise genomics, Claude Bouchard. Bouchard’s tremendous foresights
allowed him to establish exercise systems genetics in humans three decades before systems gen-
etics was recognized as a field. Bouchard began phenotyping exercise as early as 1976 (14), when
he correlated variables with skeletal maturity and chronological age in 8–​18-​year-​old boys.
A 1980 paper (12) was about estimates of sibling correlations and genetic/​heritability estimates
in a number of French-​Canadian families which was extended in 1983 by an overall review of
the genetics of physiological fitness (13). Furthermore, Bouchard pioneered the integration of
exercise training and genetics in his classical Heritage family study (10, 11). In 2007, the term
“systems genomics” began being popularized (35) to include Bouchard’s older science of “exer-
cise genomics.” In 1999, he already had performed sufficient studies to conclude that trainability
of VO2max had a significant genetic component based upon considerable heterogeneity in
degree of increase in VO2max following the same endurance-​training program in 481 sedentary,
adult Caucasians from 98 two-​generation families. With the establishment of exercise genomics
and exercise systems genomics, a wide variety of exercise characteristics can be viewed through
the lens of exercise systems genetics.
One example of exercise systems genetics is that contracting skeletal muscle releases
peptides (termed myokines (22, 39)) into capillaries. The myokines then circulate to a distant
site (i.e., cells, tissues, or organs) where they initiate a cascade to have a subsequent biochemical

7
8

F.W. Booth et al.

effect at that site (25, 37). The first investigator to identify a molecule that fitted the criteria
to have an endocrine-​like function from skeletal muscle was Pedersen, whose lab reported in
1994 that blood interleukin (IL)-​6 increased during exercise (52). Then, in 2003, Pedersen
proposed, “IL-​6 and other cytokines, which are produced and released by skeletal muscles,
exerting their effects in other organs of the body, should be named ‘myokines’ ” (38). IL-​6 is
now the prototype myokine emulating skeletal muscle-​to-​organ crosstalk, or systems genetics.
Recently, Whitham et al. (54) reported that many circulating proteins are packaged within
extracellular vesicles that also perform tissue crosstalk during exercise to produce systemic bio-
logical effects. Our notion is that components in the extracellular vesicles could form another
candidate for the therapeutic manipulation of some of the thousands of positive molecular
benefits of exercise.
The challenge of understanding the above example was reflected in a 2017 opinion piece
where 16 invited experts presented their viewpoints on the status and future of systems genetics
(5). The experts’ consensus opinion was: “deciphering genotype to phenotype relationships is a
central challenge of systems genetics and will require understanding how networks and higher-​
order properties of biological systems underlie complex traits” (5). The authors’ interpretation
of systems genetics in relationship to physical activity and physical inactivity will be considered
in the next section.

Exercise presents a high stress; however, the body is able to survive by

maintaining its homeostasis during performance
Exercise is stress. Stress has numerous definitions. The operational definition for stress used
herein is an increased blood cortisol concentration. As such, maximal exercise effort is one
of the greatest stressors to the body. An underlying reason for intense exercise, and thus
intense stress, was postulated by Charles Darwin. On a statistical basis, those who are phys-
ically unfit are more likely to be eliminated prior to their being able to reach reproductive
age. Herbert Spencer coined the phrase “survival of the fittest.” It is unlikely that the pre-​
adolescent individual of 5000 years ago, who never had the capability to perform physical
activity, could have survived to puberty to perpetuate their genes to the next generation.
Thus, their gene pool would be extinguished and not passed on to the next generation (21).
Today, the elimination of infectious diseases allows most individuals in developed com-
munities to live to puberty and have the ability to pass their genes to the next generation,
while having less than the maximal fitness of their distinct ancestors. Thus, the notion can
be raised whether “survival of the fitness” by natural selection to infectious organisms has
been bypassed, allowing individual survival to pass a “less fit” gene pool to stress on to the
next generation. Simultaneously, during the same era, we have the notion that children and
adolescents who are physically playing less will have a lower development of lifetime phys-
ical fitness, which could mean that they are passing epigenetic changes for weaker maximal
aerobic and strength fitness to their offspring. The lifetime consequence of such changes in
physical fitness would become apparent when your friends must enter assisted care living
while you do not. Along this line, we previously reviewed in greater detail our interpretation
of exercise and Darwin’s writings (7).

Exercise systems genetics versus inactivity systems genetics

Exercise systems genetics identifies the genes that: 1) prevent chronic disease, 2) treat for loss
of function due to lack of sufficient exercise, and 3) increase sports performance. For definition

8
9

Systems genetics of sport and exercise

purposes related to the previous sentence, a National Institutes of Health (NIH) website (2)
defines the Common Fund $170-​million research program:

The Molecular Transducers of Physical Activity in Humans’ effort to uncover the

molecular changes that occur in response to movement is expected to transform clin-
ical medicine’s use of physical activity as a treatment and preventive strategy. Although
researchers have demonstrated that physical activity is good for us in many ways, they
know little about the molecules that cause these improvements.

We interpret the mission of this NIH Common Fund Program to be to look for molecules
that cause the “treatment and preventive strategies” of physical activity. In this sense, the
NIH Program has made an important distinction. Stein and Bollister (47) in 2006 made
the statement, “Anabolic and catabolic pathways are usually separate.” They compared two
lists of mRNAs in papers by Lecker et al. (31) and by Fluck (23) that showed “virtually no
overlap” in skeletal muscle mRNAs for atrophy and for recovery from atrophy, respectively.
Taken together, future research will need to recognize that gene changes during exercise
training may not be identical to gene changes when athletes and those who are physically
active “take time off ” to recover from “the season,” recover from injury, or switch sports
where some genes may play a role in detraining while others play a role in training. An
example of detraining while continuing to train is that VO2max fell for 22 years between the
mid 20s and later 40s in men whose training volume declined over 22 years (50). However,
the two men in the same study who were able to maintain their exercise volume unchanged
over the 22 years had no decline in their VO2max (50). In an animal model, the Booth lab
showed that lifetime VO2peak occurred at 19 weeks of age in female rats (45, 49), which
in the same rats was independent from their lifetime peak in voluntary running distance
that happened at their younger age of 8–​9 weeks, implying the mechanisms for these two
lifetime events could be different. In humans, a study of 23 females and males who trained
one leg for 3 months, detrained for 9 months, and then retrained both legs for 3 months
(33) found 3404 gene isoforms in the vastus lateralis muscle, mainly associated with oxida-
tive function, after the first training. All training changes in these mRNAs were absent after
9 months of detraining. The experiment’s investigators concluded, “No coherent evidence
of a skeletal muscle transcriptome memory was observed, even though there were some data
indicating a training-​induced memory mechanism … These findings show the great need
for highly controlled studies of repeated exercise training.”

Physical inactivity is a necessary reference group to show healthy exercise

effects –​and turns out to be a very unhealthy comparison!
The effect of exercise can be better elucidated by exercise in the absence of gravity. Weight-​
bearing exercise is necessary to prevent bone loss. Nongravity (“lack of pounding of bones
upon impact”) exercise such as cycling is associated with a seven times increased likelihood to
have spine osteopenia than running (41). Weightlessness of spaceflight causes 10 years of bone
atrophy to occur in 1 year (53). The cardiovascular system experiences a tenfold faster loss of
functional capacity in space (53). In comparison with ground-​based models of weightlessness,
the decline of 27% in VO2max when aging from 20 to 60 years old was comparable with the
26% loss that the same men experienced after 20 days of continuous (24 hours/​day) bed rest
at the age of 20 (34). The above examples of short-​term (days to 1 year) drastic reductions in
physical activity leading to dramatic losses of physiological function provide insight into subtle

9
10

F.W. Booth et al.

reductions in physical activity over decades. Failure to have more than 150 minutes of weekly
exercise over periods of decades leads to the increased risk of 40 chronic diseases and a prema-
ture end to life (44).

A biased positive look to the future

One of the charges to the authors of this review that came from an editor of this book was to
give a 30,000-​foot view.

A projection into where technology seems to be taking physical inactivity

Until now, most genetic studies have focused on a single gene or gene clusters that dictate a
single trait in a phenotype. However, with technological advances, it is now possible to look
at a phenotype as a dynamic system utilizing the entire genome. Systems genetics integrates
transcriptomics, proteomics, and metabolomics for the different physiological systems to gain a
more complete understanding of a phenotype.
First, let us examine how systems genetics has developed until now. Human skeletal muscle
has over 6000 expressed genes leading to more than 600 secreted proteins (37). Training is able
to alter over 3000 of these gene isoforms (33), producing hundreds of different proteins in skel-
etal muscle (36), leading to over 200 metabolites secreted into the bloodstream (15), all of which
contribute to the rise of a specific phenotype.Therefore, identification of a single athlete-​model
phenotype with traditional single gene methods will prove very difficult, if not impossible.
Conversely, an approach using systems genetics could reproduce the phenomics of the athlete,
leading to the potential of improved exercise performance through gene manipulation.
The benefits of systems genetics do not just benefit champion athletes, but are useful
for the entire population. Studies associate physical inactivity with multiple negative health
consequences (8, 32). Two key approaches can be used to fight physical inactivity: 1) to iden-
tify the underlying genetics that produce a physically inactive phenotype, and 2) to look at the
underlying mechanisms of physical inactivity that contribute to the negative health effects.
Importantly, mechanisms of physical inactivity and physical activity are sometimes different (9).
Applying the holistic approach of systems genetics to physical inactivity could generate novel
gene therapies that either could prevent a physically inactive phenotype from beginning in the
first place, or could prevent the negative health consequences birthed by physical inactivity
(27, 44).

Multiple, biased, concerned looks to the future

New technology is often based upon reducing physical activity
New technologies will continue to remove physical activity from daily living.To counter loss of
physical activity because of new technology, a molecular understanding of both the motivation
to keep active and to counter motivation to be inactive will be needed. One example of a new
technology that will reduce physical activity is artificial intelligence. At one end of the spectrum,
artificial intelligence can scan massive amounts of data and not only give the specific output,
but also store that data to be called on in future problems, giving it a vast potential to revo-
lutionize public health (29). Ironically, however, these advancements in technology could also
render much of the physical work done by the public unnecessary. Studies around the world
have shown that the public and even children are becoming more physically inactive (9, 20, 43),

10
11

Systems genetics of sport and exercise

while technology is advancing and becoming more accessible to the public (26, 40). This leads
to a question to be answered: Will advancing technology (e.g., artificial intelligence) worsen
the pandemic of physical inactivity worldwide, and thus worsen chronic diseases? In summary,
advancements in technology are unlikely to stop so we must find another approach to survive
by maintaining appropriate physical activity levels for a healthy lifestyle. Stated alternatively, a
need exists to stay healthy through physical activity while our lives are getting easier.

Inheritance of low voluntary running behavior

One concern for the future to be expressed is whether physical inactivity will produce adverse
transgenerational epigenetic changes. Stated alternatively, do the offspring of physically inactive
parents inherit some of their parents’ consequential epigenetic changes due to physical inactivity
and therefore remain predisposed to physical inactivity behavior themselves? One example of
precedence to support the viability of such a hypothesis now exists. Rapid, human-​wide, epi-
genetic changes in the last 30 years of the 20th century exist in the form of an increased type 2
diabetes prevalence in Arizonian Pima Indians. In the 10–​14-​year-​old age group of Arizonian
Pima Indians, type 2 diabetes increased in boys in the 1967–​1976 time period from 0% to 1.4%
in the 1987–​1996 time period. In girls, the increase was from 0.72% in 1967–​1976 to 2.88% in
1987–​1996. Dabelea et al. (17), in reference to the above-​mentioned increase in Arizonian Pima
Indians concluded that “a vicious cycle related to an increase in the frequency of exposure to
diabetes in utero appears to be an important feature of this epidemic.”
Arizonian Pima Indians separated geographically from Mexican Pima Indians approximately
300 BC by moving to what is now Arizona (17). Schulz and Chaudhari (46) have reviewed
the lifestyle changes that began in about 1900 in the Arizonian Pima Indians after they were
forced to switch from their Mexican agricultural lifestyle to a Western diet and a sedentary
lifestyle. The change occurred because white settlers diverted water that supported the Pimas’
agriculture, forcing them to eventually move to a US government-​run reservation where they
no longer worked for their food. Schulz and Chaudhari (46) commented on the event, stating
“The timing of this significant change in lifestyle and livelihood coincides with the develop-
ment of diabetes among the Arizona Pimas.” In contrast, the Mexican Pima Indian cohort
remained farming in Mexico, where they did not experience an increase in prevalence in type
2 diabetes during the same time period. Importantly, the increased type 2 diabetes prevalence
occurred within Arizonian Pima Indians during a few generations of experiencing the new
Western lifestyle. Pima Indians have now been shown to be one of the world’s populations most
susceptible to environmentally induced type 2 diabetes (50% in adults).

Does epigenetics play a role in type 2 diabetes and in physical inactivity?

Epigenetics could play a role in the growing incidence of type 2 diabetes, as described in a recent
review paper (18). The above-​mentioned points lead us to the research question of whether
physical inactivity, which was associated with the Arizonian Pima’s development of type 2
diabetes, also caused transgenerational epigenetic changes that accentuates greater behavior of
motivation to be physically inactive, which is unknown.

Low voluntary running animal model to mimic increased human inactivity

To begin to tie the above-​discussed human studies into future generations, animal studies have
the potential to model the state of human inactivity, which arguably occurred when machines

11
12

F.W. Booth et al.

began replacing human physical labor over a century ago. In 2009, we began the selective
breeding of rats for the phenotype of low voluntary running to mimic the loss of human phys-
ical labor/​work. Our experimental design was modeled after the selective breeding models of
Garland (48) and of Britton and Koch (30). The female and the male with the lowest voluntary
running distance from each litter of 13 families were selected to be the breeding pair for the next
generation of breeding using a scheduled order of breeding. Our initial findings in 2012 were
that the newly bred rats voluntarily ran the same distances in each of the first five generations
of selective breeding compared to their wild-​type ancestors. However, litters in the sixth gen-
eration voluntarily ran significantly lower distances than their wild-​type founders (42). In 2018,
we rescued about one-​third of the voluntary running distance lost by selective breeding by a
transient overexpression of the protein kinase inhibitor alpha (PKIA) gene, which was delivered
within an adeno-​associated virus by a localized injection into the nucleus accumbens (24). The
observation supported the existence of genes with low voluntary physical activity functions.
In 2013, den Hoed et al. (19) reported that genes with sedentary functions were present in
the more sedentary pair of 772 pairs of monozygotic and dizygotic twins. The recognition
that genes exist for physical inactivity changes physical inactivity from being classified by the
US government as a “risk factor” for chronic diseases to an “actual cause” of chronic diseases.
Eventually, because genes cause physical inactivity, physical inactivity will need to be considered
as a disease that causes chronic diseases and aging.

The high economic costs of physical inactivity are contributing in their

small way to the bankruptcy of our children’s future cost of living
Why must economics be a vital part of an exercise systems genetics review? To provide
some potential answers to the question, this paragraph provides six sets of facts. First, with
accelerometers as the measurement vehicle, only 1 out of 20 US adults meet the US guidelines
(51). Alternatively stated, 19 out of 20 US adults do not meet the US physical activity guidelines.
Thus, in our opinion, 19 out of 20 US adults are not likely obtaining the health benefits from
exercise. Second, US healthcare costs of 17.9% of gross domestic product are the highest in the
world, compared to a range of 9.6–​12.4% in a select group of the ten highest-​income countries
(6).The US Congressional Budget Office (CBO) data shows that spending on major healthcare
programs which accounts for 28% of federal noninterest spending as of 2017 would increase
to 40% in 2047 (1). The CBO states: “The aging of the population and excess cost growth
are reasons for the sharp rise projected for spending on Social Security and the major federal
healthcare programs over the next 30 years” (1).Third, the US ranks 43rd (4) and 45th (3) in life
expectancy. Fourth, the average US life expectancy decreased for the past 2 years (55) –​is the
lack of exercise a contributor to loss of life expectancy? Fifth, Carlson et al. (16) of the CDC
reported, “Overall, 11.1% (95% CI: 7.3, 14.9) of aggregate, annual, healthcare expenditures were
associated with inadequate physical activity (i.e., inactive and insufficiently active designations).”
Sixth, we used the 11.1% annual cost in the previous sentence from Carlson et al. (16) to esti-
mate the accumulated annual costs of physical inactivity. In 2016, $3.3 trillion was the total
US health costs, or $10,348/​person. Thus taking 11.1% as the fraction of total healthcare cost,
physical inactivity in the US cost $366 billion in 2016. To obtain a better appreciation of the
magnitude of $366 billion, the next comparisons are made: 1) US healthcare costs for phys-
ical inactivity was $1132/​individual; 2) 8 years’ worth of healthcare costs for physical inactivity
would pay for 1 year of total healthcare; and 3) 1 year’s cost of physical inactivity in 2016 would
pay for 9.8 annual NIH budgets at $37.3 billion each.

12
13

Systems genetics of sport and exercise

As we noted in 2012, there is one potential solution to the problem (8). Kersh and Morone
(28) wrote, “Public health crusades are typically built on a scientific base … In any event, med-
ical knowledge in itself is rarely enough to stimulate a political response. Rather, the key to its
impact lies in the policy entrepreneurs who spread the medical findings” (28). Thus, we suggest
that a more impactful method to appeal to the policy entrepreneurs is to explain the impact of
exercise, sport, and physical inactivity on the economy, as we did earlier.

What systems genetics could do for humans

Systems genetics can be appreciated for its potential future molecular explanations regarding
the basis for the biological beauty that underwrites the living machinery in animals and humans
to provide them with a continuous sub-​conscious, homeostatic maintenance of their exist-
ence. Furthermore, regulatory systems invoke physiological reinforcements to allow molecular
adaptations to new environments by lowering homeostatic disruptions. Alternatively phrased
as a question: “What is the ensemble within the orchestra of genes in organs throughout the
organism that permit survival to restrict the stress of exercise within homeostatic limits?” Upon
answering this question, the next challenge will be to determine how to slow the unrelenting
reduction in maximal capacities of physiological systems whose demise is not compatible with
continued life because one or more of the dwindling capacities of organ and system capacities
to maintain homeostasis is exceeded. When we gain a firm grasp on the systems genetics that
regulate both exercise and sport responses, we will be one step closer to answering the afore-
mentioned question.

Acknowledgements: Supported by the University of Missouri (FWB, TJK, KBG, and GNR).

References
1. Congressional Budget Office. The 2017 Long-​Term Budget Outlook.Washington, DC: Congressional
Budget Office, 2017.
2. National Institute of Health. Frequently asked questions. In: Molecular Transducers of Physical
Activity in Humans. Bethesda, MD: National Institute of Health, 2016.
3. World Population Review. Life expectancy by country 2017. www.worldpopulationreview.com/​
countries/​life-​expectancy-​by-​country/​, 2017.
4. Central Intelligence Agency. The World Factbook. Washington, DC: Central Intelligence
Agency, 2018.
5. Baliga NS, Bjorkegren JL, Boeke JD, Boutros M, Crawford NP, Dudley AM, Farber CR,
Jones A, Levey AI, Lusis AJ, Mak HC, Nadeau JH, Noyes MB, Petretto E, Seyfried NT,
Steinmetz LM, and Vonesch SC. The state of systems genetics in 2017. Cell Systems 4: 7–​15, 2017.
6. Bauchner H and Fontanarosa PB. Health care spending in the United States compared with 10
other high-​income countries: What Uwe Reinhardt might have said. JAMA 319: 990–​992, 2018.
7. Booth FW, Chakravarthy MV, and Spangenburg EE. Exercise and gene expression: physio-
logical regulation of the human genome through physical activity. The Journal of Physiology
543: 399–​411, 2002.
8. Booth FW, Roberts CK, and Laye MJ. Lack of exercise is a major cause of chronic diseases.
Comprehensive Physiology 2: 1143–​1211, 2012.
9. Booth FW, Roberts CK, Thyfault JP, Ruegsegger GN, and Toedebusch RG. Role of inactivity
in chronic diseases: evolutionary insight and pathophysiological mechanisms. Physiological Reviews 97:
1351–​1402, 2017.
10. Bouchard C. Genomic predictors of trainability. Experimental Physiology 97: 347–​352, 2012.
11. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, Pérusse L, Leon AS, and Rao
DC. Familial aggregation of VO(2max) response to exercise training: results from the HERITAGE
Family Study. Journal of Applied Physiology (Bethesda, MD: 1985) 87: 1003–​1008, 1999.

13
14

F.W. Booth et al.

12. Bouchard C, Demirjian A, and Mongeau E. Sibling correlations and genetic estimates for selected
blood variables in French Canadian children. Human Genetics 54: 259–​263, 1980.
13. Bouchard C and Malina RM. Genetics of physiological fitness and motor performance. Exercise and
Sport Sciences Reviews 11: 306–​339, 1983.
14. Bouchard C, Malina RM, Hollmann W, and Leblanc C. Relationships between skeletal maturity
and submaximal working capacity in boys 8 to 18 years. Medicine and Science in Sports 8: 186–​190, 1976.
15. Burke MF, Dunbar RL, and Rader DJ. Could exercise metabolomics pave the way for
gymnomimetics? Science Translational Medicine 2: 41–​35, 2010.
16. Carlson SA, Fulton JE, Pratt M, Yang Z, and Adams EK. Inadequate physical activity and health
care expenditures in the United States. Progress in Cardiovascular Diseases 57: 315–​323, 2015.
17. Dabelea D, Hanson RL, Bennett PH, Roumain J, Knowler WC, and Pettitt DJ. Increasing
prevalence of type II diabetes in American Indian children. Diabetologia 41: 904–​910, 1998.
18. Davegardh C, Garcia-​Calzon S, Bacos K, and Ling C. DNA methylation in the pathogenesis of
type 2 diabetes in humans. Molecular Metabolism, 14: 12–​25, 2018.
19. den Hoed M, Brage S, Zhao JH, Westgate K, Nessa A, Ekelund U, Spector TD, Wareham NJ,
and Loos RJ. Heritability of objectively assessed daily physical activity and sedentary behavior. The
American Journal of Clinical Nutrition 98: 1317–​1325, 2013.
20. Duan J, Hu H, Wang G, and Arao T. Study on current levels of physical activity and sedentary
behavior among middle school students in Beijing, China. PloS One 10: e0133544, 2015.
21. Eaton SB and Eaton SB. An evolutionary perspective on human physical activity: Implications
for health. Comparative Biochemistry and Physiology Part A, Molecular & Integrative Physiology
136: 153–​159, 2003.
22. Febbraio MA and Pedersen BK. Contraction-​induced myokine production and release: is skeletal
muscle an endocrine organ? Exercise and Sport Sciences Reviews 33: 114–​119, 2005.
23. Fluck M, Schmutz S, Wittwer M, Hoppeler H, and Desplanches D. Transcriptional repro-
gramming during reloading of atrophied rat soleus muscle. American Journal of Physiology –​Regulatory,
Integrative and Comparative Physiology 289: R4–​R14, 2005.
24. Grigsby KB, Ruegsegger GN, Childs TE, and Booth FW. Overexpression of protein kinase
inhibitor alpha reverses rat low voluntary running behavior. Molecular Neurobiology, 2018. doi: 10.1007/​
s12035-​018-​1171-​0.
25. Hojman P, Gehl J, Christensen JF, and Pedersen BK. Molecular mechanisms linking exercise to
cancer prevention and treatment. Cell Metabolism 27: 10–​21, 2018.
26. Johnson GM. Internet use and child development: The techno-​microsystem. Australian Journal of
Educational and Developmental Psychology 10: 32–​43, 2010.
27. Kelly SA, Villena FP, and Pomp D. The “Omics” of voluntary exercise: Systems approaches to a
complex phenotype. Trends in Endocrinology and Metabolism 26: 673–​675, 2015.
28. Kersh R and Morone J. The politics of obesity: Seven steps to government action. Health Affairs
(Project Hope) 21: 142–​153, 2002.
29. Kiefer AW, Pincus D, Richardson MJ, and Myer GD. Virtual reality as a training tool to treat
physical inactivity in children. Frontiers in Public Health 5: 349, 2017.
30. Koch LG and Britton SL. Artificial selection for intrinsic aerobic endurance running capacity in
rats. Physiological Genomics 5: 45–​52, 2001.
31. Lecker SH, Jagoe RT, Gilbert A, Gomes M, Baracos V, Bailey J, Price SR, Mitch WE, and
Goldberg AL. Multiple types of skeletal muscle atrophy involve a common program of changes in
gene expression. FASEB Journal 18: 39–​51, 2004.
32. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, and Katzmarzyk PT. Effect of physical
inactivity on major non-​communicable diseases worldwide: an analysis of burden of disease and life
expectancy. The Lancet 380: 219–​229, 2012.
33. Lindholm ME, Giacomello S, Werne Solnestam B, Fischer H, Huss M, Kjellqvist S, and
Sundberg CJ. The impact of endurance training on human skeletal muscle memory, global isoform
expression and novel transcripts. PLoS Genetics 12: e1006294, 2016.
34. McGavock JM, Hastings JL, Snell PG, McGuire DK, Pacini EL, Levine BD, and Mitchell JH.
A forty-​year follow-​up of the Dallas Bed Rest and Training study: The effect of age on the cardio-
vascular response to exercise in men. The Journals of Gerontology Series A, Biological Sciences and Medical
Sciences 64: 293–​299, 2009.
35. Morahan G and Williams RW. Systems genetics: The next generation in genetics research? Novartis
Foundation Symposium 281: 181–​188; discussion 188–​191, 208–​189, 2007.

14
15

Systems genetics of sport and exercise

36. Padrao AI, Ferreira R, Amado F, Vitorino R, and Duarte JA. Uncovering the exercise-​related
proteome signature in skeletal muscle. Proteomics 16: 816–​830, 2016.
37. Pedersen BK and Febbraio MA. Muscles, exercise and obesity: Skeletal muscle as a secretory
organ. Nature Reviews Endocrinology 8: 457–​465, 2012.
38. Pedersen BK, Steensberg A, Fischer C, Keller C, Keller P, Plomgaard P, Febbraio M, and
Saltin B. Searching for the exercise factor: Is IL-​6 a candidate? Journal of Muscle Research and Cell
Motility 24: 113–​119, 2003.
39. Petersen AM and Pedersen BK. The anti-​ inflammatory effect of exercise. Journal of Applied
Physiology (Bethesda, MD: 1985) 98: 1154–​1162, 2005.
40. Prensky M. Digital natives, digital immigrants part 1. On the Horizon 9: 1–​6, 2001.
41. Rector RS, Rogers R, Ruebel M, and Hinton PS. Participation in road cycling vs running
is associated with lower bone mineral density in men. Metabolism: Clinical and Experimental
57: 226–​232, 2008.
42. Roberts MD, Gilpin L, Parker KE, Childs TE, Will MJ, and Booth FW. Dopamine D1 receptor
modulation in nucleus accumbens lowers voluntary wheel running in rats bred to run high distances.
Physiology & Behavior 105: 661–​668, 2012.
43. Roman B, Serra-​Majem L, Ribas-​Barba L, Perez-​Rodrigo C, and Aranceta J. How many
children and adolescents in Spain comply with the recommendations on physical activity? The Journal
of Sports Medicine and Physical Fitness 48: 380–​387, 2008.
44. Ruegsegger GN and Booth FW. Health benefits of exercise. Cold Spring Harbor Perspectives in
Medicine 8: a029694, 2018.
45. Ruegsegger GN, Toedebusch RG, Braselton JF, Childs TE, and Booth FW. Left ventricle
transcriptomic analysis reveals connective tissue accumulation associates with initial age-​dependent
decline in Vo2peak from its lifetime apex. Physiological Genomics 49: 53–​66, 2017.
46. Schulz LO and Chaudhari LS. High-​r isk populations: The Pimas of Arizona and Mexico. Current
Obesity Reports 4: 92–​98, 2015.
47. Stein TP and Bolster DR. Insights into muscle atrophy and recovery pathway based on genetic
models. Current Opinion in Clinical Nutrition and Metabolic Care 9: 395–​402, 2006.
48. Swallow JG, Carter PA, and Garland T, Jr. Artificial selection for increased wheel-​ running
behavior in house mice. Behavior Genetics 28: 227–​237, 1998.
49. Toedebusch RG, Ruegsegger GN, Braselton JF, Heese AJ, Hofheins JC, Childs TE, Thyfault
JP, and Booth FW. AMPK agonist AICAR delays the initial decline in lifetime-​apex Vo2 peak,
while voluntary wheel running fails to delay its initial decline in female rats. Physiological Genomics
48: 101–​115, 2016.
50. Trappe SW, Costill DL, Vukovich MD, Jones J, and Melham T. Aging among elite dis-
tance runners: a 22-​ yr longitudinal study. Journal of Applied Physiology (Bethesda, MD: 1985)
80: 285–​290, 1996.
51. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, and McDowell M. Physical
activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise
40: 181–​188, 2008.
52. Ullum H, Haahr PM, Diamant M, Palmo J, Halkjaer-​Kristensen J, and Pedersen BK. Bicycle
exercise enhances plasma IL-​6 but does not change IL-​1 alpha, IL-​1 beta, IL-​6, or TNF-​alpha pre-​
mRNA in BMNC. Journal of Applied Physiology (Bethesda, MD: 1985) 77: 93–​97, 1994.
53. Vernikos J and Schneider VS. Space, gravity and the physiology of aging: parallel or convergent
disciplines? A mini-​review. Gerontology 56: 157–​166, 2010.
54. Whitham M, Parker BL, Friedrichsen M, Hingst JR, Hjorth M, Hughes WE, Egan CL, Cron
L, Watt KI, Kuchel RP, Jayasooriah N, Estevez E, Petzold T, Suter CM, Gregorevic P, Kiens
B, Richter EA, James DE, Wojtaszewski JFP, and Febbraio MA. Extracellular vesicles provide a
means for tissue crosstalk during exercise. Cell Metabolism 27: 237–​251.e234, 2018.
55. Woolf SH and Aron L. Failing health of the United States. BMJ 360: k496, 2018.

15
16

2
EXPANSION OF KNOWLEDGE
AND ADVANCES IN GENETICS
FOR QUANTITATIVE ANALYSES
Penny K. Riggs

Introduction
During the nearly two decades since completion of the Human Genome Project, breakthroughs
in technologies and analytical approaches, along with deeper understanding of the genome,
have resulted in advanced research tools. Because of this progress, research efforts can be
directed toward clarifying the genetic mechanisms that affect physiological responses to exercise
activity and performance. This chapter broadly introduces genetic concepts that have emerged
since the beginning of the millennium. In addition to exploring new-​found aspects of genome
function and regulation, the role of multigenic traits and the challenges in determining causality
following identification of genetic associations are also addressed.

The Human Genome Project

Following debates and discussion among scientists, policymakers, and other interested parties
during the early 1980s, a strategy developed for conducting the Human Genome Project, which
was anticipated to be a 15-​year endeavor. With US congressionally authorized funding, the
Department of Energy (DOE) and the National Institutes of Health (NIH) agreed to partici-
pate in making the project a reality (15). A final component of funding was appropriated in
1989 to establish the National Center for Human Genome Research (PLN 101–​166 (55)).The
Center, now known as the National Human Genome Research Institute (NHGRI), along with
the DOE Joint Genome Institute, provided “headquarters” for the daunting project that offi-
cially began in October 1990. Lively accounts of events leading up to the project give detailed
looks into the personalities, policies, and substantial discussion that preceded the project’s imple-
mentation (13, 66).
Over the next decade, the Human Genome Project was on in the US, and in conjunction
with international partners. Technological advances in methodology and computing, and cor-
porate competition spurred on the process. With the turn of the century, the first draft of the
genome sequence was declared complete, and in February 2001, the International Human
Genome Project team (29) and the Celera Genomics team (74) independently published the
first sequence assembly drafts of the human genome. Completion of the draft brought some
surprise and generated discussion. That the structure of the genome contained far fewer genes

16
17

Genetics for quantitative analyses

than anticipated was unexpected (12). Political and ethical concerns for the use of genomic
information (32), and controversy over competing approaches to the project (9) brought much
attention and new legislation such as The Genetic Information Nondiscrimination Act (50).
However, enthusiasm for using this “great gift to humanity” (31) ensued, and the rapid pace of
advancements in knowledge and tools since 2000 is nothing short of amazing.

Advances in sequencing
In 2004, the euchromatic portion of the human genome sequence was finished and the “near-​
complete” human genome sequence was ready to use (28). Genomes of other species, begin-
ning with the mouse (48), became available as well. Since then, rapid progress in technology
has greatly altered the landscape. As Shendure and colleagues (67) note, automation and con-
tinual improvements in fluorescence-​based methods led to near-​maximum efficiency and cost
reduction for Sanger sequencing strategies. However, the completion of the human genome
sequence was only the beginning for the genome sequencing era. Next-​generation sequencing
(NGS) technologies appeared on the market during the 2000s as several companies developed
different approaches for “massively parallel sequencing” methods. During this time, sequencing
throughput accelerated exponentially. Strengths and limitations of various methods (8) and
thorough reviews of the historical timeline are published elsewhere (46, 57, 67). By 2015, the
cost of genome sequencing approached the once far-​fetched cost of $1000 USD per genome
(49), and much anticipated third-​generation sequencing (5, 64) technologies that enable single
molecule sequencing also became available. One of these methods, nanopore sequencing, iden-
tifies nucleotides based on the electrical disturbance created as an enzyme ratchets movement of
a nucleic acid molecule through a nanopore (37, 42). This process, commercialized by Oxford
Nanopore Technologies Limited (Oxford, UK), has resulted in small, portable sequencing
devices that produce long sequence reads, such as the minION device shown in Figure 2.1
that has now been utilized to successfully generate an entire human genome sequence (17, 30).
A smaller device designed for use with a smart phone is reportedly in development as of August

Figure 2.1 An example of a hand-​held minION sequencer in use in a laboratory setting.

Source: Photo provided by Andrew E. Hillhouse, Texas A&M University Institute of Genome Sciences
and Society.

17
18

P.K. Riggs

2018.The availability of such a device means the ability to generate sequence data anytime, any-
where, and at an affordable cost, has been realized. However, challenges still remain in storing,
analyzing, and comprehending these data.

Association studies
For many decades, researchers made maps of gene locations to use as tools for finding the genes
that were responsible for diseases, defects, or phenotypic traits of interest (58). As genomic
sequence data from more individuals became available, variations between individuals were
identified.The database of sequence variants –​single nucleotide polymorphisms (SNPs) –​could
be used for identifying genomic regions associated with measurable phenotypic traits affected
by multiple genes. The regions, known as quantitative trait loci (QTL), could be detected in
populations when the genes or sequences responsible for phenotype variation segregated within
the population studied, and SNP markers were distributed across the map at sufficient density.
With knowledge of family structure, each SNP could be tested statistically for association with
the trait of interest (10, 38).
As genome sequencing became more accessible, SNP markers reflecting common genetic
variation accumulated more rapidly. With a goal to find genes related to health and disease, the
International HapMap Consortium was formed to collect and characterize human variation on
a large scale. In phase II of the project, more than 3 million SNPs from 270 people from four
geographic regions were characterized (71). This HapMap resource was soon replaced after
the 1000 Genomes Project was completed and produced a map of variation from 2504 people
from 26 populations (68, 70). Similarly, genome resources were accumulated for mice and
numerous other species (34). With abundant characterized SNPs, and improvements in micro-
array platforms for conducting hundreds of thousands of assays simultaneously, the traditional
genetic linkage studies gave way to genome-​wide association studies (GWAS) as a popular
approach for associating genomic regions with quantitative, physical traits of interest (16, 35, 52,
56, 60). Publication of GWAS rose rapidly and has held at a fairly steady state in recent years
(Figure 2.2), although individual studies continue to increase in size. Connecting genetic vari-
ation with complex, quantitative traits remains challenging because many genetic variations,
each with small effects, influence overall phenotype.
A GWAS approach takes advantage of knowledge of common genetic variation across the
genome and requires only that the SNP be in linkage disequilibrium with the causal variant.
No prior knowledge of the genome map location is required for identification of a variant
responsible for a trait of interest (27). However, meaningful GWAS results require careful study
design, sufficient SNP density and population frequency, and sample numbers of a size to
achieve suitable statistical power for analysis (27, 54). Although powerful, this approach has not
been without controversy and debate has occurred about whether GWAS experiments could
return value worth the cost (76). The probability of success for GWAS is affected by several
factors including the number of loci that influence a trait, allele frequency within the popula-
tion, the effect on phenotype, the size of the sample population, the SNP panel composition,
and how well the phenotype can be accurately measured (77). Depending on effect size and
minor allele frequency, minimum sample sizes for detecting SNP–​trait associations could range
from a few hundred to more than 100 million (77). Thus, an investigator’s choice of analytical
tools and models, as well as treatment of rare variants, pleiotropic effects, and means for val-
idation must also be carefully considered (19, 25, 53). For complex physiological phenotypes,
GWAS might be expected to result in variable rates of success. In 2007, one of the earliest large
GWAS investigated seven diseases, with 2000 individuals for each case, along with a set of 3000

18
19

Genetics for quantitative analyses

Figure 2.2 The Medline database was queried with the keyword search terms “genome wide
association” or “genome-​wide association studies” and publication numbers per year are plotted in
this graph.

control individuals (72). Several significant associations at new regions of the genome offered
insight into mechanisms underlying disease. These early studies and subsequent data have also
supported the idea that complex traits are highly polygenic, and the path from association to
causation is often slow and complex (2). Despite drawbacks, Bouchard described the usefulness
of GWAS in exercise studies (4) and Meyre (47) argued for continued investment in appropri-
ately designed studies, complemented by whole-​genome sequencing and sequence imputation,
citing the impact of a large meta-​analysis of diabetes-​related GWAS publications (65).
Exercise, physical activity, and sport-​related GWAS studies are, to some extent, only begin-
ning. For example, Sarzynski and colleagues (62) identified four SNP loci associated with tri-
glyceride response to exercise training, as well as gene expression (transcriptomic) variants that
may contribute to variability in lipid response. In another study, Verweij et al. examined heart
rate response during and after exercise (75). The GWAS analyses returned 23 loci associated
with heart rate response resulting in 36 candidate causal genes for further investigation. Many
of the identified candidates play a role in neuron biology, and led the authors to new hypoth-
eses regarding the role of the autonomic nervous system in heart rate recovery, as well as new
directions for investigating sudden cardiac death. Several investigators are also beginning to
consider the entire body of association results for specific applications (41, 63, 78). Frau and
colleagues (22) undertook a systems genetics analysis to identify genetic variation from mul-
tiple GWAS for type 2 diabetes as well as 12 additional traits that are associated comorbidities.
To do so, they combined GWAS results, an examination of cross-​trait effects, and gene net-
work analyses to identify a set of 38 genetic variants associated with type 2 diabetes as well as
its comorbidities. These types of studies lay a strong foundation for investigating the genetic
mechanisms by which exercise and sport can play a role in overall health and disease prevention.
In addition, mouse model systems have also proven useful for identifying genetic loci associated
with voluntary physical activity (36, 40), as well as responses to exercise training (44).

19
20

P.K. Riggs

Noncoding sequences
Noncoding RNA and antisense sequences
One outcome of various genome projects, and ensuing genetic analyses, is the discovery that
the genome is far more complex than may have been anticipated. At one time, descriptions
of “junk DNA” (51) and the discovery of far fewer than expected genes within the human
genome caused surprise. As interrogation of the genome has progressed, a variety of gen-
omic regulatory elements and transcribed RNAs that do not encode gene products have
revealed that genome function is much more complicated than the “Central Dogma” of
DNA to messenger RNA to protein. This truth is observed in the outcome of GWAS. For
example, the landmark Wellcome Trust study in 2007 revealed a previously unidentified,
strong association between coronary artery disease (CAD) and a region of chromosome
9p21 (72). The associated SNPs were located in a noncoding region of the DNA, and have
strong associations with CAD in multiple populations.Yet, in 2018, the causal mechanism of
this sequence with CAD remains unknown despite progress that has led to subsequent iden-
tification of more than 150 loci associated with CAD risk (11). A noncoding RNA tran-
script, XIST, was identified as key to inactivation of the X-​chromosome in female mammals
(6, 7). In the case of callipyge sheep, affected animals exhibit muscular hypertrophy of the
loin and hindquarters, but not all muscles, due to a single SNP in an intergenic region of the
sheep genome (21, 69). This SNP results in altered expression of both sense and antisense
transcripts, with additional parent-​of-​origin effects. Although the mutation is well known,
the mechanisms by which postnatal muscle growth is altered are not completely under-
stood. Even with knowledge of the genome sequence, understanding how that sequence is
converted to phenotype remains a complex problem.

MicroRNAs
Small (~17–​27 nucleotides) RNA molecules known as microRNAs, or miRNAs, can function
as a type of molecular rheostat to modulate gene expression during various physiological
processes. MicroRNAs can act on entire networks of genes, vastly increasing the complexity of
genetic mechanisms (33). Skeletal muscle phenotypes can be regulated by miRNAs, whose dif-
ferential expression within muscle may reflect response to exercise or even variation in activity
(14, 43). In human and mouse experiments, Gastebois et al. (23) demonstrated that expression of
a miRNA, miR-​148b, played a role in decreased insulin sensitivity following a transition from
physical activity to inactivity. This miRNA was previously shown to be associated with muscle
catabolism (79), and the investigators found that inactivity resulted in increased miR-​148b and
downregulation of its target genes, resulting in reduced insulin responsiveness. In another study,
Lew and colleagues conducted an extensive review of how exercise can modulate miRNA
activity in molecular networks to protect cardiac function in diabetic patients. They proposed
that miRNAs, shown to be induced by exercise and to have cardioprotective effects, could serve
as biomarkers for determining optimal exercise strategies for diabetic patients (39).

Exosomes
Functional miRNAs are cleaved from longer precursor molecules, and the remaining “pas-
senger strand” molecule was thought to be degraded. However, in an investigation into
mechanisms of cardiac fibrosis in response to stress, and subsequent heart failure, Bang and

20
21

Genetics for quantitative analyses

colleagues discovered that cardiac fibroblasts excreted microvesicles known as exosomes (3).
They found that under stress conditions, fibroblast exosomes containing the miR-​21 passenger
strand miRNA (miR-​21*) acted on cardiac myocytes, and miR-​21* altered gene expression
that ultimately led to hypertrophic growth. This study is just one example of how miRNA
regulatory function can contribute orders of complexity to genetic studies of exercise-​related
physiological functions. Additional studies have found that physical exercise can cause skel-
etal muscle to release exosomes containing miRNAs into the bloodstream (24), potentially
functioning as trans-​acting regulators on other tissues. As noted in a discussion of exercise-​
induced circulating miRNAs, a current challenge is to move toward understanding the func-
tional role of these regulatory molecules in skeletal muscle adaptation, and other physiological
effects that occur in response to physical activity (61).

Epigenetics
In addition to variation in the sequence of the genetic code contained in the genome, epi-
genetic modifications that enable heritable alterations in phenotype without changes in DNA
sequence must also be considered (1). Modification of the histone proteins by methylation,
acetylation, and other processes, as well as methylation of stretches of guanines and cytosines
in the genome (CpG islands) can alter gene activity (often in a repressive fashion in the case
of methylation). Following the pattern of other genome mapping projects, the Epigenomics
Roadmap Project established complete epigenomes for 111 reference human genomes (59).
Epigenetic modifications occur in early development during the differentiation of specific cell
types. Epigenetics is also a component of fetal programming, in which the fetal environment can
have long-​term effects on a developing fetus, as well as its offspring. This phenomenon gained
much attention in studies of the Dutch famine birth cohort, whose mothers had experienced
starvation (18, 26). In this study, those who were subjected to prenatal famine had altered
methylation of the IGF2 gene in comparison with their siblings who were born later during
less severe conditions.
Exercise has been shown to affect epigenetic processes. An early study of histone modi-
fication following bouts of cycling exercise demonstrated a global increase in acetylation of
H3K36 (45). An increase in transcriptional elongation is a typical response to this histone modi-
fication. In one extensive review that documents exercise-​induced epigenetic effects, induced
DNA methylation and histone acetylation or methylation was shown to result in modifications
important for learning, memory, and cognition (20). Another meta-​analysis of literature to
examine the direct and indirect impact of exercise on cancer-​related biochemical pathways
was conducted by Thomas et al. (73). In this study, the authors categorized the published direct
effects on the activity of several biological pathways that likely have a positive impact on health.
They also examined current literature that has demonstrated postexercise epigenetic alterations
including histone modifications, DNA methylation, and telomere length. While these studies
provide valuable information, the review highlights specific areas where clear understanding of
the biological consequences of exercise biochemistry falls short.

Summary
The completion of the Human Genome Project represented the opening of a door into a
whole new realm of genetic technologies and advances that have continued to expand during
the past two decades. Along with the advances have come many more questions and enormous
quantities of data. What remains is how to best conduct sound experiments to interrogate

21
22

P.K. Riggs

and analyze these data, and to move from genetic variation to causal mechanisms that shape
phenotypes of interest. With careful study designs and tools at hand, the interaction of genetics
with exercise and sport activity can be better understood and utilized for improving overall
health and well-​being.

References
1. Allis CD and Jenuwein T. The molecular hallmarks of epigenetic control. Nat Rev Genet
17: 487, 2016.
2. Angel JM, Abel EL, Riggs PK, McClellan SA, and DiGiovanni J. Fine mapping reveals that
promotion susceptibility locus 1 (Psl1) is a compound locus with multiple genes that modify suscep-
tibility to skin tumor development. G3 (Bethesda), 4: 1071–​1079, 2014.
3. Bang C, Batkai S, Dangwal S, Gupta SK, Foinquinos A, Holzmann A, Just A, Remke
J, Zimmer K, Zeug A, Ponimaskin E, Schmiedl A, Yin X, Mayr M, Halder R, Fischer
A, Engelhardt S, Wei Y, Schober A, Fiedler J, and Thum T. Cardiac fibroblast-​derived
microRNA passenger strand-​ enriched exosomes mediate cardiomyocyte hypertrophy. J Clin
Invest 124: 2136–​2146, 2014.
4. Bouchard C. Overcoming barriers to progress in exercise genomics. Exerc Sport Sci Rev 39:
212–​217, 2011.
5. Branton D, Deamer DW, Marziali A, Bayley H, Benner SA, Butler T, Di Ventra M, Garaj S,
Hibbs A, Huang X, Jovanovich SB, Krstic PS, Lindsay S, Ling XS, Mastrangelo CH, Meller
A, Oliver JS, Pershin YV, Ramsey JM, Riehn R, Soni GV, Tabard-​Cossa V, Wanunu M,
Wiggin M, and Schloss JA. The potential and challenges of nanopore sequencing. Nat Biotechnol
26: 1146–​1153, 2008.
6. Brockdorff N, Ashworth A, Kay GF, McCabe VM, Norris DP, Cooper PJ, Swift S, and
Rastan S. The product of the mouse Xist gene is a 15 kb inactive X-​specific transcript containing
no conserved ORF and located in the nucleus. Cell 71: 515–​526, 1992.
7. Brown CJ, Hendrich BD, Rupert JL, Lafrenière RG, Xing Y, Lawrence J, and Willard HF.
The human XIST gene: Analysis of a 17 kb inactive X-​specific RNA that contains conserved repeats
and is highly localized within the nucleus. Cell 71: 527–​542, 1992.
8. Buermans HPJ and den Dunnen JT. Next generation sequencing technology: Advances and
applications. Biochim Biophys Acta 1842: 1932–​1941, 2014.
9. Butler D. Publication of human genomes sparks fresh sequence debate. Nature 409: 747, 2001.
10. Churchill GA and Doerge RW. Empirical threshold values for quantitative trait mapping. Genetics
138: 963–​971, 1994.
11. Clarke SL and Assimes TL. Genome-​wide association studies of coronary artery disease: Recent
progress and challenges ahead. Curr Atheroscler Rep 20: 47, 2018.
12. Claverie J-​M. What if there are only 30,000 human genes? Science 291: 1255–​1257, 2001.
13. Cook-​Deegan RM. The Human Genome Project: The formation of federal policies in the United
States, 1986–​1990. In: Biomedical Politics, edited by Hanna KE and Institute of Medicine (US)
Committee to Study Biomedical Decision Making. Washington, DC: National Academy Press, 1991,
p. 99–​168.
14. Dawes M, Kochan KJ, Riggs PK, and Lightfoot JT. Differential miRNA expression in inherently
high-​and low-​active inbred mice. Physiol Rep 3: e12469, 2015.
15. DeLisi C. Santa Fe 1986: Human genome baby-​steps. Nature 455: 876, 2008.
16. Dewan A, Liu M, Hartman S, Zhang SS, Liu DT, Zhao C, Tam PO, Chan WM, Lam DS,
Snyder M, Barnstable C, Pang CP, and Hoh J. HTRA1 promoter polymorphism in wet age-​
related macular degeneration. Science 314: 989–​992, 2006.
17. Eisenstein M. An ace in the hole for DNA sequencing. Nature 550: 285, 2017.
18. El Hajj N, Schneider E, Lehnen H, and Haaf T. Epigenetics and life-​long consequences of an
adverse nutritional and diabetic intrauterine environment. Reproduction 148: R111–​R120, 2014.
19. Evangelou E and Ioannidis JPA. Meta-​analysis methods for genome-​wide association studies and
beyond. Nat Rev Genet 14: 379, 2013.
20. Fernandes J, Arida RM, and Gomez-​Pinilla F. Physical exercise as an epigenetic modulator of
brain plasticity and cognition. Neurosci Biobehav Rev 80: 443–​456, 2017.

22
23

Genetics for quantitative analyses

21. Fleming-​Waddell JN, Olbricht GR, Taxis TM, White JD, Vuocolo T, Craig BA, Tellam RL,
Neary MK, Cockett NE, and Bidwell CA. Effect of DLK1 and RTL1 but not MEG3 or MEG8
on muscle gene expression in callipyge lambs. PLoS One 4: e7399, 2009.
22. Frau F, Crowther D, Ruetten H, and Allebrandt KV. Type-​2 diabetes-​associated variants with
cross-​trait relevance: Post-​GWAs strategies for biological function interpretation. Mol Genet Metab
121: 43–​50, 2017.
23. Gastebois C, Chanon S, Rome S, Durand C, Pelascini E, Jalabert A, Euthine V, Pialoux V,
Blanc S, Simon C, and Lefai E. Transition from physical activity to inactivity increases skeletal
muscle miR-​148b content and triggers insulin resistance. Physiol Rep 4: e12902, 2016.
24. Guescini M, Canonico B, Lucertini F, Maggio S, Annibalini G, Barbieri E, Luchetti F, Papa
S, and Stocchi V. Muscle releases alpha-​sarcoglycan positive extracellular vesicles carrying miRNAs
in the bloodstream. PLoS One 10: e0125094, 2015.
25. Hackinger S and Zeggini E. Statistical methods to detect pleiotropy in human complex traits.
Open Biol 7: 170125, 2017.
26. Heijmans BT, Tobi EW, Stein AD, Putter H, Blauw GJ, Susser ES, Slagboom PE, and Lumey
LH. Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proc
Natl Acad Sci U S A 105: 17046–​17049, 2008.
27. Hirschhorn JN and Daly MJ. Genome-​wide association studies for common diseases and complex
traits. Nat Rev Genet 6: 95, 2005.
28. International Human Genome Sequencing Consortium. Finishing the euchromatic sequence
of the human genome. Nature 431: 931–​945, 2004.
29. International Human Genome Sequencing Consortium, Lander ES, Linton LM, Birren
B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke
R, Gage D, Harris K, Heaford A, Howland J, Kann L, Lehoczky J, LeVine R, McEwan P,
McKernan K, Meldrim J, Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C, Rosetti
M, Santos R, Sheridan A, Sougnez C, Stange-​Thomann Y, Stojanovic N, Subramanian A,
Wyman D, Rogers J, Sulston J, Ainscough R, Beck S, Bentley D, Burton J, Clee C, Carter N,
Coulson A, Deadman R, Deloukas P, Dunham A, Dunham I, Durbin R, French L, Grafham
D, Gregory S, Hubbard T, Humphray S, Hunt A, Jones M, Lloyd C, McMurray A, Matthews
L, Mercer S, Milne S, Mullikin JC, Mungall A, Plumb R, Ross M, Shownkeen R, Sims S,
Waterston RH, Wilson RK, Hillier LW, McPherson JD, Marra MA, Mardis ER, Fulton LA,
Chinwalla AT, Pepin KH, Gish WR, Chissoe SL, Wendl MC, Delehaunty KD, Miner TL,
Delehaunty A, Kramer JB, Cook LL, Fulton RS, Johnson DL, Minx PJ, Clifton SW, Hawkins
T, Branscomb E, Predki P, Richardson P, Wenning S, Slezak T, Doggett N, Cheng JF, Olsen
A, Lucas S, Elkin C, Uberbacher E, et al. Initial sequencing and analysis of the human genome.
Nature 409: 860–​921, 2001.
30. Jain M, Koren S, Miga KH, Quick J, Rand AC, Sasani TA, Tyson JR, Beggs AD, Dilthey
AT, Fiddes IT, Malla S, Marriott H, Nieto T, O’Grady J, Olsen HE, Pedersen BS, Rhie A,
Richardson H, Quinlan AR, Snutch TP, Tee L, Paten B, Phillippy AM, Simpson JT, Loman
NJ, and Loose M. Nanopore sequencing and assembly of a human genome with ultra-​long reads.
Nat Biotechnol 36: 338, 2018.
31. Jasny BR and Kennedy D. The human genome. Science 291: 1153–​1153, 2001.
32. Jeffords JM and Daschle T. Political issues in the genome era. Science 291: 1249–​1251, 2001.
33. Jeffries CD, Fried HM, and Perkins DO. Additional layers of gene regulatory complexity from
recently discovered microRNA mechanisms. Int J Biochem Cell Biol 42: 1236–​1242, 2010.
34. Kirby A, Kang HM, Wade CM, Cotsapas C, Kostem E, Han B, Furlotte N, Kang EY, Rivas M,
Bogue MA, Frazer KA, Johnson FM, Beilharz EJ, Cox DR, Eskin E, and Daly MJ. Fine mapping
in 94 inbred mouse strains using a high-​density haplotype resource. Genetics 185: 1081–​1095, 2010.
35. Klein RJ, Zeiss C, Chew EY, Tsai JY, Sackler RS, Haynes C, Henning AK, SanGiovanni JP,
Mane SM, Mayne ST, Bracken MB, Ferris FL, Ott J, Barnstable C, and Hoh J. Complement
factor H polymorphism in age-​related macular degeneration. Science 308: 385–​389, 2005.
36. Kostrzewa E, Brandys MK, van Lith HA, and Kas MJH. A candidate syntenic genetic locus is
associated with voluntary exercise levels in mice and humans. Behav Brain Res 276: 8–​16, 2015.
37. Laszlo AH, Derrington IM, Ross BC, Brinkerhoff H, Adey A, Nova IC, Craig JM, Langford
KW, Samson JM, Daza R, Doering K, Shendure J, and Gundlach JH. Decoding long nanopore
sequencing reads of natural DNA. Nat Biotechnol 32: 829, 2014.

23
24

P.K. Riggs

38. Lebreton CM and Visscher PM. Empirical nonparametric bootstrap strategies in quantitative trait
loci mapping: conditioning on the genetic model. Genetics 148: 525–​535, 1998.
39. Lew JKS, Pearson JT, Schwenke DO, and Katare R. Exercise mediated protection of dia-
betic heart through modulation of microRNA mediated molecular pathways. Cardiovasc Diabetol
16: 10, 2017.
40. Lightfoot JT, Turner MJ, Pomp D, Kleeberger SR, and Leamy LJ. Quantitative trait loci for
physical activity traits in mice. Physiol Genomics 32: 401–​408, 2008.
41. Lin X, Eaton CB, Manson JE, and Liu S. The genetics of physical activity. Curr Cardiol Rep
19: 119, 2017.
42. Madoui MA, Engelen S, Cruaud C, Belser C, Bertrand L, Alberti A, Lemainque A, Wincker
P, and Aury JM. Genome assembly using nanopore-​guided long and error-​free DNA reads. BMC
Genomics 16: 327, 2015.
43. Margolis LM, Lessard SJ, Ezzyat Y, Fielding RA, and Rivas DA. Circulating MicroRNA are
predictive of aging and acute adaptive response to resistance exercise in men. J Gerontol A Biol Sci Med
Sci 72: 1319–​1326, 2017.
44. Massett MP, Avila JJ, and Kim SK. Exercise capacity and response to training quantitative trait
loci in a NZW X 129S1 intercross and combined cross analysis of inbred mouse strains. PLoS One
10: e0145741, 2016.
45. McGee SL, Fairlie E, Garnham AP, and Hargreaves M. Exercise-​induced histone modifications
in human skeletal muscle. J Physiol 587: 5951–​5958, 2009.
46. McGinn S and Gut IG. DNA sequencing –​spanning the generations. N Biotechnol 30:
366–​372, 2013.
47. Meyre D. Give GWAS a chance. Diabetes 66: 2741–​2742, 2017.
48. Mouse Genome Sequencing Consortium, Chinwalla AT, Cook LL, Delehaunty KD,
Fewell GA, Fulton LA, Fulton RS, Graves TA, Hillier LW, Mardis ER, McPherson JD,
Miner TL, Nash WE, Nelson JO, Nhan MN, Pepin KH, Pohl CS, Ponce TC, Schultz B,
Thompson J, Trevaskis E, Waterston RH, Wendl MC, Wilson RK, Yang S-​P, An P, Berry
E, Birren B, Bloom T, Brown DG, Butler J, Daly M, David R, Deri J, Dodge S, Foley K,
Gage D, Gnerre S, Holzer T, Jaffe DB, Kamal M, Karlsson EK, Kells C, Kirby A, Kulbokas
Iii EJ, Lander ES, Landers T, Leger JP, Levine R, Lindblad-​Toh K, Mauceli E, Mayer JH,
McCarthy M, Meldrim J, Meldrim J, Mesirov JP, Nicol R, Nusbaum C, Seaman S, Sharpe
T, Sheridan A, Singer JB, Santos R, Spencer B, Stange-​Thomann N, Vinson JP, Wade
CM, Wierzbowski J, Wyman D, Zody MC, Birney E, Goldman N, Kasprzyk A, Mongin
E, Rust AG, Slater G, Stabenau A, Ureta-​Vidal A, Whelan S, Ainscough R, Attwood J,
Bailey J, Barlow K, Beck S, Burton J, Clamp M, Clee C, Coulson A, Cuff J, Curwen V,
Cutts T, Davies J, Eyras E, Grafham D, Gregory S, Hubbard T, Hunt A, Jones M, Joy
A, Leonard S, et al. Initial sequencing and comparative analysis of the mouse genome. Nature
420: 520–​562, 2002.
49. National Human Genome Research Institute. The cost of sequencing a human genome. www.
genome.gov/​27565109/​the-​cost-​of-​sequencing-​a-​human-​genome/​ 2016. Accessed July 19, 2018.
50. National Human Genome Research Institute. Legislative history of GINA. www.genome.gov/​
27568535/​legislative-​history-​of-​gina/​ 2017. Accessed August 19, 2018.
51. Ohno S and Yomo T. The grammatical rule for all DNA: Junk and coding sequences. Electrophoresis
12: 103–​108, 1991.
52. Ozaki K, Ohnishi Y, Iida A, Sekine A, Yamada R, Tsunoda T, Sato H, Sato H, Hori M,
Nakamura Y, and Tanaka T. Functional SNPs in the lymphotoxin-​α gene that are associated with
susceptibility to myocardial infarction. Nat Genet 32: 650, 2002.
53. Palmer C and Pe’er I. Statistical correction of the Winner’s Curse explains replication variability in
quantitative trait genome-​wide association studies. PLoS Genet 13: e1006916, 2017.
54. Pearson TA and Manolio TA. How to interpret a genome-​ wide association study. JAMA
299: 1335–​1344, 2008.
55. Public Law Number 101–​156. H.R. 3566. Departments of Labor, Health and Human Services,
and Education, and related agencies appropriations. 101st Congress of the United States, 1989.
56. Rankinen T, An P, Pérusse L, Rice T, Chagnon YC, Gagnon J, Leon AS, Skinner JS, Wilmore
JH, Rao DC, and Bouchard C. Genome-​wide linkage scan for exercise stroke volume and cardiac
output in the HERITAGE Family Study. Physiol Genomics 10: 57–​62, 2002.
57. Reuter Jason A, Spacek DV, and Snyder Michael P. High-​throughput sequencing technologies.
Molecular Cell 58: 586–​597, 2015.

24
25

Genetics for quantitative analyses

58. Riggs PK and Gill CA. Molecular mapping and marker-​assisted breeding for muscle growth and
meat quality. In: Applied Muscle Biology and Meat Science, edited by Du M and McCormick RJ. Boca
Raton: CRC Press, 2009, p. 287–​310.
59. Roadmap Epigenomics Consortium, Kundaje A, Meuleman W, Ernst J, Bilenky M, Yen
A, Heravi-​Moussavi A, Kheradpour P, Zhang Z, Wang J, Ziller MJ, Amin V, Whitaker JW,
Schultz MD, Ward LD, Sarkar A, Quon G, Sandstrom RS, Eaton ML, Wu Y-​C, Pfenning
A, Wang X, ClaussnitzerYaping Liu M, Coarfa C, Alan Harris R, Shoresh N, Epstein CB,
Gjoneska E, Leung D, Xie W, David Hawkins R, Lister R, Hong C, Gascard P, Mungall AJ,
Moore R, Chuah E, Tam A, Canfield TK, Scott Hansen R, Kaul R, Sabo PJ, Bansal MS, Carles
A, Dixon JR, Farh K-​H, Feizi S, Karlic R, Kim A-​R, Kulkarni A, Li D, Lowdon R, Elliott G,
Mercer TR, Neph SJ, Onuchic V, Polak P, Rajagopal N, Ray P, Sallari RC, Siebenthall KT,
Sinnott-​Armstrong NA, Stevens M, Thurman RE, Wu J, Zhang B, Zhou X, Abdennur N, Adli
M, Akerman M, Barrera L, Antosiewicz-​Bourget J, Ballinger T, Barnes MJ, Bates D, Bell RJA,
Bennett DA, Bianco K, Bock C, Boyle P, Brinchmann J, Caballero-​Campo P, Camahort R,
Carrasco-​Alfonso MJ, Charnecki T, Chen H, Chen Z, Cheng JB, Cho S, Chu A, Chung W-​Y,
Cowan C, Athena Deng Q, Deshpande V, Diegel M, Ding B, Durham T, Echipare L, Edsall L,
Flowers D, et al. Integrative analysis of 111 reference human epigenomes. Nature 518: 317–​330, 2015.
60. Rosenthal A. Editorial overview. Genomics and proteomics. Curr Opin Mol Ther 1: 669–​670, 1999.
61. Russell AP and Lamon S. Exercise, skeletal muscle and circulating microRNAs. Prog Mol Biol Transl
Sci 135: 471–​495, 2015.
62. Sarzynski MA, Davidsen PK, Sung YJ, Hesselink MK, Schrauwen P, Rice TK, Rao DC,
Falciani F, and Bouchard C. Genomic and transcriptomic predictors of triglyceride response to
regular exercise. Br J Sports Med 49: 1524–​1531, 2015.
63. Sarzynski MA, Ghosh S, and Bouchard C. Genomic and transcriptomic predictors of response
levels to endurance exercise training. J Physiol 595: 2931–​2939, 2017.
64. Schadt EE, Turner S, and Kasarskis A. A window into third-​generation sequencing. Hum Mol
Genet 19: R227–​R240, 2010.
65. Scott RA, Scott LJ, Mägi R, Marullo L, Gaulton KJ, Kaakinen M, Pervjakova N, Pers TH,
Johnson AD, Eicher JD, Jackson AU, Ferreira T, Lee Y, Ma C, Steinthorsdottir V, Thorleifsson
G, Qi L, Van Zuydam NR, Mahajan A, Chen H, Almgren P, Voight BF, Grallert H, Müller-​
Nurasyid M, Ried JS, Rayner NW, Robertson N, Karssen LC, van Leeuwen EM, Willems
SM, Fuchsberger C, Kwan P, Teslovich TM, Chanda P, Li M, Lu Y, Dina C, Thuillier D,
Yengo L, Jiang L, Sparso T, Kestler HA, Chheda H, Eisele L, Gustafsson S, Frånberg M,
Strawbridge RJ, Benediktsson R, Hreidarsson AB, Kong A, Sigurðsson G, Kerrison ND,
Luan Ja, Liang L, Meitinger T, Roden M, Thorand B, Esko T, Mihailov E, Fox C, Liu C-​T,
Rybin D, Isomaa B, Lyssenko V, Tuomi T, Couper DJ, Pankow JS, Grarup N, Have CT,
Jørgensen ME, Jørgensen T, Linneberg A, Cornelis MC, van Dam RM, Hunter DJ, Kraft P,
Sun Q, Edkins S, Owen KR, Perry JRB, Wood AR, Zeggini E, Tajes-​Fernandes J, Abecasis
GR, Bonnycastle LL, Chines PS, Stringham HM, Koistinen HA, Kinnunen L, Sennblad
B, Mühleisen TW, Nöthen MM, Pechlivanis S, Baldassarre D, Gertow K, Humphries SE,
Tremoli E, Klopp N, Meyer J, Steinbach G, et al. An expanded genome-​wide association study
of type 2 diabetes in Europeans. Diabetes 66: 2888–​2902, 2017.
66. Shapiro R. The Human Blueprint: The Race to Unlock the Secrets of our Genetic Script. New York: St.
Martin’s Press, 1991.
67. Shendure J, Balasubramanian S, Church GM, Gilbert W, Rogers J, Schloss JA, and Waterston
RH. DNA sequencing at 40: past, present and future. Nature 550: 345–​353, 2017.
68. Sudmant PH, Rausch T, Gardner EJ, Handsaker RE, Abyzov A, Huddleston J, Zhang Y,
Ye K, Jun G, Hsi-​Yang Fritz M, Konkel MK, Malhotra A, Stütz AM, Shi X, Paolo Casale
F, Chen J, Hormozdiari F, Dayama G, Chen K, Malig M, Chaisson MJP, Walter K, Meiers
S, Kashin S, Garrison E, Auton A, Lam HYK, Jasmine Mu X, Alkan C, Antaki D, Bae T,
Cerveira E, Chines P, Chong Z, Clarke L, Dal E, Ding L, Emery S, Fan X, Gujral M, Kahveci
F, Kidd JM, Kong Y, Lameijer E-​W, McCarthy S, Flicek P, Gibbs RA, Marth G, Mason CE,
Menelaou A, Muzny DM, Nelson BJ, Noor A, Parrish NF, Pendleton M, Quitadamo A,
Raeder B, Schadt EE, Romanovitch M, Schlattl A, Sebra R, Shabalin AA, Untergasser A,
Walker JA, Wang M, Yu F, Zhang C, Zhang J, Zheng-​Bradley X, Zhou W, Zichner T, Sebat
J, Batzer MA, McCarroll SA, The Genomes Project C, Mills RE, Gerstein MB, Bashir A,
Stegle O, Devine SE, Lee C, Eichler EE, and Korbel JO. An integrated map of structural variation
in 2,504 human genomes. Nature 526: 75–​81, 2015.

25
26

P.K. Riggs

69. Takeda H, Caiment F, Smit M, Hiard S, Tordoir X, Cockett N, Georges M, and Charlier C.
The callipyge mutation enhances bidirectional long-​range DLK1-​GTL2 intergenic transcription in
cis. Proc Natl Acad Sci U S A 103: 8119–​8124, 2006.
70. The Genomes Project Consortium, Auton A, Abecasis GR, Altshuler DM, Durbin RM,
Abecasis GR, Bentley DR, Chakravarti A, Clark AG, Donnelly P, Eichler EE, Flicek P,
Gabriel SB, Gibbs RA, Green ED, Hurles ME, Knoppers BM, Korbel JO, Lander ES, Lee
C, Lehrach H, Mardis ER, Marth GT, McVean GA, Nickerson DA, Schmidt JP, Sherry ST,
Wang J, Wilson RK, Gibbs RA, Boerwinkle E, Doddapaneni H, Han Y, Korchina V, Kovar C,
Lee S, Muzny D, Reid JG, Zhu Y, Wang J, Chang Y, Feng Q, Fang X, Guo X, Jian M, Jiang H,
Jin X, Lan T, Li G, Li J, Li Y, Liu S, Liu X, Lu Y, Ma X, Tang M, Wang B, Wang G, Wu H, Wu
R, Xu X, Yin Y, Zhang D, Zhang W, Zhao J, Zhao M, Zheng X, Lander ES, Altshuler DM,
Gabriel SB, Gupta N, Gharani N, Toji LH, Gerry NP, Resch AM, Flicek P, Barker J, Clarke
L, Gil L, Hunt SE, Kelman G, Kulesha E, Leinonen R, McLaren WM, Radhakrishnan R, Roa
A, Smirnov D, Smith RE, Streeter I, Thormann A, Toneva I, Vaughan B, Zheng-​Bradley X,
Bentley DR, Grocock R, Humphray S, James T, Kingsbury Z, Lehrach H, Sudbrak R, et al.
A global reference for human genetic variation. Nature 526: 68–​74, 2015.
71. The International HapMap Consortium, Frazer KA, Ballinger DG, Cox DR, Hinds DA,
Stuve LL, Gibbs RA, Belmont JW, Boudreau A, Hardenbol P, Leal SM, Pasternak S, Wheeler
DA, Willis TD, Yu F, Yang H, Zeng C, Gao Y, Hu H, Hu W, Li C, Lin W, Liu S, Pan H, Tang
X, Wang J, Wang W, Yu J, Zhang B, Zhang Q, Zhao H, Zhao H, Zhou J, Gabriel SB, Barry R,
Blumenstiel B, Camargo A, Defelice M, Faggart M, Goyette M, Gupta S, Moore J, Nguyen
H, Onofrio RC, Parkin M, Roy J, Stahl E, Winchester E, Ziaugra L, Altshuler D, Shen Y, Yao
Z, Huang W, Chu X, He Y, Jin L, Liu Y, Shen Y, Sun W, Wang H, Wang Y, Wang Y, Xiong
X, Xu L, Waye MMY, Tsui SKW, Xue H, Wong JT-​F, Galver LM, Fan J-​B, Gunderson K,
Murray SS, Oliphant AR, Chee MS, Montpetit A, Chagnon F, Ferretti V, Leboeuf M, Olivier
J-​F, Phillips MS, Roumy S, Sallée C, Verner A, Hudson TJ, Kwok P-​Y, Cai D, Koboldt DC,
Miller RD, Pawlikowska L, Taillon-​Miller P, Xiao M, Tsui L-​C, Mak W, Qiang Song Y, Tam
PKH, Nakamura Y, Kawaguchi T, Kitamoto T, Morizono T, Nagashima A, et al. A second
generation human haplotype map of over 3.1 million SNPs. Nature 449: 851–​861, 2007.
72. The Wellcome Trust Case Control Consortium, Burton PR, Clayton DG, Cardon LR,
Craddock N, Deloukas P, Duncanson A, Kwiatkowski DP, McCarthy MI, Ouwehand WH,
Samani NJ, Todd JA, Donnelly P, Barrett JC, Burton PR, Davison D, Donnelly P, Easton
D, Evans D, Leung H-​T, Marchini JL, Morris AP, Spencer CCA, Tobin MD, Cardon LR,
Clayton DG, Attwood AP, Boorman JP, Cant B, Everson U, Hussey JM, Jolley JD, Knight
AS, Koch K, Meech E, Nutland S, Prowse CV, Stevens HE, Taylor NC, Walters GR, Walker
NM, Watkins NA, Winzer T, Todd JA, Ouwehand WH, Jones RW, McArdle WL, Ring
SM, Strachan DP, Pembrey M, Breen G, St Clair D, Caesar S, Gordon-​Smith K, Jones L,
Fraser C, Green EK, Grozeva D, Hamshere ML, Holmans PA, Jones IR, Kirov G, Moskvina
V, Nikolov I, O’Donovan MC, Owen MJ, Craddock N, Collier DA, Elkin A, Farmer A,
Williamson R, McGuffin P, Young AH, Ferrier IN, Ball SG, Balmforth AJ, Barrett JH,
Bishop DT, Iles MM, Maqbool A, Yuldasheva N, Hall AS, Braund PS, Burton PR, Dixon
RJ, Mangino M, Stevens S, Tobin MD, Thompson JR, Samani NJ, Bredin F, Tremelling M,
Parkes M, Drummond H, Lees CW, Nimmo ER, Satsangi J, Fisher SA, Forbes A, Lewis CM,
et al. Genome-​wide association study of 14,000 cases of seven common diseases and 3,000 shared
controls. Nature 447: 661–​678, 2007.
73. Thomas RJ, Kenfield SA, and Jimenez A. Exercise-​induced biochemical changes and their poten-
tial influence on cancer: A scientific review. Br J Sports Med 51: 640–​644, 2017.
74. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M,
Evans CA, Holt RA, Gocayne JD, Amanatides P, Ballew RM, Huson DH, Wortman JR,
Zhang Q, Kodira CD, Zheng XH, Chen L, Skupski M, Subramanian G, Thomas PD, Zhang
J, Gabor Miklos GL, Nelson C, Broder S, Clark AG, Nadeau J, McKusick VA, Zinder N, Levine
AJ, Roberts RJ, Simon M, Slayman C, Hunkapiller M, Bolanos R, Delcher A, Dew I, Fasulo
D, Flanigan M, Florea L, Halpern A, Hannenhalli S, Kravitz S, Levy S, Mobarry C, Reinert
K, Remington K, Abu-​Threideh J, Beasley E, Biddick K, Bonazzi V, Brandon R, Cargill M,
Chandramouliswaran I, Charlab R, Chaturvedi K, Deng Z, Francesco VD, Dunn P, Eilbeck
K, Evangelista C, Gabrielian AE, Gan W, Ge W, Gong F, Gu Z, Guan P, Heiman TJ, Higgins
ME, Ji R-​R, Ke Z, Ketchum KA, Lai Z, Lei Y, Li Z, Li J, Liang Y, Lin X, Lu F, Merkulov GV,

26
27

Genetics for quantitative analyses

Milshina N, Moore HM, Naik AK, Narayan VA, Neelam B, Nusskern D, Rusch DB, Salzberg
S, Shao W, Shue B, Sun J, Wang ZY, Wang A, Wang X, Wang J, Wei M-​H, Wides R, Xiao C,
Yan C, et al. The sequence of the human genome. Science 291: 1304–​1351, 2001.
75. Verweij N, van de Vegte YJ, and van der Harst P. Genetic study links components of the autono-
mous nervous system to heart-​rate profile during exercise. Nat Communications 9: 898, 2018.
76. Visscher Peter M, Brown Matthew A, McCarthy Mark I, and Yang J. Five years of GWAS dis-
covery. Am J Hum Genet 90: 7–​24, 2012.
77. Visscher PM, Wray NR, Zhang Q, Sklar P, McCarthy MI, Brown MA, and Yang J. 10 years of
GWAS discovery: Biology, function, and translation. Am J Hum Genet 101: 5–​22, 2017.
78. Williams CJ, Williams MG, Eynon N, Ashton KJ, Little JP, Wisloff U, and Coombes JS. Genes
to predict VO(2max) trainability: a systematic review. BMC Genomics 18: 831, 2017.
79. Zhang J, Fu SL, Liu Y, Liu YL, and Wang WJ. Analysis of microRNA expression profiles in
weaned pig skeletal muscle after lipopolysaccharide challenge. Int J Mol Sci 16: 22438–​22455, 2015.

27
28

3
HUMAN SYSTEMS GENETIC
MODELING USED IN EXERCISE
Jaakko Kaprio

Introduction
Nearly all human characteristics and behaviors run in families, that is to say, relatives and people
living together tend to resemble each other more than two individuals chosen at random from
the same population. Such characteristics can be structural (height, bone structure, obesity) or
physiological (muscle strength), behavioral (physical activity), or life-​course related (such as
number of children or lifespan). Correspondingly, abnormal conditions and disease often also
run in families. Familial aggregation is thus a well-​established observation, but the nature and
causes of it depends on the trait in question. Familial aggregation is also well known for physical
activity and exercise characteristics (2, 5, 21; see also Chapter 6). Given that there are multiple
modes of physical activity and exercise, this chapter uses physical activity traits as the primary
phenotypes in the examples described below.
Genetic modeling is a method to test hypotheses about the causes of familial aggregation and
to seek understanding of its mechanisms. Until recently, the role of genetic factors was inferred
on the basis of indirect evidence from family relationships –​greater resemblance between gen-
etically close persons was taken as evidence for genetic evidence but was not proof of it. Indeed,
among humans, almost all the literature on the role of genetic factors in physical activity comes
from modeling of family data, including special family designs such as twins. Modern molecular
genetic methods are only now, in the past few years, providing robust, direct evidence for gen-
etic factors to account for interindividual differences in human exercise and physical activity in
the population at large.
This chapter will not go deeply into the theoretical framework or detailed modeling of
quantitative genetics or molecular genetics. For this the reader is referred to textbooks and
review articles that cover these aspects more thoroughly and some suggested further reading is
provided at the end of the chapter. Further, Table 3.1 provides commonly accepted definitions
of the most frequently used terms in this area. The chapter will cover family and twin designs,
as they have been integral in providing evidence for genetic factors, while the chapter will
conclude by summarizing molecular genetic designs, particularly the genome-​wide association
study (GWAS) and its recent developments. A couple of very new studies are reviewed to show
how GWAS is being applied to exercise and physical activity.

28
29

Human systems genetic modeling

Table 3.1 Definitions of selected terms

Additive genetic variance: The component of variance of a phenotype that is attributable to the additive
effect of both alleles at all relevant loci
Allele: One of two or more states in which either copy of a gene can exist. This can be a single base-​
pair difference or a more extensive change in the genomic sequence
Assortative mating: A tendency for individuals with similar genotypes to mate
Genetic variance due to dominance: The component of variance of a phenotype that is attributable to the
interactions between alternative alleles at a locus over all relevant loci
Epistatic genetic variance: The component of variance of a phenotype that is attributable to the
interactive effect of two or more genes
Shared environmental variance: The component of variance of a phenotype that is attributable to the
experiences and exposures shared by family members making them more similar over the expected
similarity based on their degree of gene relatedness
Unshared environmental exposure: The component of variance of a phenotype that is attributable to all
environmental factors specific to the individual considered, and includes also measurement error

Source: Based partly on the “Glossary” of Thomas DC. Statistical methods in genetic epidemiology. Oxford
University Press, New York 2004, see “Further reading.”

Study of families
What are families?
The core biological and social unit for studies of families consists of two parents, who have one
or more offspring.They are thus biologically related due to transmission of shared chromosomes
and genes from parents to their children. Parents and their children are defined as first-​degree
relatives, and through other biological relatives, more extensive biological relationships can be
identified. These include grandparents, aunts and uncles, and cousins as second-​degree relatives.
By sharing genes in common, family members can be assumed to resemble each other due to
genes affecting the structures and functions that underlie physical activity.
We may further characterize the nuclear family as consisting of parents and children who
share the same environment for varying amounts of time. In an idealized situation, both parents
raise their children together, creating a family environment with common material conditions
(such as a shared home and finances), time spent together, and common rearing and family values.
This family environment then acts throughout the development of the child into adolescence
and at least until the child moves out of the parental home. One can then assume that the children
would adopt behaviors, such as sports, that their parents engage in.The children would also adopt
common parental values (e.g., the value of exercise).Thus, the parents and children resemble each
other due to the nature of the time they spend together.Together with the effect of shared genes,
shared social influences within families can result in familial aggregation. Overall, nongenetic
environmental influences are more important during the time family members share a common
household, i.e., during childhood and adolescence, but these influences can be maintained in
later life either socially or even through epigenetic mechanisms. If we study nuclear families
(parents and children), we can observe familial aggregation, but cannot be confident in ascribing
where it arises from –​is it common genes, shared exposures and experiences, or both?
In human society, there is a lot of variation in the structure and function of family units.This
variation has also provided opportunities for research designs that help to distinguish the effects
of genes from the rearing environment. The most used of these has been the twin study and its

29
30

J. Kaprio

extensions. However, there are increasing numbers of diverse family relationships, such as single-​
parent families in which the other parent has been present for only a short period of the child’s
life, if at all. On the other hand, blended families consist of a couple, the children they have had
together, and their children from previous relationships. Such families offer opportunities but
also considerable challenges for study of the behavioral determinants of physical activity and
engagement in sports.
Family studies focused on first-​degree relatives assess only overall “familiality,” i.e., the
proportion of variance attributable to both genetic and nongenetic familial influences
as well as their interactions (8). Family studies can include nuclear families (parents and
their offspring) or more-​extended pedigrees (grandparents, parents, offspring, aunts, uncles,
cousins, etc.). Exceptionally large multigenerational pedigrees can be studied with very dis-
tantly related persons descendant from early ancestors even some centuries back (6) –​such as
studies conducted among the Amish, Mormons, Icelanders, and other geographically defined
isolates with very little immigration until recent times. In larger pedigrees, there is more
scope for distinguishing genetic and nongenetic factors, but most have focused on diseases
and other conditions and there appears to be a dearth of studies of exercise characteristics
in these large pedigrees.

Quantifying family data

For quantitative, continuous traits, familial correlations can be estimated for pairs of relatives.
The intraclass correlation coefficient estimates the degree of resemblance between two family
members of the same generation (such siblings or cousins), while interclass correlation estimates
the degree of resemblance between two family members from different generations (such
as parents and children, and grandparents and children). Higher values of these correlations
are evidence of more familial resemblance; first-​degree relatives are expected to show larger
correlations than second-​degree relatives.
Categorical traits such as the presence or absence of participation in a selected sport, or
ordered categories such as the level of activity (e.g., asked as inactive, moderately active, or very
active) can be analyzed using a threshold model of liability (1). In the model, the liability to the
behavior is assumed to be normally distributed, and there are certain latent cut points that dis-
tinguish one category from the next.The latent liability is assumed to arise from multiple causa-
tive factors, each with a small effect, either genetic or environmental, which then give rise to the
variation in liability in the population. The model precludes major genes with large effects, but
these have not been found for physical activity or other exercise characteristics.The assumption
of bivariate normality (for a relative pair) can be tested when there are three or more categories
of the study variable, and the familial resemblance is then computed using a polychoric correl-
ation. For a binary trait, the assumption cannot be tested.
In order to understand the roots of family aggregation, it is necessary to distinguish between
effects arising from the rearing environment, and those arising from shared genes. By studying
nuclear families alone, this is very difficult. For example, a recent major meta-​analysis (26) of
112 studies of parent–​child physical activity derived estimates ranging from 0.19 for mother–​
son pairs to 0.29 for father–​son pairs, with parent–​daughter correlations taking intermediate
values. As parents and children share 50% of their segregating genes, the parent–​offspring correl-
ation can be doubled to yield an upper-​bound estimate of the proportion of variance ascribed
to genetic factors (i.e., 38–​58% of the population variance ascribed to genetic factors) under the
assumption that nongenetic factors play no role. However, the actual role of genetic factors may
be much smaller, and social modeling, as proposed by the authors, might be more meaningful.

30
31

Human systems genetic modeling

Without further information from other relative types, we cannot estimate the true propor-
tion of variance accounted for by genetic effects in nuclear families. One option is to extend
family studies to second-​degree relatives. However, biologically less-​related family members
(such as aunts and nephews or cousins) are also less likely to have spent time in a common
rearing environment, not permitting incisive study of relative contribution of genes and rearing
environment. Until the advent of modern molecular genetics and the ability to genotype
millions of genetic variants in large samples, adoption, twin, and twin-​family designs were
almost the only way to provide more insight about the relative contribution of genetic and
nongenetic contributions to familiality.Therefore, in order to disentangle genes and experience,
we have studied special family groups: twins who share experiences but differ in shared genes,
or adoptees and their biological and foster parents who differ in their shared experience.

Twin studies
There is a massive twin literature on human behaviors and traits of all kinds as summarized in
a recent comprehensive review article (17) indicating that nearly all studied traits have some
degree of genetic influence, but environmental influences are also ubiquitous.Thus, twin studies
have been the workhorse of behavioral genetic studies of exercise and physical activity.

Biology of twins and twinning

The biology of twinning and twins is complex. Genetically, two types of twins exist in humans.
First, monozygotic (MZ) twins, as the result of an early division of the zygote into two indi-
viduals share the same genomic sequence and hence are genetically identical, thus often called
identical twins. Whole-​genome sequencing indicates that at most only a handful of base-​pair
differences are seen between twins in young MZ pairs, but with age and in certain conditions
they do accumulate somatic mutations and other changes that may account for discordance
between twins (12, 27). Other external influences can act on them in utero to create phenotypic
differences. A notable potential source is the variation in pregnancy experiences. MZ twins,
depending on the timing of the division of the zygote, may develop with their own chorion
and placenta, or share the same placenta and chorion either with or without a common amnion.
There is little available evidence that these placentation differences affect the results of twin
studies of exercise behaviors, but few studies have rigorously assessed this due to a lack of reli-
able placentation information. After birth, multiple factors can generate both similarities and
differences between the MZ twins.
Dizygotic (DZ) twinning is more heritable than MZ twinning and arises from the simultan-
eous release and fertilization of two eggs. Thus, DZ twins are genetically full siblings and each
twin has their own placenta during pregnancy. Twin pregnancies result in the birth of the two
individuals, generally with lower birth weight and shorter gestation than singletons. Despite
the lower birth weight, twins catch up with singletons in development quickly, and as adults
are very similar to singletons in mortality and virtually all behavior measures that have been
examined. This supports the generalizability of twin study results to the general population.

The classical twin model

The realization of the existence of two types of twins more than a century ago has led to com-
parison of the similarity of MZ versus DZ twins for estimates on the role and relative contribu-
tion of genetic factors to interindividual differences in behavior. If both types of pairs are overall

31
32

J. Kaprio

more similar than two individuals chosen at random, this is evidence for familial aggregation.
The absence of differences in mean similarity of MZ versus DZ co-​twins suggests the absence
of genetic influences. As MZ twins share the same genomic sequence and DZ twins share only
50% of their segregating genes on average, then in the presence of genetic factors, MZ twins
would be expected to resemble each other for the trait in question more than DZ twins.
This inference about the role of genetic factors from the twin model holds under certain
basic assumptions. The first is that the environmental variances in MZ and DZ pairs are equal,
i.e., that MZ and DZ twins are equally correlated in the exposure to environmental experiences
and factors that are relevant for the behavior being studied. This assumption has been found to
hold generally (9). Differences could arise from placentation and in utero effects, or from differ-
ential parental treatment of MZ versus DZ twins. Assessment of relevant exposures and their
similarity in MZ and DZ pairs is needed to test the assumption, and violation of the assumption
may give rise to spuriously high estimates of genetic effects. The analysis is particularly challen-
ging as it has been shown that measures of what are often considered environmental exposures
may also have a genetic component to them. For example, smoking is considered to be an
environmental exposure, and actual cigarette smoke is a true environmental toxicant exposure.
However, there are interindividual differences in the amount smoked by smokers that relate to
genetic differences (3).
The other key assumption of the twin model is the random mating of the parents with
respect to the behavior being studied. It is assumed that the DZ pairs, like full siblings, share
50% of their segregating genes. However, if the parents resemble each other more than expected
in a particular characteristic and the characteristic has a genetic basis, then this would make
the siblings more similar than expected under random mating. For example, if both parents are
elite athletes, the biological characteristics underlying that elite ability may be expected to be
enriched in their children. This assortative mating then biases the family and twin models.

Basic concepts of the twin model

The twin model assumes that we are dealing with behaviors in which the genetic component
is due to multiple genes, each of quite small effect. For example, new studies with molecular
genetic approaches indicate that this is indeed the case for physical activity behaviors and major
genes with large effects are rare or nonexistent (25).Twin study designs, basic analyses, and mod-
eling approaches are reviewed elsewhere (4, 18, 23).
As indicated above, the genetic similarity of MZ twins is expected to be twice that of DZ
twins, and so the expected genetic correlation for additive effects of MZ pairs is unity (1.0) and
that of DZ twins is 0.5; for a derivation of the expectation and the underlying genetics, please
see, for example, Neale and Cardon (1992) or Thomas (2003) (see “Further reading”). This
source of variation in the phenotype is known as additive genetic variance (A) and it is that part
of genetic effects that is transmitted from parents to offspring. Genetic effects may also be due
to dominance (D), i.e., the sum of all nonlinear effects of alleles at a locus.The expected genetic
correlation reflecting dominance effects is unity in MZ but only 0.25 in DZ and sibling-​pairs,
and this is known as genetic variance due to dominance; dominance effects do not contribute to
parent–​offspring similarity. Finally, there may be gene–​gene interactions (i.e., epistasis) affecting
phenotypes, but the classical twin model assumes that these are not present.
Nongenetic variance in a trait is divided into that shared by the twin siblings, i.e., those
experiences and exposures that make them similar –​termed environmental effects in common
(C), and those that are not shared, i.e., unique to each twin (E). These are distinguished by
whether the effects of the experiences and exposures are shared and have equal effects on both

32
33

Human systems genetic modeling

twins, not by the actual environmental factor. Also, measurement error and random effects are
part of E (i.e., unique environment exposure), so E is included in all models.

The univariate twin model

Figure 3.1 provides a path diagram of a twin model for estimation of A, C, and E effects, com-
monly known as the ACE model. Based on the expectations listed above, it is possible to model
data from MZ and DZ twin pairs to derive estimates of the relative contribution of genetics.
This model also yields an estimate of the heritability of a trait, defined as the proportion of
total variance for a trait accounted for by genetic factors. This heritability estimate may be A/​
A+C+E but sometimes also A+D/​A+D+E. Heritability is a relative measure, so changes in the
environmental variance can change it even when no genetic changes occur. Also, heritability is
not a fixed characteristic of the behavior, but rather a population-​level estimate taken at a given
time in a given population. For an extensive discussion of the concept of heritability, there are
two excellent reviews (14, 24).
Modeling permits evaluation of which models best account for the observed variance in
a trait, providing the best statistical fit. Thus, we can evaluate which of several models fits the
data when a single behavior is looked at. The simplest model is unique environment (E), thus
rejecting all evidence for familial effects; this model rarely happens. An AE model would specify
that the pattern of twin similarity in MZ and DZ models fits a purely polygenic additive model
with environmental effects unique to each twin. It would have no shared environmental effects
(C) and no genetic effects due to dominance (D). The alternative models (CE, ACE, and ADE)
can also be specified and tested. By comparing the fit of two models, such as ACE and AE, one

rMZ=1, rDZ=0.5

1.00

E A C C A E

e a c c a e

Twin 1 Twin 2

Figure 3.1 Univariate twin model path diagram for decomposition of variance into additive genetic
factors (A), environmental factors shared by the twins (i.e., are in common C), and environmental
factors unique (E) to each twin. The three latent variables A, C, and E affect the measured phenotypes
(as shown by single-​headed arrows) of twin 1 and twin 2 (the rectangles). The C effects are by definition
fully correlated in both monozygotic (MZ) and dizygotic (DZ) twin pairs, while the expectation of
the degree of additive genetic relatedness is unity (1.00) in MZ pairs and 0.5 in DZ pairs under the
assumptions of random mating of the parents for the trait in question and satisfaction of the equal
environments assumption. In a model accounting for nonadditive effects, C is replaced by D (effects
due to dominance), and the expected correlation for D is 1.00 in MZ pairs and 0.25 in DZ pairs. The
regression coefficients (a, c, e) provide information on the degree of relationship between the latent
and observed. In the ACE model, the observed correlation for MZ pairs has an expected value (based
on the paths through A and C of both twins) of a2 + c2, and for DZ pairs 1/​2a2 + c2.Solving these two
equations provides a crude estimate of the additive genetic variance (a2).

33
34

J. Kaprio

can decide whether shared environment (C) effects are statistically needed to account for the
data.Very often the pattern of MZ and DZ correlations is used as a starting point to guide the
modeling –​in a univariate case the choice is often straightforward, but not for multivariate
models. Customized software for conducting twin modeling is available, but it is recommended
to seek the advice and guidance of an experienced statistical geneticist knowledgeable about
the model, particularly if one aims to conduct multivariate models. A number of postgraduate
courses in statistical genetics, behavioral genetics, or genetic epidemiology are regularly offered
that include twin modeling.

Multivariate twin models

At present, many types of multivariate and longitudinal quantitative genetic models are possible.
The simplest form is to run the univariate model in strata of the data, say, separately among
young and old participants or among women and men, to see if the variance components A,
C, and E differ between strata. Twin models can address sources of sex differences in more
detail if data on opposite-​sex twins have been collected. The so-​called sex-​limitation models
ask whether different sets of genes influence the behavior in men and women. If the observed
correlation for the behavior is much lower in opposite-​sex (male–​female) DZ pairs than in
same-​sex DZ pairs of either sex, this result can indicate that there are sex-​specific effects. For
example, such effects were observed in the very large twin analysis of physical activity in the
GenomEUtwin study by Stubbe et al. (21).
Multivariate models permit answering questions about the degree of shared genetic or envir-
onmental effects across related traits or over time. For example, do the same genetic and envir-
onmental effects affect physical activity in adolescence and adulthood, providing more detail on
the contributors to stability and change of the behavior than can be obtained from standard lon-
gitudinal surveys and follow-​up studies of unrelated individuals? While genes do not change in
structure over time, their expression and activity do. These alterations in expression and activity
permit novel genetic effects to arise as people age and develop. Another multivariate question
may address whether two or more facets of exercise behavior are correlated due to shared genes
or due to shared nongenetic influences. Such genetic correlations have been estimated until
recently solely by family and twin data (15), but now molecular genetic data can provide more
information on the genetic structure of related phenotypes through GWAS studies (see later
section). This information can then guide phenotype development, in terms of combining or
otherwise constructing variables with the most genetic information.
A basic starting point for multivariate modeling is the Cholesky decomposition or lower-​
triangle model. Figure 3.2 provides a schematic of the model, and the legend to the figure
provides more details. An advantage of the model is that it is solvable and provides a saturated
model against which models with fewer parameters and/​or paths can be tested. Also, it is easy to
apply to data sets if variables or measurement points of the same variable are relatively few. The
model produces genetic and environmental correlation matrices of the study variables, which
can then be compared to the genetic correlations derived from GWAS analyses. Because the
model decomposes variances in each component from left to right into more restricted factors,
the ordering of variables is important to consider.
This schematic illustrates a full model and can be used to derive matrices of the genetic and
environmental correlations between variables. These may represent different variables, such as
levels of physical activity during work, leisure, and sports (9) or the same variable at different
points in time; for an illustrative full model of four measures of physical activity from adoles-
cence to young adulthood see ­Figure 1 in Aaltonen et al. (1)

34
35

Human systems genetic modeling

A1 A2 A3
a13
a23
a11
a12 a22 a33

VAR1 VAR2 VAR3

e12 e22 e33

e11
e23
e13
E1 E2 E3

Figure 3.2 Schematic diagram of the essential feature of the Cholesky decomposition multivariate
model. The figure shows an AE model applied to three variables (VAR1,VAR2,VAR3) for which
twin data are available. The univariate twin model in Figure 3.1 is extended so that paths from the
latent genetic (A) and environmental (E) variation also cover relationships between variables. Paths
a11, a22, and a33 represent variable-​specific effect of genes on the trait (variable), and correspondingly
a12 denotes the degree to which genetic effects affecting VAR1 also affect VAR2, and so on. The
corresponding relationships hold for the environmental effects.

In addition, the full model can be tested against models that specify selected values for the
paths. For example, setting path a12 to zero would test whether there a significant genetic cor-
relation between VAR1 and VAR2.
For longitudinal models using the Cholesky approach, time/​age is the natural ordering given
that all factors impact only on current and later time points. Given its nature, the Cholesky
decomposition can accommodate different patterns of change but is not directly falsifiable. In
longitudinal contexts, the Cholesky model provides information about the relative contribution
of genetic and environmental factors to the tracking of trait.Thus, it is a useful starting point for
analyses when little is known about the longitudinal process and contributing variables. From
this exploratory tool one can expand to other longitudinal models such as the simplex model
or the longitudinal growth model (7). The longitudinal growth model allows us to ask whether
genetic factors contribute to variation in both initial level as well as rate of change of the study
variable and are those correlated. In other words, do the same genes influence both initial level
(say, initial fitness) and rate of change (improvement in fitness over a training period)? In con-
trast to the Cholesky decomposition, the longitudinal growth model enables predictions on
future time points.
Multivariate models may also be much more specific in their construct. Among the best
known are the common pathways model and the independent pathways model. The common
pathways model specifies that there is one common latent trait through which the genetic
and environment variances act. For example, there may be four or more correlated physical
activity measures. In the common pathway model, the latent variable has regression paths to all
measured variables, and this common latent variable is then decomposed into genetic and envir-
onmental factors. Thus, unlike the Cholesky, the genetic and environmental components do
not act directly on the variables but run through the common latent variable. In addition, each
variable has residual genetic and environmental components to them. The common pathway
model is a useful tool for measurement construction and theory development (for more details,
see Neale and Cardon (1992) in “Further reading”). The independent pathway model can be

35
36

J. Kaprio

considered a special case of the common pathway model. It posits three latent variables for A, C,
and E effects, such that they all correlate with the measured variables. Other models that have
been developed for use with twin data include the direction of causation model, a sibling inter-
action model, and models to assess rater bias (for more details, see Neale and Cardon (1992) in
“Further reading”).The latter can be used to assess the effect of having two more raters provide
information on several family members, typically parents reporting on both twins.
Finally, these models permit assessment of gene–​environment interactions, asking whether
a known exposure modifies the impact of genetic variation. An example is the well-​replicated
finding that physical activity buffers the impact of genes on obesity; among sedentary persons,
genes account for a much larger fraction of variance in body mass index (BMI) than among
physically active persons. This observation from twin studies (16) has now been extensively
corroborated using measured genotypes associated with BMI (11, 20).

MZ discordant pairs to study causal associations

When studying the association of a putative risk factor or exposure on an outcome, the asso-
ciation may be causal, i.e., implying that reducing the exposure would lead to a reduction in
the outcome. Alternatively, it can be due to confounding. Measurement of known confounders
and adjustment for them in statistical models has been the standard approach in observational
epidemiology, be they cohort or case–​control studies. However, not all confounders are known
or can be measured. Genetic factors underlying athletic ability may be shared with genetic
effects on obesity. Thus, the association of exercise with weight gain may be causal or it may be
accounted for by known and unknown confounders. As exercise itself is, in part, heritable and
now that genes for various aspects of exercise behavior are beginning to be identified, there is
potentially confounding due to shared genes, as illustrated by the very latest molecular genetic
studies.
A study design to examine nongenetic (possibly causal) associations that control for genetic
variation uses exposure-​discordant twin pairs. As MZ twins share the same genomic sequence,
all difference between the twins arise from nongenetic causes, taken very broadly to include all
small random events. If we can identify twin pairs in which one exercises and the other does
not, then a test of the causal hypothesis of the association between exercise and weight gain
would be to study weight development in a large number of such pairs discordant for exercise.
If the MZ co-​twins who are physically inactive have significantly more weight gain than their
co-​twins who are active, strong evidence would be provided to support a causal hypothesis.The
design controls for genetic background, but also for sex and age effects as well as the exposures
that both twins have shared, such as numerous childhood and adolescent exposures from their
common childhood home. The challenge in such studies is often finding sufficient numbers
of pairs who are truly discordant. Exposing one twin to an experimental intervention while
keeping the other twin as a control is a strong design that combines twins with an intervention.
Another design is to examine the children of discordant pairs, to see if the discordance is trans-
mitted in equal measures to the children of DZ pairs (biologically cousins) and of MZ pairs
(biologically half-​siblings but socially cousins).

Adoption studies
Theoretically, adoption studies are a very powerful design for disentangling genetic and
nongenetic influences. A biological parent–​adoptee correlation, when the adopted child has
been raised since birth by unrelated adoptive parents, is strong evidence for genetic effects.

36
37

Human systems genetic modeling

However, in practice the adoption may not have occurred immediately after birth, and exposures
during pregnancy may be transmitted epigenetically later into life. Adoptive parent–​adoptee
correlations are indicators of the effects of the rearing environment. Adoption studies of physical
activity and related traits are rare.

Twin family designs

A central limitation of the twin design is that it does not directly tell about the transmission of
effects from one generation to the next. Therefore, there are a number of designs combining
family and twins, where the twins can be either in the offspring generation or as parents.
Including two generations permits one to not only model the transmission of the study trait
from parents to offspring, but also to include the relationship of the parents. As discussed above,
the twin model assumes that there is random mating with respect to the trait being studied.
When this is not the case, the twin model results may be biased. Parents may resemble each
other more than expected if they come from a microenvironment that is more homogeneous
than in the general population. This could, for example, be a religious affiliation, social strata
(of deprivation, for example), or geographical locality. This results in what is known as social
homogamy, whereby the parents are more alike for social reasons rather than genetic. If spouses
have chosen their mates on the basis of their actual personal characteristics, this phenotypic
assortment means that there is also some degree of genetic relatedness. For example, spouses
have more similar heights than expected, i.e., height shows phenotypic assortment and this
would increase the expected genetic relatedness with respect to height in their offspring. One
concern in two-​generational studies results from the fact that the parents and offspring are often
studied at different ages, and the actual behavior may have different genetic and environmental
determinants at different ages; simple adjustment for age may not suffice. Ideally, the study
behavior should be assessed at the same age in that case, but that may mean waiting for several
decades in a longitudinal study of such families. Three-​generation studies are even rarer (2).

Genome-​wide association studies

In contrast to family studies, molecular genetic studies permit the study of the role of genetic
factors in unrelated persons. Until the sequencing of the human genome, the available tools
were limited and findings regarding the genetic basis of physical activity sparse and inconsistent.
Some candidate genes have been put forward and studied, but on their basis, it has not been
possible to assess how much human genetic variation actually contributes to interindividual
differences in exercise ability, physical activity, and related phenotypes (19). In the past decade,
the experimental design of GWAS has led to increasing findings on the role of specific genes and
genetic variation overall in many diseases as well as normal physiological and behavioral traits.
Peter Visscher and his colleagues reviewed recently the first 10 years of GWAS discoveries (25).
The design of GWAS is fairly straightforward as current genotyping technologies permit
the efficient and cheap genotyping of hundreds of thousands of genetic variants, mostly SNPs,
across the human genome. These serve as markers for blocks of DNA that are mostly inherited
together and therefore there is no need to fully sequence all the genomes under study. GWAS
is informative of a large fraction of human genetic variation and permits identification of small
regions of association. Advances in technology, statistical genetics, and genomic biology have
permitted researchers to make strong advances in the use of GWAS for understanding popu-
lation and complex-​traits genetics and the biology of the conditions. Two central advances
have been the availability of ever larger sample sizes on the one hand, and better characterized

37
38

J. Kaprio

phenotypes on the other. Unlike studies done on obesity, diabetes, height, and education which
have yielded hundreds of genetic loci linked to these traits, the study of physical activity has
been challenging, and until recently very little progress in GWAS of physical activity was
achieved (19).
Using data on nearly half a million persons from the UK Biobank (22), Tikkanen et al.
identified 64 loci associated with handgrip strength that replicated using 223,315 individ-
uals for discovery and 116,610 for replication. The strongest association was with FTO, a
locus first associated with obesity. The handgrip loci findings extended those reported earlier
in the CHARGE consortium (13) and will help to understand the reasons for variation
in handgrip strength in the population. We can expect that as other objectively measured
indices of exercise characteristics such as time spent on various physical activities become
available for genomic studies, more loci will be found due to more accuracy in the studied
phenotype. The genome-​wide significant loci for handgrip strength accounted for only
1.7% of variance in measured physical activity, but overall common variants accounted for
some 13% of variance. This is substantial but considerably less than implied by twin and
family studies.
The UK Biobank data was also at the core of another large GWAS study (10) that
extended the analyses with data from the ARIC study, with robust findings for eight loci
for moderate to vigorous physical activity. The strongest associations were seen with APOE,
a lipid and dementia-​associated gene, and with CADM2, previously associated with risk-​
taking behaviors. The authors also found that the physical activity measures show genetic
correlations with educational attainment and various obesity-​related traits. In summary, large-​
scale GWAS of self-​reported and objectively measured physical activity traits are now pro-
ducing the first consistent and replicated associations with a handful of genes. They also
show that the common genetic variants overall account for as little as 10% of the variance
in these traits, a share that may rise somewhat as more variants are genotyped and imputed
more accurately and phenotyping improves. Finally, the examined physical activity GWAS
analyses reveal expected genetic correlations with other traits such as education, obesity, and
risk-​taking behaviors.

The future in the era of molecular genetics

When considering overall genetic influences, it should be kept in mind that heritability of a
trait in a population is not necessarily static. For example, if important, influential environmental
and behavioral exposures change over time and explain more or less of the population variance
in a trait. As such, with all else being equal, there will be inverse variations in the heritability of
the trait. In addition, heritability of a trait may vary by sex and age and other covariates. While
heritability is an important concept related to the overall magnitude of genetic versus environ-
mental influences on a trait at the population level, it must be recognized that there is a highly
complex interplay between genetic and environmental factors, as is seen in epigenetics and gene
expression, for example.
Beyond heritability estimates, classic twin studies with multivariate analyses have allowed
the examination of shared genetic influences between phenotypes to test hypotheses regarding
shared etiologies or possible pathways of genetic influences on clinical phenotypes. Finally,
molecular genetics are now bringing a host of new approaches and tools to better under-
stand the genetic architecture of exercise characteristics and gain insights to relevant biological
processes.

38
39

Human systems genetic modeling

Further reading
Knopik VS,Neiderhiser JM,DeFries JC,Plomin R. Behavioral Genetics (7th ed.).New York: Worth, 2017.
Lynch M, Walsh B. Genetics and Analysis of Quantitative Traits. Sunderland, MA: Sinauer, 1998.
Neale BM, Ferreira M. AR, Medland SE and Posthuma D. Statistical Genetics: Gene Mapping Through
Linkage and Association. New York: Taylor & Francis, 2008.
Neale MC, Cardon LR. Methodology for Genetic Studies of Twins and Families. NATO ASI series
D: Behavioural and Social Sciences,Vol. 67. Dordecht: Kluwer, 1992.
Thomas DC. Statistical Methods in Genetic Epidemiology. New York: Oxford University Press, 2004.

References
1. Aaltonen S, Ortega-​Alonso A, Kujala UM, and Kaprio J. Genetic and environmental influences
on longitudinal changes in leisure-​time physical activity from adolescence to young adulthood. Twin
Research and Human Genetics 16: 535–​43, 2013.
2. Aarnio M, Winter T, Kujala UM, and Kaprio J. Familial aggregation of leisure-​time physical
activity –​a three generation study. International Journal of Sports Medicine 18: 549–​56, 1997.
3. Benowitz NL. Nicotine addiction. New England Journal of Medicine 362: 2295–​2303, 2010.
4. Boomsma D, Busjahn A, and Peltonen L. Classical twin studies and beyond. Nature Reviews
Genetics 3: 872–​82, 2002.
5. de Geus EJ, Bartels M, Kaprio J, Lightfoot JT, and Thomis M. Genetics of regular exercise and
sedentary behaviors. Twin Research and Human Genetics 17: 262–​71, 2014.
6. Kaplanis J, Gordon A, Shor T, Weissbrod O, Geiger D, Wahl M, Gershovits M, Markus B,
Sheikh M, Gymrek M, Bhatia G, MacArthur DG, Price AL, and Erlich Y. Quantitative analysis
of population-​scale family trees with millions of relatives. Science 360: 171–​175, 2018.
7. Kaprio J and Silventoinen K. Advanced methods in twin studies. Methods in Molecular Biology
713: 143–​52, 2011.
8. Kendler KS and Neale MC. “Familiality” or heritability. Archives of General Psychiatry 66: 452–​3, 2009.
9. Kendler KS, Neale MC, Kessler RC, Heath AC, and Eaves LJ. A test of the equal-​environment
assumption in twin studies of psychiatric illness. Behavior Genetics 23: 21–​7, 1993.
10. Klimentidis YC, Raichlen DA, Bea J, Garcia DO, Wineinger NE, Mandarino LJ, Alexander
GE, Chen Z, and Going SB. Genome-​wide association study of habitual physical activity in
over 377,000 UK Biobank participants identifies multiple variants including CADM2 and APOE.
International Journal of Obesity 42: 1161–​1176, 2018.
11. Kilpeläinen TO, Qi L, Brage S, Sharp SJ, Sonestedt E, Demerath E, Ahmad T, Mora S,
Kaakinen M, Sandholt CH, Holzapfel C, Autenrieth CS, Hyppönen E, Cauchi S, He M,
Kutalik Z, Kumari M, Stančáková A, Meidtner K, Balkau B, Tan JT, Mangino M, et al.
Physical activity attenuates the influence of FTO variants on obesity risk: a meta-​analysis of 218,166
adults and 19,268 children. PLoS Medicine 8: e1001116, 2011.
12. Martin N, Boomsma D, and Machin G. A twin-​pronged attack on complex traits. Nature Genetics
17: 387–​92, 1997.
13. Matteini AM, Tanaka T, Karasik D, Atzmon G, Chou WC, Eicher JD, Johnson AD, Arnold
AM, Callisaya ML, Davies G, Evans DS, Holtfreter B, Lohman K, Lunetta KL, Mangino M,
Smith AV, Smith JA, Teumer A, Yu L, Arking DE, Buchman AS, Chibinik LB, et al. GWAS
analysis of handgrip and lower body strength in older adults in the CHARGE consortium. Aging Cell
15: 792–​800, 2016.
14. Mayhew AJ and Meyre D. Assessing the heritability of complex traits in humans: methodological
challenges and opportunities. Current Genomics 18: 332–​340, 2017.
15. Mustelin L, Joutsi J, Latvala A, Pietiläinen KH, Rissanen A, and Kaprio J. Genetic influences
on physical activity in young adults: a twin study. Medicine and Science in Sports and Exercise
44: 1293–​301, 2012.
16. Mustelin L, Silventoinen K, Pietiläinen K, Rissanen A, and Kaprio J. Physical activity reduces
the influence of genetic effects on BMI and waist circumference: a study in young adult twins.
International Journal of Obesity 33: 29–​36, 2009.
17. Polderman TJ, Benyamin B, de Leeuw CA, Sullivan PF, van Bochoven A, Visscher PM, and
Posthuma D. Meta-​analysis of the heritability of human traits based on fifty years of twin studies.
Nature Genetics 47: 702–​9, 2015.

39
40

J. Kaprio

18. Posthuma D, Beem AL, de Geus EJ, van Baal GC, von Hjelmborg JB, Iachine I, and
Boomsma DI. Theory and practice in quantitative genetics. Twin Research 6: 361–​76, 2003.
19. Sarzynski MA, Loos RJ, Lucia A, Pérusse L, Roth SM, Wolfarth B, Rankinen T, and
Bouchard C. Advances in exercise, fitness, and performance genomics in 2015. Medicine and Science
in Sports and Exercise 48: 1906–​16, 2016.
20. Shungin D, Deng WQ, Varga TV, Luan J, Mihailov E, Metspalu A; GIANT Consortium,
Morris AP, Forouhi NG, Lindgren C, Magnusson PKE, Pedersen NL, Hallmans G, Chu
AY, Justice AE, Graff M, Winkler TW, Rose LM, Langenberg C, Cupples LA, Ridker PM,
Wareham NJ, et al. Ranking and characterization of established BMI and lipid associated loci as
candidates for gene-​environment interactions. PLoS Genetics 13: e1006812, 2017.
21. Stubbe JH, Boomsma DI, Vink JM, Cornes BK, Martin NG, Skytthe A, Kyvik KO, Rose
RJ, Kujala UM, Kaprio J, Harris JR, Pedersen NL, Hunkin J, Spector TD, and de Geus EJ.
Genetic influences on exercise participation in 37,051 twin pairs from seven countries. PLoS One
1: e22, 2006.
22. Tikkanen E, Gustafsson S, Amar D, Shcherbina A, Waggott D, Ashley EA, and Ingelsson
E. Biological insights into muscular strength: Genetic findings in the UK Biobank. Scientific Reports
8: 6451, 2018.
23. van Dongen J, Slagboom PE, Draisma HH, Martin NG, and Boomsma DI. The continuing
value of twin studies in the omics era. Nature Reviews Genetics 13: 640–​53, 2012.
24. Visscher PM, Hill WG, and Wray NR. Heritability in the genomics era –​concepts and
misconceptions. Nature Reviews Genetics 9: 255–​66, 2008
25. Visscher PM, Wray NR, Zhang Q, Sklar P, McCarthy MI, Brown MA, and Yang J. 10
Years of GWAS discovery: Biology, function, and translation. American Journal of Human Genetics
101: 5–​22, 2017.
26. Yao CA and Rhodes RE. Parental correlates in child and adolescent physical activity: A meta-​
analysis. International Journal of Behavioral Nutrition and Physical Activity 12: 10, 2015.
27. Ye K, Beekman M, Lameijer EW, Zhang Y, Moed MH, van den Akker EB, Deelen J,
Houwing-​Duistermaat JJ, Kremer D, Anvar SY, Laros JF, Jones D, Raine K, Blackburne
B, Potluri S, Long Q, Guryev V, van der Breggen R, Westendorp RG, ‘t Hoen PA, den
Dunnen J, van Ommen GJ, et al. Aging as accelerated accumulation of somatic variants: whole-​
genome sequencing of centenarian and middle-​aged monozygotic twin pairs. Twin Research and
Human Genetics 16: 1026–​32, 2013.

40
41

4
THE TRANSLATION OF
SYSTEMS GENETICS OF
EXERCISE TO EVERYDAY LIFE
Molly Bray

Introduction
Whether in the context of cardiovascular fitness, weight loss/​maintenance, metabolic disease
risk, or mental health, the benefits of physical activity and exercise for most individuals are
undeniable. Conversely, physical inactivity is among the leading causes of death in the US,
second only to tobacco use (5). Sedentary behavior has been shown to have a significant dose–​
response effect in increasing risk for cardiovascular disease, type 2 diabetes, obesity, and all-​cause
mortality (47, 71). In adults aged 40–​69 years, almost 10% of deaths have been attributed directly
to insufficient levels of physical activity (14). Despite the health benefits associated with physical
activity, many individuals find it difficult to maintain participation in regular physical activity,
regardless of their interest or intent. Approximately 25% of adults do not engage in any leisure-​
time physical activity (45), and only 21% of adults meet the 2008 physical activity guidelines of
2.5 or more hours per week, with the lowest prevalence of physical activity participation among
Hispanics and non-​Hispanic black individuals (16). How might an understanding of the genetic
underpinnings of physical activity and exercise response be used to improve intervention effi-
cacy and turn the rising tide of sedentarism in the world? A primary goal of exercise genomics
is to determine what makes a person “naturally” physically active and to use this knowledge to
help others to achieve this natural state through enhanced interventions and more efficacious
and personalized exercise prescription.

Theoretical foundations of exercise interventions

Physical activity and exercise are complex behaviors driven by both extrinsic and intrinsic
factors. Attitudes, perceived barriers, enjoyment, self-​efficacy, and self-​motivation have all been
associated with exercise adherence (9, 18, 46, 63).Various theoretical frameworks explain the asso-
ciation between these concepts and exercise adherence or dropout in different ways. According
to social-​cognitive theory, those who have high self-​efficacy about physical activity would per-
ceive fewer barriers to their physical activity or be less influenced by them, be more likely to act
on their expectations of desirable outcomes of being physically active (i.e., extrinsic motives),
and be more likely to enjoy physical activity (intrinsic motives; 3, 4). Self-​determination theory
and schema theory are complementary to social-​cognitive theory and suggest that intrinsic

41
42

M. Bray

motives for physical activity and a physically active self-​identity, which develop during late
adolescence and young adulthood, interact with physical and social environments to influence
physical activity (54). The transtheoretical model proposes that when people try to increase
their physical activity, they use neurocognitive processes to move through stages of change
(48). The theory of reasoned action states that behaviors are initially determined by intentions,
which are in turn driven by both extrinsic motives for physical activity, such as social affiliation
and appearance, and intrinsic motives, such as physical competence and enjoyment (15). Most
interventions targeted at improving health through changes in diet and/​or physical activity
are grounded in behavior theory, which can be useful for formulating strategies for changing
behavior but generally lack a biological component. Exercise genomics can inform intervention
studies by identifying the genes and pathways that underlie and drive the theoretical behavior.
A common observation among exercise and physical activity intervention studies is a high
level of initial compliance and concomitant positive responses to exercise, followed by poor
long-​term adherence and eventual dropout. Many exercise programs have reported a 50–​60%
dropout rate within the first 3–​6 months (19, 20, 67). Variables reported to be associated with
exercise dropout include age, education, sex, ethnicity, previous activity, dietary habits, smoking,
occupation, and social support, but no single variable explains all the variance in subject attrition
(9, 12, 52, 63, 66). Additional research has shown that exercise dropouts are more likely to have
higher body fat and body weight than those who continue with an exercise program (28); con-
versely, participants who are physically active and have a moderate level of aerobic fitness are the
most likely to continue to exercise (66). Rothman (53) notes that behavior change differs from
behavior maintenance or adherence. Factors driving the initiation of a behavior are strongly
linked to anticipated positive effects, while the factors contributing to behavior maintenance are
strongly dependent upon perceived satisfaction with the outcome of the behavioral change (53).
Though multiple psychological and physical traits have been identified that may be predictive
of exercise and physical activity behavior, what is lacking in these investigations is the incorp-
oration of putative biological drivers of behavior, such as neural signaling and metabolism that
may be revealed through genetic studies.

Translating exercise genomics into practice

As described above, behavioral theories can provide a strong foundation for the development
of physical activity-​based interventions. As shown in Figure 4.1, understanding the biological
processes that underlie observed behavior gleaned from genetic studies has the potential to
greatly enhance intervention efficacy through: 1) improvement of exercise adherence; 2) opti-
mization of exercise prescriptions; and 3) early identification of individuals at risk for dropout.
In terms of exercise adherence, genetic studies can be used to identify the neural pathways
that influence motivation, persistence, and attitudes toward exercise. To optimize exercise pre-
scription, genes identified for physical activity may reveal new information about biological
processes that enhance or inhibit positive physiological responses to exercise, which in turn
influence exercise adherence. Genetic variation may also explain the variability in response to
environmental changes. For example, while strategies such as modifying the built environment
to encourage physical activity are important at a community level, genetic information may
reveal how and why some individuals are drawn to use environmental resources such as parks,
cycle lanes, and trails and some are not. Importantly, genetic information may be extremely
useful in identifying those individuals who may be at risk for dropout or nonresponse and, thus,
may need more encouragement or instruction to successfully incorporate physical activity as a
lifestyle.

42
43

Systems genetics of exercise

Figure 4.1 Genetic information can be used to: 1) improve exercise adherence; 2) optimize exercise
prescriptions; and 3) identify individuals at risk for dropout.

Using genetic information to enhance exercise adherence

Like most dietary interventions, exercise programs are generally effective when people stick
to them, and the best way to increase intervention efficacy is through improved adherence.
Adherence can be defined as completion of or compliance with a prescribed protocol or course
of action. As described above, attitudes and beliefs play a substantial role in physical activity
behavior and exercise dropout, and at least some of the heritability of physical activity may
represent genetic variation associated with the neural processing underlying exercise attitudes
and adherence. Huppertz and colleagues examined the heritability of attitudes toward exercise
in adult twin pairs, with a focus on six domains: perceived benefits; lack of skills, support, and/​or
resources; time constraints; lack of energy; lack of enjoyment; and embarrassment (34). All atti-
tude domains were significantly correlated with exercise behavior and exhibited significant her-
itability. Interestingly, the highest heritabilities observed were for intrinsic factors such as lack of
enjoyment (males: 0.47; females: 0.44), embarrassment (males: 0.42; females: 0.49), and lack of
skills (males: 0.45; females: 0.48), while external factors such as time constraints and perceived
benefits were more strongly explained by environmental factors in both men and women (0.70–​
0.79). Together, these attitudes explained 28% of variance in exercise behavior (34). Heritability
studies provide support for a genetic basis for the psychological constructs associated with
physical activity, and ultimately, it is important to identify the genes and pathways underlying
the genetic basis of physical activity behavior in order to make more informed decisions about
how to intervene.
Altering physical and behavioral traits through selective animal breeding is one of the
most powerful experimental approaches for demonstrating that a trait is genetic. Genes

43
44

M. Bray

that underlie voluntary physical activity in animals or humans likely influence both phys-
ical (e.g., exercise performance and ability) and psychological (e.g., motivation, adherence)
components of exercise behavior, which reinforce each other in a feed-​forward manner (35).
As described in Chapter 5, a number of animal models have been created with extremely high
levels of physical activity, with rodents running both further and faster than their wildtype
counterparts (17, 21). Interestingly, animals bred to be physically active demonstrate increased
signs of stress when the running wheel is removed, enhanced preference for running in lieu
of other activities, and increased neuronal activity associated with physical activity, supporting
the hypothesis that neurocognitive traits can be selected for in the breeding process (51).
Though a limited number of syntenic quantitative trait loci regions between mouse and
human have been identified through gene mapping studies in selectively bred animals, it is
surprising how little overlap has been reported, although critical differences in study design
may have contributed to this lack of replication (38). The lack of replication for genes iden-
tified in animal studies points to the tremendous importance of considering genetic back-
ground, environmental cues and conditions, stress, and other mental states when assessing
physical activity behavior.
Many genes have been identified that appear to regulate both spontaneous and planned physical
activity, as well as exercise response. As outlined in Chapter 8, genes identified to date provide com-
pelling evidence that physical activity behavior is under substantial neural control.Variation within
the dopaminergic, serotonergic, and endocannabinoid systems may alter satisfaction, pleasure, and
reward derived from exercise and physical activity, and in turn, influence persistence in phys-
ical activity behavior. In support of this hypothesis, Bryan et al. demonstrated that affective state,
perceived exertion, mood, and heart rate were a function of whether and how long a person had
been performing an acute bout of exercise and that changes in these attitudes were mediated by
a variant (Val66Met, rs6265) in the brain-​derived neurotropic factor (BDNF) gene (11). Positive
affect increased with greater duration of acute exercise, and carriers of the BDNF Met66 allele
showed a more pronounced increase in positive affect compared to Val/​Val homozygotes. Increases
in heart rate during exercise were also associated with both alterations in mood and variation in
BDNF, suggesting that affective mood may be influenced by the interaction between physiologic
changes and genetic variation (11). In a follow-​up study, the BDNF Met66 allele was also associated
with intrinsic motivation for exercise, with carriers being more likely to remain on the treadmill,
even when given the option to discontinue the exercise session (13).
Although many studies have examined the interaction between genes and physical activity/​
exercise on disease risk, energy balance, and exercise response, few studies have been specific-
ally designed to examine how genetic variation influences exercise adherence. Thompson et al.
examined an insertion/​deletion (I/​D) polymorphism in the angiotensin-​converting enzyme
(ACE) gene for an association with exercise response and adherence, which was defined by
session attendance in a progressive aerobic exercise-​training program.While the ACE I/​D poly-
morphism was not associated with exercise response (in terms of aerobic fitness, anthropomet-
rics, or blood pressure), adherence to the exercise training protocol was significantly higher in
I carriers than in D homozygotes (61). No differences were observed in average exercise inten-
sity between the I/​D genotypes, however, suggesting that simply accounting for attendance
without considering other parameters of the workout sessions does not sufficiently define
adherence (61). The Training Interventions and Genetics of Exercise Response (TIGER) study
was developed to investigate the genetic underpinnings of exercise adherence and response
in young adults (55). A total of 3665 subjects in the study underwent 15 weeks of aerobic
exercise training, for three days per week for at least 30 minutes at an intensity between 65%

44
45

Systems genetics of exercise

and 85% of maximum heart rate reserve. Every exercise session was documented and tracked
using computerized heart rate monitors. Compliance with the study protocol was quantified
by adjusting the duration of each exercise session by the average intensity of the session and
summing over all sessions to create a total score of exercise intensity-​minutes, called the heart
rate physical activity score (HRPAS) (44). Exercise adherence was defined as meeting the min-
imum prescribed HRPAS. HRPAS is also considered a measure of exercise tolerance, since it is
a function of how hard and how long the participants exercised within the range of the exercise
prescription. Whole-​genome association analysis using HRPAS as a primary outcome identi-
fied SNPs in ten genes that exceeded a genome-​wide significance of P <10−4.5, including the
fructosamine 3 kinase related protein (FN3KRP), brain-​derived neurotropic factor (BDNF),
fat mass and obesity related transcript (FTO), receptor interacting serine/​threonine kinase 2
(RIPK2), ATP binding cassette subfamily B member 11 (ABCB11), brain and reproductive
organ-​expressed (BRE), zinc fingers and homeoboxes 3 (ZHX3), insulin degrading enzyme
(IDE), tubulin folding cofactor D (TBCD), and centrosomal protein 112 (CEP112) genes (30).
Pathways contributing to lipid metabolism, neural signaling, muscle contraction, and adiposity
were significantly represented by SNPs with a nominal P <0.0001. The strongest single gene
association with exercise adherence/​tolerance was for FN3KRP (P <10−11), which functions
in gene regulation through deglycation of target proteins (60). Though no other studies to
date have reported associations between FN3KRP and physical activity or exercise, it has been
identified in gene mapping studies for multiple cardiometabolic risk traits (body mass index,
waist, hip, systolic blood pressure) (7, 56), and allele-​specific expression of FN3KRP driven
by miR-​34a is associated with variation in hemoglobin A1c levels (29), supporting its role in
glycation-​based regulation. The functional role of the FN3KRP gene in exercise adherence
is not yet known, and it is possible that it is in disequilibrium with other functional genes or
regulatory sequence.
At the far end of the exercise adherence spectrum is exercise addiction. Similar to other
types of addiction, exercise addiction is associated with increasing tolerance, withdrawal anx-
iety, lack of control, and reduction of other activities (27). As described in earlier chapters,
exercise is known to excite areas of the brain and systems associated with pleasure and reward,
such as dopamine, endorphins, and cannabinoids, in a manner similar to that of addictive
substances. Genes that have been associated with physical activity, including serotonin receptor
2A (HTR2A), serotonin receptor 1B (HTR1B), dopamine receptor D2 (DRD2), and dopamine
receptor D4 (DRD4), have also been associated with alcohol, cocaine, and opioid dependence
(31). Strategies to reduce addictive behavior in other areas may inform the conversion of exer-
cise addicts to a normal level of adherence. In addition, such studies also have the potential to
reveal important mechanisms that motivate behavior.
What is currently lacking in terms of genetic studies of exercise adherence are prospective
studies of individuals selected on genotype and powered to detect a difference between genotype
groups to validate association studies. One difficulty in identifying genes for exercise adherence
retrospectively in existing cohorts is that subjects are often urged to remain in the interven-
tion by extrinsic rewards or motivation such as monetary compensation and/​or prompting and
encouragement from study investigators, which may influence the natural process of attrition.
To identify and replicate genes for exercise adherence, studies should be performed in which
optimal exercise interventions are provided and subjects are allowed to drop from the study
without external interference. In addition, future genetic studies should examine the interaction
between genes and other factors, such as those listed above, that have been associated with exer-
cise dropout to simultaneously characterize the multivariate predictors of exercise adherence.

45
46

M. Bray

Using genetic information to enhance exercise prescription and intervention

Currently, exercise prescription is a largely a one-​size-​fits-​all undertaking, with the argument
being that “exercise is good for everyone.” Nevertheless, significant variability has been observed
in response to exercise training, with some individuals experiencing negative, rather than posi-
tive effects of exercise, depending on the outcome being studied (10). As identified in other
chapters, numerous genes have been associated with physical activity and exercise behavior,
providing potential targets for developing gene-​based exercise prescriptions. In addition, the
interaction between genes and exercise has been robustly demonstrated to influence response to
exercise training for multiple traits (muscular strength and power, cardiorespiratory fitness and
endurance performance, body weight and adiposity, insulin and glucose metabolism, lipid and
lipoprotein metabolism, and hemodynamic traits) (57). As with many other complex traits, the
genes identified to date explain only a small proportion of the total variability in the exercise-​
related outcomes studied, and many have suggested that much larger cohorts than those cur-
rently being examined are needed to robustly develop a set of genetic variants that can predict
exercise adherence and response (24).
In developing interventions targeted at improving health behaviors, the systems genetics of
exercise points to the fact that genes play a plausible role in influencing many components of
health behaviors. Reiss and colleagues note that genes can modify how individuals respond to
the social and physical environments by influencing response to environmental stress, sensi-
tivity to changes in the environment, compatibility with the environment, and environment-​
specific responses (50). Thus, changing the built environment to promote physical activity may
cause stress in some individuals and excitement or challenge in others. For example, the sero-
tonin system, which has been shown to be involved in physical activity behavior, has also been
implicated in differential cardiovascular reactivity in response to social stress (65), providing a
potential link between physical activity behavior and environmentally induced stress. Making
assumptions that individuals who do not take advantage of such environmental alterations are
simply lazy or unmotivated ignores a potential key biological underpinning of inactivity.
While animal studies provide evidence that disrupting candidate genes may influence phys-
ical activity and other metabolic traits, few human studies have been designed to directly test the
effects of exercise training by selecting subjects based on genotype. In a study by Thompson and
colleagues, balanced groups of subjects were selected based on their apolipoprotein E (APOE)
genotype to have sufficient power to detect differences in lipid response to exercise training (62).
The investigators were able to demonstrate significant differences between genotype groups in
exercise response, not only in lipid traits (total cholesterol/​high-​density lipoprotein ratio, low-​
density lipoprotein/​high-​density lipoprotein ratio) but also in aerobic capacity, with significantly
lower VO2max in subjects with the 3/​3 genotype compared to the other genotype groups (62).
Similar to other studies, the BDNF signaling pathway emerged as a central factor linking
multiple other pathways, and highlighting neural signaling as a target for exercise adherence in
the TIGER study (30). BDNF provides an example of the potential use of genetic information
to guide the development of behavioral exercise interventions as well as to direct the use of
exercise and physical activity for therapy for other types of conditions. BDNF is a neurotrophin
that has been shown to play a central role in neuronal plasticity and long-​term potentiation
associated with learning and memory (40). The Val66Met variant in the BDNF gene has been
associated with depression and multiple neurodegenerative disorders (e.g., Parkinson’s disease,
multiple sclerosis, Alzheimer’s disease; 37, 59), in addition to physical activity. Importantly, both
acute and chronic aerobic exercise have been shown to increase peripheral BDNF expression
(33), and exercise has been shown to attenuate the negative effects of BDNF sequence variation

46
47

Systems genetics of exercise

on memory loss, depression, and hippocampal atrophy (23). Increases in BDNF expression
associated with exercise training have been proposed as one potential mechanism for the action
of exercise in reducing depressive symptoms, enhancing self-​esteem, improving mood and sleep,
and increasing resilience to stress (25, 33), making BDNF not only a plausible candidate gene
but also a likely target for intervention. One way to apply this knowledge would be to use
BDNF as a marker to identify individuals who are sensitive to the effects of exercise versus those
who are not. This could involve measuring plasma levels of BDNF or measuring gene variants
in BDNF; given new technologies for characterizing genetic variation, the latter might be the
easier of the two. As in the Thompson study (62), subjects could be selected based on BDNF
protein levels or genotype, and then examined for exercise response and adherence a priori. Such
a strategy would represent a translational approach to a critical issue in exercise research.

Using genetic information to identify individuals at risk for exercise dropout

Of all the ways in which genetic information may be utilized to improve health, genetic risk
prediction is among the most straightforward. Genetic risk alleles/​scores have been shown to
inform risk prediction for cardiovascular disease (e.g., markers at 9p21 for incident coronary
artery disease), cancer (e.g., BRCA1 and BRCA2 for breast cancer), venous thromboembolism
(e.g., factor V Leiden), drug response (e.g., multilocus risk score at NAT2, CYP2D6, CYP2C19,
CYP2C9, and CYP4F2 for coumarin response), Alzheimer’s disease (e.g., APOE), and others
(32, 39, 68, 70), with large-​effect alleles providing the most reproducible and stable predictions
of disease risk. Nevertheless, for complex diseases and traits, such as exercise adherence or
dropout, in which multiple small-​effect variants likely explain most of the variance, genetic pre-
diction has been less robust. Despite the fact that many genome-​wide association study (GWAS)
findings have been convincingly replicated, Marigorta and colleagues caution that even when
combining all functional SNPs for a complex trait into a genetic risk score, the ability to predict
risk is limited by the heritability of the trait (43). Nevertheless, GWAS data could be combined
with other types of “omics” data (e.g., transcriptome, methylome, metabolome, and phenome)
to better understand biological interactions and more robustly predict behavioral outcomes like
exercise adherence and dropout.
In the TIGER study (described previously), the FTO gene was associated with exercise
adherence, with combinations of any three risk-​raising alleles in the FTO locus more than
doubling exercise dropout risk (odds ratio=2.22; 95% confidence interval: 1.45–​3.40) (30).
In a follow-​up study, subjects were selected based on genetic risk score for FTO (less than
three risk-​raising alleles or more than six risk-​raising alleles), and differences in gene expression
from leukocytes before and after exercise and between groups were examined. After control-
ling for multiple testing, the diacylglycerol kinase zeta (DGKZ) gene was found to be signifi-
cantly differentially expressed between pre-​and postexercise time points in the high-​r isk group
(P <0.04) (22). Interestingly, DGKZ has been implicated in muscle hypertrophy, insulin sen-
sitivity, energy metabolism, and addictive behavior (8, 42, 69), providing a plausible biological
mechanism for its potential action in exercise adherence. How the effect of DGKZ is modified
by variation in FTO is not yet known.
To date, exercise and physical activity have been primarily assessed as behavioral/​lifestyle
means of reducing genetic risk for metabolic diseases, rather than in the context of genetic risk
for the behaviors themselves. For example, the FTO gene is one of the most highly replicated
findings in obesity genetics (26, 58), and it has also repeatedly been reported that physical
activity attenuates the increased risk associated with deleterious FTO variation (1, 49, 64).
Physical activity has been shown to substantially reduce multigenic disease risk for obesity

47
48

M. Bray

and coronary heart disease (36, 41) and to be more important in determining risk than gen-
etic factors for type 2 diabetes (2, 6). These examples demonstrate that health behaviors are
important in the context of genetic risk and that people are not “doomed” by their DNA
sequence. Because healthy behaviors such as diet and exercise can offset the risk-​raising effect of
gene variation, it is more important than ever to develop a better understanding of the factors
that influence exercise adherence.

Conclusion
The systems genomics approach has been extremely effective in identifying genes associated
with physical activity behavior and exercise response, many of which have been robustly
replicated. The list of potential candidate genes associated with physical activity behavior and
exercise response is extensive and growing with each new study. Genes identified to date point
to a strong neural basis for exercise behavior that is supported by studies of the genetics of
exercise adherence. To translate the findings from exercise and physical activity studies to inter-
vention and practice, statistical replication should be followed by functional analysis in the
laboratory along with well-​powered studies in humans to test the effects and interactions of
genotype a priori. The message that “exercise and physical activity are good for everyone” has
clearly not been effective in promoting physical activity, and a stronger basis in biology is needed
to devise more efficacious and personalized exercise interventions and prescriptions. Given the
success of genetic risk prediction in many areas and the increasing ease with which genetic data
can be collected, it is reasonable to envision a straightforward genetic testing protocol that can
be used to identify individuals at risk for dropout or nonresponse.

Acknowledgement: Thanks to Matthew Lehrer, MS, for his tremendous help with summar-
izing the references for this chapter.

References
1. Andreasen CH, Mogensen MS, Borch-​Johnsen K, Sandbaek A, Lauritzen T, Sorensen TI,
Hansen L, Almind K, Jorgensen T, Pedersen O, and Hansen T. Non-​replication of genome-​wide
based associations between common variants in INSIG2 and PFKP and obesity in studies of 18,014
Danes. PLoS One 3: e2872, 2008.
2. Ardisson Korat AV, Willett WC, and Hu FB. Diet, lifestyle, and genetic risk factors for type 2
diabetes: A review from the Nurses’ Health Study, Nurses’ Health Study 2, and Health Professionals’
Follow-​up Study. Curr Nutr Rep 3: 345–​354, 2014.
3. Bandura A. Health promotion by social cognitive means. Health Educ Behav 31: 143–​164, 2004.
4. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs,
NJ: Prentice Hall, 1986.
5. Bauer UE, Briss PA, Goodman RA, and Bowman BA. Prevention of chronic disease in the 21st
century: elimination of the leading preventable causes of premature death and disability in the USA.
Lancet 384: 45–​52, 2014.
6. Bellou V, Belbasis L, Tzoulaki I, and Evangelou E. Risk factors for type 2 diabetes mellitus: An
exposure-​wide umbrella review of meta-​analyses. PLoS One 13: e0194127, 2018.
7. Benton MC, Lea RA, Macartney-​Coxson D, Carless MA, Goring HH, Bellis C, Hanna M,
Eccles D, Chambers GK, Curran JE, Harper JL, Blangero J, and Griffiths LR. Mapping eQTLs
in the Norfolk Island genetic isolate identifies candidate genes for CVD risk traits. Am J Hum Genet
93: 1087–​1099, 2013.
8. Benziane B, Borg ML, Tom RZ, Riedl I, Massart J, Bjornholm M, Gilbert M, Chibalin AV,
and Zierath JR. DGKzeta deficiency protects against peripheral insulin resistance and improves
energy metabolism. J Lipid Res 58: 2324–​2333, 2017.

48
49

Systems genetics of exercise

9. Booth ML, Owen N, Bauman A, Clavisi O, and Leslie E. Social-​cognitive and perceived envir-
onment influences associated with physical activity in older Australians. Prev Med 31: 15–​22, 2000.
10. Bouchard C and Rankinen T. Individual differences in response to regular physical activity. Med Sci
Sports Exerc 33: S446–​S451; discussion S452–​S443, 2001.
11. Bryan A, Hutchison KE, Seals DR, and Allen DL. A transdisciplinary model integrating genetic,
physiological, and psychological correlates of voluntary exercise. Health Psychol 26: 30–​39, 2007.
12. Buckworth J and Wallace LS. Application of the transtheoretical model to physically active adults.
J Sports Med Phys Fitness 42: 360–​367, 2002.
13. Caldwell Hooper AE, Bryan AD, and Hagger MS. What keeps a body moving? The brain-​
derived neurotrophic factor val66met polymorphism and intrinsic motivation to exercise in humans.
J Behav Med 37: 1180–​1192, 2014.
14. Carlson SA, Adams EK, Yang Z, and Fulton JE. Percentage of deaths associated with inadequate
physical activity in the United States. Prev Chronic Dis 15: E38, 2018.
15. Chatzisarantis N and Brickell T. Functional signifcance of psychological variables that are included
in the theory of planned behaviour: A self-​determination theory approach to the study of attitudes,
subjective norms, perceptions of control and intentions. Eur J Soc Psychol 28: 303–​322, 1998.
16. Centers for Disease Control. Facts about physical activity. www.cdc.gov/​physicalactivity/​data/​
facts.htm, 2014.
17. Copes LE, Schutz H, Dlugosz EM, Acosta W, Chappell MA, and Garland T, Jr. Effects of
voluntary exercise on spontaneous physical activity and food consumption in mice: Results from an
artificial selection experiment. Physiol Behav 149: 86–​94, 2015.
18. Dishman RK. Biologic influences on exercise adherence. Res Q Exerc Sport 52: 143–​159, 1981.
19. Dishman RK. Exercise compliance: A new view for public health. Phys Sportsmed 14: 127–​145, 1986.
20. Dishman RK and Ickes W. Self-​motivation and adherence to therapeutic exercise. J Behav Med
4: 421–​438, 1981.
21. Dlugosz EM, Chappell MA, McGillivray DG, Syme DA, and Garland T, Jr. Locomotor trade-​
offs in mice selectively bred for high voluntary wheel running. J Exp Biol 212: 2612–​2618, 2009.
22. Dong F, Vazquez A, Dishman R, O’Connor P, Jackson A, Fernandez J, and Bray M. The
interaction between FTO variation and exercise in influencing gene expression and DNA methyla-
tion. In preparation.
23. Erickson KI, Miller DL, and Roecklein KA. The aging hippocampus: interactions between exer-
cise, depression, and BDNF. Neuroscientist 18: 82–​97, 2012.
24. Eynon N, Voisin S, Lucia A, Wang G, and Pitsiladis Y. Preface: Genomics and biology of exercise
is undergoing a paradigm shift. BMC Genomics 18: 825, 2017.
25. Fox KR. The influence of physical activity on mental well-​being. Public Health Nutr 2: 411–​418, 1999.
26. Frayling TM. Genome-​wide association studies provide new insights into type 2 diabetes aetiology.
Nat Rev Genet 8: 657–​662, 2007.
27. Freimuth M, Moniz S, and Kim SR. Clarifying exercise addiction: Differential diagnosis, co-​
occurring disorders, and phases of addiction. Int J Environ Res Public Health 8: 4069–​4081, 2011.
28. Gale JB, Eckhoff WT, Mogel SF, and Rodnick JE. Factors related to adherence to an exercise
program for healthy adults. Med Sci Sports Exerc 16: 544–​549, 1984.
29. Ghanbari M, Franco OH, de Looper HW, Hofman A, Erkeland SJ, and Dehghan A. Genetic
variations in microRNA-​ binding sites affect microRNA-​ mediated regulation of several genes
associated with cardio-​metabolic phenotypes. Circ Cardiovasc Genet 8: 473–​486, 2015.
30. Herring M, Vazquez A, Dong F, Dishman R, O’Connor P, Jackson A, and Bray M. Genome-​
wide association for exercise adherence and tolerance in the TIGER study. In preparation.
31. Herring MP, Sailors MH, and Bray MS. Genetic factors in exercise adoption, adherence and
obesity. Obes Rev 15: 29–​39, 2014.
32. Huang SS, Liu Y, Jing ZC, Wang XJ, and Mao YM. Common genetic risk factors of venous
thromboembolism in Western and Asian populations. Genet Mol Res 15: 15017644, 2016.
33. Huang T, Larsen KT, Ried-​Larsen M, Moller NC, and Andersen LB. The effects of physical
activity and exercise on brain-​derived neurotrophic factor in healthy humans: A review. Scand J Med
Sci Sports 24: 1–​10, 2014.
34. Huppertz C, Bartels M, Jansen IE, Boomsma DI, Willemsen G, de Moor MH, and de Geus
EJ. A twin-​sibling study on the relationship between exercise attitudes and exercise behavior. Behav
Genet 44: 45–​55, 2014.
35. Kelly SA and Pomp D. Genetic determinants of voluntary exercise. Trends Genet 29: 348–​357, 2013.

49
50

M. Bray

36. Khera AV, Emdin CA, Drake I, Natarajan P, Bick AG, Cook NR, Chasman DI, Baber U,
Mehran R, Rader DJ, Fuster V, Boerwinkle E, Melander O, Orho-​Melander M, Ridker PM,
and Kathiresan S. Genetic risk, adherence to a healthy lifestyle, and coronary disease. N Engl J Med
375: 2349–​2358, 2016.
37. Kishi T, Yoshimura R, Ikuta T, and Iwata N. Brain-​derived neurotrophic factor and major
depressive disorder: Evidence from meta-​analyses. Front Psychiatry 8: 308, 2017.
38. Kostrzewa E and Kas MJ. The use of mouse models to unravel genetic architecture of physical
activity: A review. Genes Brain Behav 13: 87–​103, 2014.
39. Labos C and Thanassoulis G. Genetic risk prediction for primary and secondary prevention of
atherosclerotic cardiovascular disease: An update. Curr Cardiol Rep 20: 36, 2018.
40. Leal G, Comprido D, and Duarte CB. BDNF-​induced local protein synthesis and synaptic plasti-
city. Neuropharmacology 76: 639–​656, 2014.
41. Li S, Zhao JH, Luan J, Ekelund U, Luben RN, Khaw KT, Wareham NJ, and Loos RJ. Physical
activity attenuates the genetic predisposition to obesity in 20,000 men and women from EPIC-​
Norfolk prospective population study. PLoS Med 7: e1000332, 2010.
42. Mancino S, Burokas A, Gutierrez-​Cuesta J, Gutierrez-​Martos M, Martin-​Garcia E, Pucci
M, Falconi A, D’Addario C, Maccarrone M, and Maldonado R. Epigenetic and prote-
omic expression changes promoted by eating addictive-​ like behavior. Neuropsychopharmacology
40: 2788–​2800, 2015.
43. Marigorta UM, Rodriguez JA, Gibson G, and Navarro A. Replicability and prediction: Lessons
and challenges from GWAS. Trends Genet 34: 504–​517, 2018.
44. Miller FL OCD, Herring MP, Sailors MH, Jackson AS, Dishman RK, Bray MS. Exercise
dose, exercise adherence, and associated health outcomes in the TIGER study. Med Sci Sports Exerc,
46: 69–​75, 2013.
45. Moore LV, Harris CD, Carlson SA, Kruger J, and Fulton JE. Trends in no leisure-​time physical
activity –​United States, 1988–​2010. Res Q Exerc Sport 83: 587–​591, 2012.
46. Oman RF and King AC. Predicting the adoption and maintenance of exercise participation using
self-​efficacy and previous exercise participation rates. Am J Health Promot 12: 154–​161, 1998.
47. Patterson R, McNamara E, Tainio M, de Sa TH, Smith AD, Sharp SJ, Edwards P, Woodcock
J, Brage S, and Wijndaele K. Sedentary behaviour and risk of all-​cause, cardiovascular and cancer
mortality, and incident type 2 diabetes: A systematic review and dose response meta-​analysis. Eur J
Epidemiol, 33: 811–​829, 2018.
48. Prochaska JO, Velicer WF, Rossi JS, Goldstein MG, Marcus BH, Rakowski W, Fiore C,
Harlow LL, Redding CA, Rosenbloom D, and et al. Stages of change and decisional balance for
12 problem behaviors. Health Psychol 13: 39–​46, 1994.
49. Rampersaud E, Mitchell BD, Pollin TI, Fu M, Shen H, O’Connell JR, Ducharme JL, Hines S,
Sack P, Naglieri R, Shuldiner AR, and Snitker S. Physical activity and the association of common
FTO gene variants with body mass index and obesity. Arch Intern Med 168: 1791–​1797, 2008.
50. Reiss D, Leve LD, and Neiderhiser JM. How genes and the social environment moderate each
other. Am J Public Health 103 Suppl 1: S111–​S121, 2013.
51. Rhodes JS, Garland T, Jr, and Gammie SC. Patterns of brain activity associated with variation in
voluntary wheel-​running behavior. Behav Neurosci 117: 1243–​1256, 2003.
52. Ross CE. Walking, exercising, and smoking: Does neighborhood matter? Soc Sci Med
51: 265–​274, 2000.
53. Rothman AJ. Toward a theory-​ based analysis of behavioral maintenance. Health Psychol
19: 64–​69, 2000.
54. Ryan R and Deci E. Self-​determination theory and the facilitation of intrinsic motivation, social
development, and well-​being. Am Psychologist 55: 68–​78, 2000.
55. Sailors MH, Jackson AS, McFarlin BK, Turpin I, Ellis KJ, Foreyt JP, Hoelscher DM, and
Bray MS. Exposing college students to exercise: the Training Interventions and Genetics of Exercise
Response (TIGER) study. J Am Coll Health 59: 13–​20, 2011.
56. Sajuthi SP, Sharma NK, Chou JW, Palmer ND, McWilliams DR, Beal J, Comeau ME, Ma
L, Calles-​Escandon J, Demons J, Rogers S, Cherry K, Menon L, Kouba E, Davis D, Burris
M, Byerly SJ, Ng MC, Maruthur NM, Patel SR, Bielak LF, Lange LA, Guo X, Sale MM,
Chan KH, Monda KL, Chen GK, Taylor K, Palmer C, Edwards TL, North KE, Haiman
CA, Bowden DW, Freedman BI, Langefeld CD, and Das SK. Mapping adipose and muscle

50
51

Systems genetics of exercise

tissue expression quantitative trait loci in African Americans to identify genes for type 2 diabetes and
obesity. Hum Genet 135: 869–​880, 2016.
57. Sarzynski MA, Loos RJ, Lucia A, Perusse L, Roth SM, Wolfarth B, Rankinen T, and
Bouchard C. Advances in exercise, fitness, and performance genomics in 2015. Med Sci Sports Exerc
48: 1906–​1916, 2016.
58. Scuteri A, Sanna S, Chen WM, Uda M, Albai G, Strait J, Najjar S, Nagaraja R, Orru M, Usala
G, Dei M, Lai S, Maschio A, Busonero F, Mulas A, Ehret GB, Fink AA, Weder AB, Cooper
RS, Galan P, Chakravarti A, Schlessinger D, Cao A, Lakatta E, and Abecasis GR. Genome-​
wide association scan shows genetic variants in the FTO gene are associated with obesity-​related
traits. PLoS Genet 3: e115, 2007.
59. Shen T, You Y, Joseph C, Mirzaei M, Klistorner A, Graham SL, and Gupta V. BDNF poly-
morphism: A review of its diagnostic and clinical relevance in neurodegenerative disorders. Aging Dis
9: 523–​536, 2018.
60. Szwergold BS. Fructosamine-​6-​phosphates are deglycated by phosphorylation to fructosamine-​
3,6-​bisphosphates catalyzed by fructosamine-​3-​kinase (FN3K) and/​or fructosamine-​3-​kinase-​
related-​protein (FN3KRP). Med Hypotheses 68: 37–​45, 2007.
61. Thompson PD, Tsongalis GJ, Ordovas JM, Seip RL, Bilbie C, Miles M, Zoeller R, Visich
P, Gordon P, Angelopoulos TJ, Pescatello L, and Moyna N. Angiotensin-​converting enzyme
genotype and adherence to aerobic exercise training. Prev Cardiol 9: 21–​24, 2006.
62. Thompson PD, Tsongalis GJ, Seip RL, Bilbie C, Miles M, Zoeller R, Visich P, Gordon
P, Angelopoulos TJ, Pescatello L, Bausserman L, and Moyna N. Apolipoprotein E geno-
type and changes in serum lipids and maximal oxygen uptake with exercise training. Metabolism
53: 193–​202, 2004.
63. Trost SG, Owen N, Bauman AE, Sallis JF, and Brown W. Correlates of adults’ participation in
physical activity: Review and update. Med Sci Sports Exerc 34: 1996–​2001, 2002.
64. Vimaleswaran KS, Li S, Zhao JH, Luan J, Bingham SA, Khaw KT, Ekelund U, Wareham NJ,
and Loos RJ. Physical activity attenuates the body mass index-​increasing influence of genetic vari-
ation in the FTO gene. Am J Clin Nutr 90: 425–​428, 2009.
65. Way BM and Taylor SE. A polymorphism in the serotonin transporter gene moderates cardiovas-
cular reactivity to psychosocial stress. Psychosom Med 73: 310–​317, 2011.
66. Wier LT and Jackson AS. Factors affecting compliance in the NASA/​Johnson Space Center fitness
programme. Sports Med 8: 9–​14, 1989.
67. Wilson K and Brookfield D. Effect of goal setting on motivation and adherence in a six-​week exer-
cise program. Int J Sport Exercise Physiol 7: 89–​100, 2009.
68. Wunderle M, Olmes G, Nabieva N, Haberle L, Jud SM, Hein A, Rauh C, Hack CC, Erber
R, Ekici AB, Hoyer J, Vasileiou G, Kraus C, Reis A, Hartmann A, Schulz-​Wendtland R,
Lux MP, Beckmann MW, and Fasching PA. Risk, prediction and prevention of hereditary
breast cancer –​large-​scale genomic studies in times of big and smart data. Geburtshilfe Frauenheilkd
78: 481–​492, 2018.
69. You JS, Dooley MS, Kim CR, Kim EJ, Xu W, Goodman CA, and Hornberger TA. A
DGKzeta-​FoxO-​ubiquitin proteolytic axis controls fiber size during skeletal muscle remodeling. Sci
Signal 11: eaao6847, 2018.
70. Zhang G and Nebert DW. Personalized medicine: Genetic risk prediction of drug response.
Pharmacol Ther 175: 75–​90, 2017.
71. Zhang Y, Liu J, Yao J, Ji G, Qian L, Wang J, Zhang G, Tian J, Nie Y, Zhang YE, Gold MS,
and Liu Y. Obesity: Pathophysiology and intervention. Nutrients 6: 5153–​5183, 2014.

51
52
53

SECTION 2

Systems genetics of
physical activity

Introduction
Over the last 25 years, it has been accepted that physical activity is a public health issue. In literally
hundreds of thousands of well-​conducted studies, physical activity or physical inactivity has been
linked with positive or negative health outcomes, respectively. However, as noted several times in
the coming chapters, global physical activity levels measured in almost every population are much
less than is appropriate for positive health benefits despite the tremendous educational efforts that
have been made worldwide to encourage physical activity. This disconnect –​between knowing
that we must be active and not being active –​is at the root of the six chapters in this section.
Specifically, these chapters comprehensively consider what is known about the systems genetics
of physical activity, ranging from the basic question of whether genetic systems regulate physical
activity to how the environment may affect the genetic systems regulating physical activity.
The section starts with a consideration of the basic question “Do genetics regulate phys-
ical activity level?” This critical question must be answered before further mechanistic studies
and application can be considered. Perhaps surprisingly, this question has probably generated
more studies than any other systems genetics question in the exercise and sport science discip-
lines. Chapter 5, written by Dr. Scott Kelly, tackles the question of whether genetics regulate
physical activity in animal models and provides an initial hypothesis that both central and per-
ipheral mechanisms are involved. This chapter is followed by one written by Matthijs van der
Zee and Dr. Eco De Geus, that carries the question into human models and results in some
elegant theoretical models showing that there are enough data to support a systems genetics
regulation of physical activity levels. With the foundation from Chapters 5 and 6, we are then
open to consider other important aspects of genetic regulation of physical activity. Chapter 7,
written by Dr. J. Timothy Lightfoot, Ayland Letsinger, and Jorge Granados, consider what
factors might have resulted in the evolution of genetic control of physical activity –​in other
words, why should genetics regulate physical activity? In Chapter 8, Dr. Justin Rhodes begins
the discussion of where the genetic mechanisms that control physical activity are located, with
an examination of the data that point to potential genetic mechanisms in the brain that regu-
late physical activity. Another approach to the “Where are the mechanisms located?” question
is presented in Chapter 9 by Dr. David Ferguson that focuses on potential peripheral gen-
etic regulating mechanisms of activity. While this book is focused on systems genetics, most

53
54

Systems genetics of physical activity

scientists understand that the environment can interact with and play a role in regulating genetic
mechanisms. Thus, Drs. Heather Vellers and Emily Schmitt in Chapter 10 address the possibility
that if physical activity is regulated by genetic mechanisms, perhaps these genetic mechanisms
can be altered by unique environmental exposures which lead to alterations in physical activity.
In total, these chapters provide a comprehensive review of the literature and thinking on
the systems genetics regulation of physical activity. We are pleased that these chapters have been
written by those investigators most involved and published in the topic, and as such, the chapters
provide the most current ideas and theories on the topic. As a result, we hope that these chapters
provide a foundation and starting point for increased investigation and identification of the crit-
ical controlling genetic factors of physical activity.

54
55

5
IS PHYSICAL ACTIVITY
REGULATED BY GENETICS?
EVIDENCE FROM
ANIMAL MODELS
Scott A. Kelly

What is physical activity in an animal model?

As in human populations, physical activity can represent a range of behaviors in animal models.
Consequently, context matters when discussing the genetic regulation of physical activity.
For example, assessing exercise endurance in rodents typically utilizes forced running on a
motorized treadmill equipped with a shock grid for a negative stimulus (48). Alternatively,
assessing daily movement distance may involve measuring the number of revolutions an animal
voluntarily turns a running wheel in a given amount of time (16). These two types (forced vs.
voluntary) of physical activity are a result of, and result in, at least some variation in physio-
logical and molecular responses. And, most relevant to this chapter, these exercise types may be
regulated by different genetic architectures or genomic mechanisms. I emphasize this early in
the text because results (e.g., genetic mapping, transcriptomic, and proteomic) across animal
models and between studies may appear inconsistent, and, in part, this lack of replication may
simply be a reflection of measuring different phenotypes, however subtle. This is especially
pronounced within the broad category of voluntary physical activity, where subtypes are
numerous (e.g., wheel running, home-​cage activity, open-​field activity, spontaneous physical
activity, and nonexercise activity thermogenesis). So, context certainly matters when comparing
results obtained from a variety of measures, but may not be as important if we are broadly
looking for genetic regulation of physical activity that transcends specific activity types (i.e., a
set of master activity regulators). Thus, in this chapter, I will attempt to provide examples in the
context of specific measurements.
I will not attempt to describe each of the voluntary activity types listed above given that
operational definitions and methodology of the measurements have been previously reviewed
(12, table 1 in 19). Given that this chapter will not attempt to be exhaustive, I will primarily
discuss the evidence for the genetic regulation of physical activity in the context of studies
utilizing voluntary wheel running. For more exhaustive reviews covering a range of activity
types in both rodents and humans, there are numerous studies available (12, 19, 22, 30, 31). This
chapter will focus on voluntary wheel running as an activity measure because of the import-
ance it has played in the early understanding of the genetic regulation of physical activity, and
its hypothesized parallels to human exercise behavior.

55
56

S.A. Kelly

Running wheels have long been used to measure voluntary activity in rodents (44), although
precisely why rodents engage in wheel-​running behavior and how applicable it is to human
voluntary exercise has been rightly debated. In an exhaustive review of wheel-​running behavior,
Sherwin (42) concluded that “wheel running has no directly analogous naturally occurring
behavior, it is (sometimes) performed for its own sake per se rather than as a redirected or
substitute activity, and studies on motivation show that wheel running is self-​reinforcing and
perceived by animals as ‘important’.” Sherwin (42) further argued that wheel-​running behavior
is most likely an “artefact of captive environments,” but regardless is “one of great interest to
behavioural science.” Counter to Sherwin’s contention that wheel-​running behavior is an arti-
fact of captive environments, Meijer and Robbers (34) demonstrated that when running wheels
are placed in nature, mice not only utilize them but also do so in bout lengths that match
captive mice. Eikelboom (7), in a response to Sherwin (42), argued: “Many similarities exist
between this intrinsically motivated human exercise and the wheel running seen in animals,
so analyzing them together may lead to theories explaining both.” Regardless, it can probably
be agreed upon that voluntary wheel-​running behavior requires ability and motivation, can be
self-​reinforcing, and as described by Novak et al. (37), “is not solely reflective of the tendency
to be physically active, but is a complex and dynamic behavior that interacts with genetics and
the environment.” For these reasons, I, and others have argued that voluntary wheel-​running
behavior in rodents adequately models human voluntary exercise behavior. However, a notable
discrepancy between human and rodent voluntary physical activity occurs between the sexes.
In human populations, males tend to engage in more physical activity, while among rodents,
females tend to be more active. The cause of this discrepancy has been discussed elsewhere (2;
and see Chapter 10). It is also important to note that when discussing the genetic architecture
of voluntary wheel-​running behavior I am referencing the predisposition to initiate physical
activity and not the response to exercise or trainability (e.g., 41), although it is certainly possible
the architectures may overlap and/​or interact.

Heritability
“Are the individual differences in spontaneous activity inherent?” (40). Although others had
suggested that behaviors might be regulated by “internal drives” (e.g., 39), Rundquist (40)
appears to be the first to have expressly designed an experiment to answer the question pertaining
to voluntary wheel-​running activity. Rundquist (40) used the qualifier “spontaneous” to distin-
guish between an internal drive as opposed to external motivators (presumably environmental
factors) of activity. In an attempt to answer the question, Rundquist (40) selectively bred rats
for 12 generations utilizing rotating drums, a design previously utilized by Stewart (44). The
rotating drums were solid surface running wheels with the home cage suspended inside the
drum (see ­Figure 1 in 44). Rundquist (40) commented on the “unbelievable” interindividual
variation observed through the course of selective breeding and noted that he handled “a male
rat which travelled approximately sixty feet in fifteen days, and at the other extreme, a female
which travelled over two hundred miles during the same period.” After 12 generations of selective
breeding, Rundquist (40) concluded that: “It is, then, quite safe to ascribe the major role in the
production of the individual differences in this activity to inheritance.” A logical follow-​up to
Rundquist (40), and at least partially the focus of the following studies, was to ask how much
of the total variance observed in voluntary physical activity is attributable to the genetic vari-
ance –​or, what is the exact “heritability” of voluntary wheel running?
Festing (10) quantified voluntary activity in 26 mouse strains, exposing them to wheels
(85 mm in diameter) over a 48-​hour (5 mice from each of 23 strains) and 7-​day (6 mice from

56
57

Physical activity animal models

each of 27 strains) period. In each experiment, mean daily activity was calculated (revolutions
per 24 hours in thousands). Despite high individual variation, Festing (10) observed statistically
significant interstrain differences in wheel activity in both experiments, with closely related
strains having similar activity levels. Lerman et al. (26) examined the genetic variability in vol-
untary exercise in seven inbred mouse strains. Running duration and distance was examined
over a 2-​week period using running wheels (11.5 cm diameter) placed inside individual cages.
Lerman et al. (26) observed statistically significant interstrain differences in wheel running dur-
ation, distance, and speed. Lerman et al. (26), using the coefficient of genetic determination (g2),
estimated broad-​sense heritability to be 0.42, 0.39, and 0.24 for running duration, distance, and
average speed, respectively. Lightfoot et al. (27), using 13 strains of mice, also estimated broad-​
sense heritability of wheel-​running activity. Lightfoot et al. (27) monitored wheel (127 mm)
running for 21 consecutive days. Heritability estimates, using the coefficient of genetic deter-
mination (g2), were 0.18 for average daily distance (km), 0.25 for average daily exercise time or
duration (minutes), and 0.14 for average daily exercise velocity (m/​minute). When examining
heritability among female mice only, estimates were 0.12 (distance), 0.12 (duration), and 0.44
(velocity). For males estimates were 0.31 (distance), 0.44 (duration), and 0.49 (velocity).
Coming full circle, back to the approach used by Rundquist (40), the most convincing case
that physical activity is a heritable trait in rodents (i.e., exhibits significant narrow-​sense herit-
ability or, significant additive variance) resulted from an artificial selection experiment for high
levels of voluntary wheel running in mice (45). The work was initiated by Theodore Garland Jr.
and colleagues, is currently ongoing, and has resulted in over 150 publications to date character-
izing the morphological, physiological, and behavioral traits that have evolved in concert with
high levels of voluntary activity. The replicated selection experiment began from a base popu-
lation of outbred Hsd:ICR mice (Mus domesticus). Utilizing 224 mice (112 male, 112 female)
eight closed lines (i.e., once established, individuals were only bred within a line) were generated.
Four replicate lines were selectively bred for high voluntary wheel running (high running lines,
HR) and four lines were maintained as controls (C lines). The C lines were exposed to the same
experimental protocol as the HR lines but were bred without regard to the amount of wheel
running. Individual mice, at 6–​8 weeks of age, were given access to running wheels (1.12 m
circumference, 35.7 cm diameter) for 6 days and the mean number of total revolutions on days
5 and 6 was the criterion for which the HR lines were selectively bred. After ten generations,
realized heritability estimates, adjusted for within-​family selection, ranged from 0.18–​0.32 across
the four HR lines, with an average of 0.28. By generation 16, selection had resulted in an
approximate 2.5-​to 3.0-​fold increase in total revolutions per day in HR lines (four replicates)
as compared to the control lines (C lines, four replicates). At generation 16, the difference in
number of revolutions was primarily caused by HR mice running faster rather than for more
minutes each day, but the relative importance of the two components differed between the sexes
with males showing a significant increase in amount of time spent running (45). Recent work
by Garland and colleagues has demonstrated variable selection signatures (“genomic regions
showing excessive differentiation between treatments”) in each of the four HR replicate lines
(49). These findings support the hypothesis that physical activity is a phenotype that may be
amenable to multiple independent evolutionary solutions (discussed further in Chapter 7).
Although estimates vary (Table 5.1), the studies just described have conclusively shown that
voluntary wheel running (and its components) in rodents is a heritable behavior. Additionally,
the heritability estimates in these studies match those of human populations (see Chapter 6 and
references therein).The remainder of this chapter, while reinforcing the heritability of voluntary
wheel-​running behavior, will focus on identifying specific chromosomal regions and potential
candidate genes responsible for physical activity.

57
58

S.A. Kelly

Table 5.1 Heritability estimates of voluntary wheel-​running traits in three different mouse models

Trait* Heritability estimate Reference

Running distance 0.39 Lerman et al. (27)

0.18 Lightfoot et al. (28)
0.28** Swallow et al. (46)
Running duration 0.42 Lerman et al. (27)
0.25 Lightfoot et al. (28)
0.14** Swallow et al. (46)
Running speed 0.24 Lerman et al. (27)
0.14 Lightfoot et al. (28)
0.28** Swallow et al. (46)

* Generalized categories. See text for additional methodological details (e.g., wheel size, number of
measurement days, etc.)
** Represent narrow-​sense heritability estimates. Running duration and average running speed are
estimated using a midparent offspring regression and are not adjusted for within-​family artificial
selection.

Genetic architecture
Genetic mapping
To date, numerous, quantitative trait loci (QTL) which are chromosomal regions correlating
with variation in a phenotype, have been mapped for a variety of physical activity traits util-
izing several different mouse models. Regardless of the exact physical activity trait or mouse
model, the ultimate goal of these studies is to identify specific genes and mutations within these
regions controlling the behavior of interest.These approaches have relied on second-​generation
intercrosses between inbred populations (F2), backcrosses (BC), advanced intercross lines (AIL),
and large panels of recombinant inbred lines (e.g., the Collaborative Cross (CC)). Below, I will
discuss studies that have successfully utilized each of these approaches to identify QTL under-
lying voluntary wheel running behavior in mice. As stated at the beginning of this chapter,
physical activity can be very broadly defined and the following examples are not meant to be
representative of the entirety of the behavior. And, indeed, there have been QTL identified for
other physical activity traits such as home-​cage activity (14), open-​field behavior (6), and tread-
mill running (4, 48). By focusing on a few representative studies, all targeting the same behavior,
direct comparisons, although difficult in isolated mapping populations, are made somewhat
easier when the phenotype is narrowly defined to voluntary wheel running.
Lightfoot et al. (28) documented the first QTL controlling voluntary wheel running in a
F2 population. The F2 population (n=310) was generated from previously characterized high-​
active (C57L/​J) and low-​active (C3H/​HeJ) inbred strains (27). F2 mice were given access to a
running wheel (solid-​surface, 145 mm diameter) for 21 consecutive days and average running
distance (km), duration (minutes), and speed (m/​minute) were calculated. As is frequently
observed in rodent models of exercise (2), F2 female mice ran greater distances, for a longer
period of time, and at greater speeds than males. Using 129 single nucleotide polymorphisms
(SNPs) evenly spaced across the genome, four statistically significant (5% experiment-​wise
threshold) QTL were identified underlying the variation in running distance (DIST13.1), dur-
ation (DUR13.1), and speed (SPD9.1, SPD13.1). Three of the QTL colocalized to the same

58
59

Physical activity animal models

region of chromosome 13 and each of these three accounted for approximately 6% of the total
phenotypic variation in a given trait with an additional running speed QTL (SPD9.1), located
on chromosome 9, accounting for 11.3% of the total phenotypic. All four of these QTL were
replicated with additional haplotype mapping. In addition to the four statistically significant
QTL, Lightfoot and colleagues identified 14 suggestive (5% chromosome-​wise threshold) QTL
underlying the voluntary exercise traits. None of the QTL identified significantly interacted
with sex. Despite a relatively small population size and low marker density, this original mapping
study was highly impactful. Lightfoot et al. (28) reinforced that voluntary locomotion was a
heritable trait and the additive genetic components matched narrow-​sense heritability estimates
from prior studies (for comparison see 45). Additionally, for the first time, specific chromo-
somal regions underlying the variation in voluntary exercise traits were identified. Although
the chromosomal regions identified were large, the results facilitated functional hypotheses (by
Lightfoot and others) about the role of specific genes underlying these QTL in the biological
regulation of activity.
Following on the work of Lightfoot and colleagues, Nehrenberg et al. (36) utilized a BC
population (n=384) initiated between mice selectively bred for high voluntary wheel running
(HR) and the inbred strain C57BL/​6J, an average runner among inbred strains (27). The HR
mice utilized to create the F1 population originated from one of the four replicated selection
lines and were fixed for a recessive mutation causing an approximately 50% reduction in hind-
limb muscle mass (the “mini-​muscle” phenotype (MM)) (20). F1 male mice were backcrossed
to HR female parents producing a population of 384 mice with a 50:50 ratio of normal to
MM phenotypes. BC mice were given access to a running wheel (nonsolid surface, circum-
ference=1.1 m) for 6 consecutive days. Mean values of distance (m), time (minutes), average
speed (m/​minute), and maximum speed in any one-​minute running interval (m/​minute), were
calculated for days 5 and 6 of the 6-​day test. These values were calculated and chosen for
mapping, as the mean number of total revolutions on days 5 and 6 was the criterion for which
the HR line was selectively bred (45). In the BC population, female mice ran more than
males and mice with the MM allele ran faster and further than mice without the mutation
(13). In the BC population, using 154 evenly spaced genome-​wide SNPs, statistically signifi-
cant QTL were detected for average running speed (chromosome 7) and maximum running
speed (chromosomes 6 and 7), while none were detected for running distance or duration.
Additionally, a sex-​specific QTL, for the male subpopulation, was detected for running time
on chromosome 2. Nehrenberg et al. (37) reinforced that QTL for voluntary wheel could be
detected and that, in at least one case, the QTL can be sex specific. The absence of replica-
tion of the findings of the Lightfoot study (28) was also a major contribution. As pointed out
by Nehrenberg et al. (36), the lack of replication reinforced the importance of differences in
measurement methods (e.g., different running wheels, different days of wheel access), general
laboratory-​to-​laboratory variation, age differences at the time of measurement, differences in
genetic background of mice utilized, and differences in the genetic cross design (F2 vs. BC).
Continuing to build on the work of the previous two studies, Kelly et al. (16) generated
an AIL between one HR line and the inbred strain C57BL/​6J (B6). Although Kelly et al.
(16) utilized a different replicate HR line, the AIL was an opportunity to attempt to repli-
cate the findings of previously identified QTL in a nearly similar genetic background used
by Nehrenberg et al. (36). The AIL also allowed for the theoretical reduction in confidence
intervals surrounding any mapped QTL, thus allowing greater inference of potential candi-
date genes underlying loci. AILs are generated through random intercrossing over multiple
generations to accumulate recombination events and provide increased mapping resolution
compared with F2 and BC populations (5).The progenitor lines utilized by Kelly and colleagues

59
60

S.A. Kelly

underwent a reciprocal breeding protocol to produce a F1 generation and the subpopulations

(HR♀ × B6♂ and B6♀ × HR♂) were subsequently kept separate (15). Following the F3 gen-
eration, a large G4 population (N=815) was utilized to measure a variety of phenotypes related
to voluntary exercise behavior (16) and body composition prior to and following exercise (17).
G4 individuals were given access to running wheels for 6 consecutive days and revolutions were
recorded in 1-​minute intervals for 23–​24 hours of each day. Mice were genotyped using 764
fully informative SNPs evenly spaced across the genome. As expected, the additional mating
through generations three and four resulted in an approximate threefold genetic map expansion
(averaged across all chromosomes) and increased mapping resolution. One of the two reciprocal
crosses (B6♀ × HR♂) was maintained through a G10 and subsequent phenotyping and mapping
(with 2058 fully informative SNPs) was performed in an attempt to replicate QTL identified
in the G4.
In the G4 population, 32 significant and 13 suggestive QTL were identified underlying vol-
untary wheel-​running behavior. These QTL collectively represented running distance (total
daily revolutions), time spent running (i.e., cumulative 1-​minute intervals in which at least
one revolution was recorded), average speed (total revolutions/​time spent running), and max-
imum speed (highest number of revolutions in any 1-​minute interval within a 24-​hour period).
Additionally, the QTL were representative of the slope (i.e., trajectory of running on a temporal
scale) and intercept (i.e., starting point of temporal trajectory) of a linear regression across all
6 days of running. The primary conclusions of Kelly and colleagues (16) were threefold. First,
reinforcing the findings of Lightfoot et al. (28) and Nehrenberg et al. (36), the genetic archi-
tecture underlying voluntary wheel-​running behavior is complex and most likely governed by
many genes, each having a relatively small effect. The largest QTL effect discovered in the AIL
accounted for only 6.6% of the total phenotypic variation. Second, QTL were often unique to
single days of running, a phenomena not examined by the Lightfoot (28) or Nehrenberg (36)
studies, but later replicated by Leamy and colleagues (24).This was especially pronounced when
comparing the initial days (1 and 2) of wheel exposure with the remaining days (3–​6). Third,
QTL associated with the trajectory of running across all 6 days (quantified by the slope of a
linear regression) largely did not colocalize with chromosomal regions identified on individual
days. Kelly and colleagues concluded that their results “reinforce a genetic basis for the predis-
position to engage in voluntary exercise,” but unique regions may govern “initiation, continu-
ation, and temporal pattern of voluntary activity in mammals” (16).
As a follow-​up to the G4 study, a similar study was repeated on one (B6♀ × HR♂) of the
two reciprocal crosses in the G10 population (25). In theory, replication of the G4 QTL would
be high and confidence intervals would be reduced even further, increasing the likelihood of
identifying a narrowed set of candidate genes. Leamy et al. (25) discovered that, on average, QTL
confidence intervals were reduced from 23 Mb in the G4 to 11.3 Mb in the G10 population.
However, there was only a 40% replicate rate of QTL in the G10 compared to the G4. Several
potential explanations for the lack of replication are discussed by Leamy et al. (25), the most
likely being that one-​half of the G10 mice were fed a high-​fat diet, whereas all G4 mice were
fed a standard diet. A small sample size and lack of statistical power prevented mapping with G10
individuals only fed a standard diet.
Despite the contribution of novel QTL discoveries in the studies presented above, limitations
persist. First, the above studies (and others not identified here) have failed to consistently rep-
licate individual loci. Perhaps this is not surprising given the subtle, but important interstudy
differences, and the infinitesimal regulation of complex traits such as physical activity (11). But,
it does make it difficult for future studies to target prioritized regions. Second, the studies typic-
ally identify QTL with large confidence intervals (e.g., 11 Mb or larger) containing potentially

60
61

Physical activity animal models

hundreds of candidates, again, making it difficult to prioritize functional genes and their physio-
logical effects. Third, given the relatively limited genetic background of the strains utilized,
the broad applicability of the findings may be limited. The final two QTL studies discussed
below have specifically attempted to minimize the size of the confidence intervals using large
populations and high marker densities.They also attempt to address the lack of genetic diversity
in the previous studies by utilizing populations representative of multiple inbred and recently
wild derived strains.
Mathes et al. (33) utilized CC mice to identify QTL for wheel-​running distance on days 5
and 6 and days 11 and 12 of a 12-​day exposure to voluntary running wheels (the same wheel
model utilized by Kelly et al. (16)). The CC are a large panel of recombinant inbred mouse
lines derived from a genetically diverse set of eight founder strains, thus providing high levels
of broadly representative phenotypic and genetic diversity (1). Within the CC mapping popu-
lation, wheel-​running variation exceeded that observed in the eight founder strains and ranged
from about 1 km per night up to about 16 km per night on days 5 and 6 of a 12-​day test. The
maximum values for the CC mice were comparable to those observed in mice selectively bred
for high voluntary wheel running (HR mice discussed above). Mathes et al. (33) identified only
one statistically significant QTL for mean running distance on days 11 and 12, but this QTL
accounted for 17% of the phenotypic variation. In a separate population, haplotype association
mapping across 38 inbred strains failed to replicate the findings of the CC (29), but did identify
three significant QTL for running distance among the entire population on chromosomes 12,
18, and 19. Additionally, five sex-​specific QTL were identified among males (chromosomes 5,
6, 8 (distance), and 6 (speed)), and four among females (chromosomes 8 and 11 (running dis-
tance), 11 (running speed), and X (running duration)). Given the genetic diversity present in
the two studies previously described (30, 33), there is undoubtedly a greater ability to generalize
the results to additional mouse models and translate the findings to human populations. Indeed,
Lightfoot et al. (30) discuss the encouraging overlap between their findings, other mouse studies
discussed above (16, 36), and the findings of human studies (e.g., 3, 43). Additionally, some QTL
identified by Lightfoot et al. (29) had narrow confidence intervals containing as few as eight
predicted genes, facilitating the identification of potential candidates regulating physical activity
behavior. Mathes et al. (33) were also able to identify candidate genes within QTL, but by using
a complementary expression (eQTL) approach.

Fine-​mapping approaches
As discussed above, the identification of chromosomal regions containing QTL has been
successful, but the identification of genes underlying these regions that regulate voluntary exer-
cise behavior has been more difficult. One approach to identify potential candidate genes has
been to combine QTL mapping with gene expression analysis, or the mapping of eQTL.
Limitations to this approach are discussed elsewhere (47), but are primarily associated with the
tendency to favor cis eQTL over trans eQTL and low microarray probe density. In short, expres-
sion level of individual transcripts is treated as the phenotype, and variation in expression levels
is correlated with chromosomal regions (as in QTL mapping). Genes that are mapped under
loci for previously identified QTL and are locally controlled (cis-​acting eQTL) are prioritized.
Using a G4 AIL population (discussed earlier), Kelly et al. (18, 21) examined the transcrip-
tome in both brain and muscle tissue. These two tissues were chosen as voluntary physical
activity has been shown to be composed of both motivation (neurobiological) and ability (mus-
cular). Both of these studies (18, 21) identified eQTL colocalizing with previously identified
wheel-​running QTL. In brain tissue, Kelly et al. (18) discussed six plausible candidate genes

61
62

S.A. Kelly

(Insig2, Socs2, DBY, Arrdc4, Prcp, IL15) and in muscle (21) three (Insig2, Prcp, Sparc) that poten-
tially may play a role in regulating voluntary activity. These candidates were chosen based on
the strength of the effect observed, their known biological functions, and their correlation with
distance run, time spent running, or speed of running. However, in each of these studies there
were many additional eQTL mapped under wheel-​running QTL that were not discussed. For
example, in brain tissue, on average, 30 eQTL were mapped under each of 43 previously iden-
tified wheel-​running QTL. Although this approach does narrow the list of candidate genes it
still leaves many more to sift through than is desired.
More commonplace than eQTL approaches is the examination of the global transcriptome
using mouse models exhibiting discordant phenotypes. Significant differential gene expression
(typically in physiologically relevant tissues) between groups with different phenotypes is then
used to prioritize potential candidate genes underlying the phenotype. If genes that are sig-
nificantly differentially expressed happen to fall under previously identified QTL then even
higher priority is given. For example, Mathes et al. (32) examined gene expression in the dorsal
striatum and nucleus accumbens in mice selectively bred for high voluntary wheel running or
for obesity. Mathes et al. (32) chose these target tissues and these strains of mice to determine
how dopamine may potentially contribute to opposing phenotypes. Mathes et al. (32) found
significant differences in gene expression between the strains in both brain areas within the
dopamine pathway, concluding that central reward pathways are important in both obesity and
excessive exercise.
Although both of the above approaches have been successful, neither predicts the pro-
tein products that result from the identified genes. Proteomic approaches, however, do iden-
tify proteins associated with a particular phenotype. Proteomic approaches have been utilized
to identify end-​protein expression, in both brain and muscle tissue, in high-​and low-​active
inbred strains of mice (8, 9). Ferguson et al. (8) examined protein expression in the soleus and
extensor digitorum longus muscles and found higher expression of proteins relating to calcium
regulation and the Krebs cycle pathways in high-​active mice. Low-​active mice overexpressed
proteins associated with cytoskeletal structure and electron transport pathways. Furthermore,
transient knockdown of proteins annexin A6 and calsequestrin 1 protein of high-​active mice
significantly reduced physical activity levels. Taken together Ferguson et al. (8) conclude, “each
pattern contributes to the peripheral capability to be either high-​or low-​active …” and “spe-
cific mechanisms regulate activity leading to the high-​or low activity status of the animal.” In
a follow-​up study on the nucleus accumbens, Ferguson et al. (9) identified seven differentially
expressed proteins associated with neural stress (stress 70 protein and V type proton ATPase cata-
lytic subunit A) and metabolism (creatine kinase B, succinyl-​CoA ligase), overexpressed in the
low-​and high-​active strains respectively. Ferguson et al. (9) also reinforce the systems approach
outlined in ­Figure 4 of Pomp et al. (38), suggesting that protein signatures should be a part of
efforts to identify mechanisms regulating physical activity.

Future directions
The studies described here and elsewhere have clearly demonstrated that voluntary physical
activity (wheel running) in rodent models is a heritable behavior. These studies have also iden-
tified chromosomal regions underlying the behavior. However, overall, there has been very little
progress identifying the specific genes and variants underlying identified QTL. The identifica-
tion of these genes and understanding how they function physiologically to regulate activity
levels must be a priority of future investigations. Once specific genes are identified, studies
can begin to examine whether or not these genes also directly impact disease states through

62
63

Physical activity animal models

pleiotropy. Exacerbating this challenge is that studies thus far have demonstrated that the genetic
architecture regulating physical activity is exceedingly complex (e.g., 23). The genetic architec-
ture of exercise behavior has been shown to be population specific (or mouse model-​dependent)
with very low replication of QTL across populations and even within populations. The “next-​
generation” mouse models (e.g., Collaborative Cross and Diversity Outbred populations) will
hopefully alleviate the need to create isolated mapping populations and be used universally in
studies attempting to uncover the genetic architecture of exercise behavior. Whether or not
interstudy replication of QTL increases and translation to human populations becomes easier
with these new mouse models remains unanswered.
Not only does the genetic architecture of physical activity vary by population, but it also
interacts with environmental factors such as sex, age, and diet, each of which may also dir-
ectly influence the predisposition to engage in voluntary activity. The overall complexity of the
genetic, environmental, and interactive effects on the predisposition to voluntarily exercise is
depicted in Figure 5.1 and elsewhere with accompanying references (19). As was stated at the
beginning of the chapter, physical activity is not a single trait, and the context in which we dis-
cuss the genetic architecture matters. Highlighting this fact is that at least two studies (16, 24)
have demonstrated that the genetic architecture underlying a specific type of activity (voluntary
wheel running) shifts over the course of an exercise regimen (or wheel exposure period).These
findings may inform clinicians as to why individuals vary in their willingness to initiate and
adhere to prescribed exercise programs. These findings also reinforce the point that the genetic
regulation of activity depends not just on the type of activity, but when the activity occurs rela-
tive to the rest of the activity. Further research is needed to establish the genetic factors contrib-
uting to the temporal stability of voluntary exercise levels.
In this chapter, I focused on the genetic architecture of voluntary wheel running. There
are many studies that have examined the genetic architecture of other physical activity types.
Although I have stressed narrowly defining the type of physical activity when comparing studies,
I also want to emphasize the importance of trying to discover a common genetic architecture
across all physical activity types. To my knowledge, there is little overlap in the genetic regions

Genetic architecture Environment

Motivation and ability

PHYSICAL ACTIVITY

Feedback

Physiological
response

Figure 5.1 Voluntary physical activity is simultaneously influenced by genetic architecture, the
environment, and gene-​by-​environment interactions. As a result of physical activity there is typically
a physiological response, the magnitude of which may itself be influenced by the genome. And, this
physiological response may “feedback” to increase or decrease subsequent physical activity. This feedback
(negative or positive) may act through the two components that comprise voluntary physical activity
behavior, motivation, and ability.

63
64

S.A. Kelly

identified across activity types (e.g., home-​cage activity, open-​field activity, wheel running, and
treadmill running), but I am also unaware of a study that has intentionally sought to map loci
across a broad range of behaviors in a single population. A related outstanding question is
whether the same genes are responsible for both elevated activity and inactivity and, if different,
if is there an imbalance in the number of genes that regulate each. The importance of pursuing
these unanswered questions cannot be overstated given the relative decline in human activity
(e.g., 46) and the potential human-​health impact of this decline (e.g., 35).

References
1. Aylor DL, Valdar W, Foulds-​Mathes W, Buus RJ, Verdugo RA, Baric RS, Ferris MT, Frelinger
JA, Heise M, Frieman MB, Gralinski LE, Bell TA, Didion JD, Hua K, Nehrenberg DL, Powell
CL, Steigerwalt J, Xie Y, Kelada SN, Collins FS, Yang IV, Schwartz DA, Branstetter LA,
Chesler EJ, Miller DR, Spence J, Liu EY, McMillan L, Sarkar A, Wang J, Wang W, Zhang Q,
Broman KW, Korstanje R, Durrant C, Mott R, Iraqi FA, Pomp D, Threadgill D, de Villena
FP, Churchill GA. Genetic analysis of complex traits in the emerging Collaborative Cross. Genome
Res 21: 1213–​1222, 2011.
2. Bowen RS, Turner MJ, Lightfoot JT. Sex hormone effects on physical activity levels: why doesn’t
Jane run as much as Dick? Sports Med 41: 73–​86, 2011.
3. Cai G, Cole SA, Butte N, Bacino C, Diego V, Tan K, Goring HH, O’Rahilly S, Farooqi IS,
Comuzzie AG. A quantitative trait locus on chromosome 18q for physical activity and dietary
intake in Hispanic children. Obesity 14: 1596–​1604, 2006.
4. Courtney SM, Massett MP. Identification of exercise QTL using association mapping in inbred
mice. Physiol Genomics 44: 948–​955, 2012.
5. Darvasi A, Soller M. Advanced intercross lines, an experimental population for fine genetic
mapping. Genetics 141: 1199–​1207, 1995.
6. Delprato A, Algeo MP, Bonheur B, Bubier JA, Lu L, Williams RW, Chesler EJ, Crusio WE.
QTL and systems genetics analysis of mouse grooming and behavioral responses to novelty in an
open field. Genes Brain Behav 16: 790–​799, 2017.
7. Eikelboom R. Human parallel to voluntary wheel running: exercise. Anim Behav 57: F11–​F12, 1999.
8. Ferguson DP, Dangott LJ, Schmitt EE, Vellers HL, Lightfoot JT. Differential skeletal muscle
proteome of high-​and low-​active mice. J Appl Physiol 116: 1057–​1067, 2014.
9. Ferguson DP, Dangott LJ, Vellers HL, Schmitt EE, Lightfoot JT. Differential protein expres-
sion in the nucleus accumbens of high-​and low-​active mice. Behav Brain Res 291: 283–​288, 2015.
10. Festing MFW. Wheel activity in 26 strains of mouse. Lab Anim 11: 257–​258, 1977.
11. Fisher RA. The Genetical Theory of Natural Selection. Oxford: Oxford University Press, 1930.
12. Garland T Jr, Schutz H, Chappell MA, Keeney BK, Meek TH, Copes LE, Acosta W, Drenowatz
C, Maciel RC, van Dijk G, Kotz CM, Eisenmann JC. The biological control of voluntary exer-
cise, spontaneous physical activity, and daily energy expenditure in relation to obesity: human and
rodent perspectives. J Exp Biol 214: 206–​229, 2011.
13. Hannon RM, Kelly SA, Middleton KM, Kolb EM, Pomp D, Garland T Jr. Phenotypic effects
of the mini-​muscle allele in a large HR × C57BL/​6J mouse backcross. J Hered 99: 349–​354, 2008.
14. Kas MJH, de Mooij-​van Malsen JG, de Krom M, van Gassen KLI, van Lith HA, Olivier B,
Oppelaar H, Hendriks J, de Wit M, Groot Koerkamp MJA, Holstege FCP, van Oost BA, de
Graan PNE. High-​resolution mapping of a novel genetic locus regulating voluntary physical activity
in mice. Genes Brain Behav 11: 113–​24, 2012.
15. Kelly SA, Nehrenberg DL, Hua K, Gordon RR, Garland T Jr, Pomp D. Parent-​of-​origin effects
on voluntary exercise levels and body composition in mice. Physiol Genomics 40: 111–​120, 2010.
16. Kelly SA, Nehrenberg DL, Peirce JL, Hua K, Steffy BM, Wiltshire T, Pardo-​Manuel de
Villena F, Garland T Jr, Pomp D. Genetic architecture of voluntary exercise in an advanced inter-
cross line of mice. Physiol Genomics 42: 190–​200, 2010.
17. Kelly SA, Nehrenberg DL, Hua K, Garland T Jr, Pomp D. Exercise, weight loss, and changes
in body composition in mice: phenotypic relationships and genetic architecture Physiol Genomics
43: 199–​212, 2011.

64
65

Physical activity animal models

18. Kelly SA, Nehrenberg DL, Hua K, Garland T Jr, Pomp D. Functional genomic architecture of pre-
disposition to voluntary exercise in mice: Expression QTL in the brain. Genetics 191: 643–​654, 2012.
19. Kelly SA, Pomp D. Genetic determinants of voluntary exercise. Trends Genet 29: 348–​357, 2013.
20. Kelly SA, Bell TA, Selitsky SR, Buus RJ, Hua K, Weinstock GM, Garland T Jr, de Villena
FPM, Pomp D. A novel intronic single nucleotide polymorphism in the myosin heavy polypeptide
4 gene is responsible for the mini-​muscle phenotype characterized by major reduction in hind-​limb
muscle mass in mice. Genetics 195: 1385–​1395, 2013.
21. Kelly SA, Nehrenberg DL, Hua K, Garland T Jr, Pomp D. Quantitative genomics of voluntary
exercise in mice: Transcriptional analysis and mapping of expression QTL in muscle. Physiol Genomics
46: 593–​601, 2014.
22. Kostrzewa E, Kas MJ. The use of mouse models to unravel genetic architecture of physical activity.
Genes Brain Behav 13: 87–​103, 2014.
23. Leamy LJ, Pomp D, Lightfoot JT. An epistatic genetic basis for physical activity traits in mice. J
Hered 99: 639–​646, 2008.
24. Leamy LJ, Pomp D, Lightfoot JT. A search for quantitative trait loci controlling within-​individual
variation of physical activity traits in mice. BMC Genet 11: 83, 2010.
25. Leamy LJ, Kelly SA, Hua K, Pomp D. Exercise and diet affect quantitative trait loci for body
weight and composition traits in an advanced intercross population of mice. Physiol Genomics
44: 1141–​1153, 2012.
26. Lerman I, Harrison BC, Freeman K, Hewett TE, Allen DL, Robbins J, Leinwand LA. Genetic
variability in forced and voluntary endurance exercise performance in seven inbred mouse strains. J
Appl Physiol 92: 2245–​2255, 2002.
27. Lightfoot JT, Turner MJ, Daves M, Vordermark A, Kleeberger SR. Genetic influence on daily
wheel running activity level. Physiol Genomics 19: 270–​276, 2004.
28. Lightfoot JT, Turner MJ, Pomp D, Kleeberger SR, Leamy LJ. Quantitative trait loci (QTL) for
physical activity traits in mice. Physiol Genomics 32: 401–​408, 2008.
29. Lightfoot JT, Leamy L, Pomp D, Turner MJ, Fodor AA, Knab A, Bowen RS, Ferguson
D, Moore-​Harrison T, Hamilton A. Strain screen and haplotype association mapping of wheel
running in inbred mouse strains. J Appl Physiol 109: 623–​634, 2010.
30. Lightfoot JT. Current understanding of the genetic basis for physical activity.J Nutr 141: 526–​530, 2011.
31. Lightfoot JT, De Geus EJC, Booth FW, Bray MS, den Hoed M, Kaprio J, Kelly SA, Pomp
D, Saul MC, Thomis MA, Garland T Jr, Bouchard C. Biological/​genetic regulation of physical
activity level: consensus from GenBioPAC. Med Sci Sports Exerc 50: 863–​873, 2018.
32. Mathes WF, Nehrenberg DL, Gordon R, Hua K, Garland T Jr, Pomp D. Dopaminergic
dysregulation in mice selectively bred for excessive exercise or obesity. Behav Brain Res
210: 155–​163, 2010.
33. Mathes WF, Aylor DL, Miller DR, Churchill GA, Chesler EJ, de Villena FP, THreadgill DW,
Pomp D. Architecture of energy balance traits in emerging lines of the Collaborative Cross. Am J
Physiol Endocrinol Metab 6: E1124–​1134, 2011.
34. Meijer JH, Robbers Y. Wheel running in the wild. Proc Biol Sci 281: 20140210, 2014.
35. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in the United States,
2000. JAMA 291: 1238–​1245, 2004.
36. Nehrenberg DL, Wang S, Hannon RM, Garland T Jr, Pomp D. QTL underlying voluntary
exercise in mice: interactions with the “mini-​muscle” locus and sex. J Hered 101: 42–​53, 2010.
37. Novak CM, Burghardt PR, Levine JA. The use of a running wheel to measure activity in
rodents: Relationship to energy balance, general activity, and reward. Neurosci Biobehav Rev
36: 1001–​1014, 2012.
38. Pomp D, Nehrenberg D, Estrada-​Smith D. Complex genetics of obesity in mouse models. Annu
Rev Nutr 28: 331–​345, 2008.
39. Richter CP. Animal behavior and internal drives. Q Rev Biol 2:307–​343, 1927.
40. Rundquist EA. Inheritance of spontaneous activity in rats. J Comp Psych 16: 415–​438, 1933.
41. Sarzynski MA, Ghosh S, Bouchard C. Genomic and transcriptomic predictors of response levels
to endurance exercise training. J Physiol 595: 2931–​2939, 2017.
42. Sherwin CM.Voluntary wheel running: a review and novel interpretation. Anim Behav 56: 11–​27, 1998.
43. Simonen R, Rankinen T, Perusse L, Rice T, Rao DC, Chagnon Y, Bouchard C. Genome-​
wide linkage scan for physical activity levels in the Quebec Family study. Med Sci Sports Exerc
35: 1355–​1359, 2003.

65
66

S.A. Kelly

44. Stewart CC. Variations in daily activity produced by alcohol and by changes in barometric pressure
and diet, with a description of recording methods. Am J Physiol 1: 40–​56, 1898.
45. Swallow JG, Carter PA, Garland T Jr. Artificial selection for increased wheel-​running behavior in
house mice. Behav Genet 28: 227–​237, 1998.
46. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, Mcdowell M. Physical activity in the
United States measured by accelerometer. Med Sci Sports Exerc 40: 181–​188, 2008.
47. Verdugo RA, Farber CR, Warden CH, Medrano JF. Serious limitations of the QTL microarray
approach for QTL gene discovery. BMC Biol 8: 96, 2010.
48. Ways JA, Smith BM, Barbato JC, Ramdath RS, Pettee KM, DeRaedt SJ, Allison DC, Koch
LG, Lee SJ, Cicila GT. Congenic strains confirm aerobic running capacity quantitative trait loci on
rat chromosome 16 and identify possible intermediate phenotypes. Physiol Genomics 29: 91–​97, 2007.
49. Xu S, Garland T Jr. A mixed model approach to genome-​wide association studies for selection
signatures, with application to mice bred for voluntary exercise behavior. Genetics 207: 785–​799, 2017.

66
67

6
IS PHYSICAL ACTIVITY
REGULATED BY GENETICS?
EVIDENCE FROM STUDIES
IN HUMANS
Matthijs D. van der Zee and Eco de Geus

Introduction
Physical activity (PA) is a broad concept containing a variety of different human behaviors that
share the common denominator of expected beneficial effects on mental and somatic health
(1). Typically, daily PAs take up about 30% of total energy expenditure and reflect a mixture of
obligatory activities related to transportation, work, or household chores and self-​chosen PA
behaviors including sports and exercise activities in leisure time. Objective measurements of
daily PA in studies conducted in Europe and the US have confirmed the existence of large indi-
vidual differences in activity levels (26, 27). These differences are observed in men and women
alike, and persist after stratifying for age which is known to have a strong effect on average PA
level (49).
To increase the success of intervention on this important health behavior, much research has
been devoted to the causes of the individual differences seen in PA in the general population.
In this chapter, we focus on the role played by genetic factors. Broadly, the genetic research on
PA phenotypes can be classified into a) twin studies that partition the observed variance in PA
phenotypes into environmental and genetic components based on well-​established biometric
models of human inheritance (48), and b) gene-​finding studies that use either candidate gene-​
based approaches based on the known biological role of proteins (e.g., in neurotransmission) or
agnostic whole-​genome searches using genome-​wide microsatellite (linkage) or single nucleo-
tide polymorphism (SNP) markers to detect genomic loci harboring variants associated with
PA. We will summarize the current evidence from each of these types of studies to address the
question of whether PA is regulated by genetics.

Twin studies
PA behaviors appear to “run in the family,” for example, the chance of one family member
being a regular exerciser increases the chance of all other family members to be, or to become,
an exerciser. Familial aggregation of PA can be investigated by computing correlations among
relatives such as siblings, and parents and their offspring. However, siblings among each other,
and parents and their offspring, share not just half of their genes, they also share a house-
hold, socioeconomic status, the neighborhood, and various other aspects of belonging to the

67
68

M.D. van der Zee and E. de Geus

same family or living in the same neighborhood (the so-​called shared environment), including
parenting behaviors, family functioning, shared peers, etc.
Twin studies can separate the two mechanisms of familial aggregation by comparing the
resemblance in monozygotic (MZ) or identical twins to the resemblance in dizygotic (DZ) or
fraternal twins (48). When twins are reared together, the amount of this sharing of the family
environment is the same for MZ and DZ twins. The important difference between MZ and
DZ twins is that the former share most, if not all, of their genotypes, whereas the latter share on
average only half of the genotypes segregating in that family. If the resemblance in PA within
MZ pairs is larger than that in DZ pairs, this suggests that additive genetic factors (A) influence
PA. If MZ resemblance is more than double as large, it suggests the influence of nonadditive
(D) genetic factors. Additive genetic factors represent the sum of all linear effects of the gen-
etic loci that influence the trait of interest. Nonadditive factors include dominance and epi-
static interaction effects. If, however, the resemblance in PA in DZ twins is as large as it is in
MZ twins, this points to shared environmental factors (C) as the cause of twin resemblance.
Furthermore, the extent to which MZ twins do not resemble each other is ascribed to the
unique environmental factors (E). These include all person-​specific experiences such as differ-
ential jobs or lifestyles, accidents or other life events, and in childhood, differential treatment by
the parents, going to different schools, and having nonshared friends and peers. Measurement
error will also be subsumed by the unique environmental factor. For more details on the mod-
eling of twin and family resemblance, see Chapter 3.
We have previously reviewed all twin and family studies on the heritability of total PA
and regular sports and exercise behaviors in childhood or adolescent samples (50). In younger
children, the shared environmental factors (C) explain the largest part of the variation in PA.
However, the importance of these shared environmental factors decreases in adolescence and
young adulthood, where genetic effects become the dominant factor explaining individual
differences in both total PA and regular sports and exercise behaviors. A sample size-​weighted
meta-​analysis showed PA heritability estimates of 20% (95% confidence interval (CI) 13–​27%)
in children, 35% (95% CI 17–​52%) in early adolescents, and 53% (95% CI 47–​59%) in late
adolescents. This increase in heritability ran in parallel to a gradual decrease in the import-
ance of shared environmental factors, which are the main cause of individual differences in PA
and exercise behavior in young childhood, but then gradually become overwhelmed by the
importance of genetic factors during adolescence. Using longitudinal follow-​up of twins from
childhood to adolescence, we showed that the increase in heritability, at least for voluntary exer-
cise behaviors, was due to a strong increase in the genetic variance in the course of adolescence
that was not paired to a similar increase in environmental variance (31). Whether the trend of
increasing heritability is continued or curbed in young, middle, or late adulthood remains to
be established.
Sufficient studies in adult twins (>18 years of age) have now accrued to allow a similar meta-​
analytic approach and to arrive at solid estimates of additive genetic and shared environmental
variance for PA phenotypes in adulthood. We searched publications in the English language on
human subjects in PubMed and Web-​of-​Science from January 1980 to December 2017 using
the keywords (“Physical Activity” OR Exercise OR Sports OR Lifestyle) AND (Genetic OR
Genes OR Linkage OR QTL OR Twin OR Family OR Familial OR Heritability) AND
“Humans” [MeSH terms]. From the 850 putative papers, title and abstract analysis was used to
select only those publications reporting MZ and DZ/​sib correlations (in at least 30 complete
pairs) and/​or estimates of (non-​)additive genetic, shared, and/​or unique environmental vari-
ance components. Reference sections of selected papers were used to identify additional papers
missed by these search terms. We then removed samples reporting on twins with a mean age

68
69

Physical activity studies in humans

less than 18. Next we removed partly overlapping datasets from the same twin cohorts. When
the same phenotype was used in largely the same age group, we only used the study reporting
the largest dataset, which typically would be the most recent study. A final set of 27 adult twin
studies (2, 3, 8, 11, 14–​16, 20–​24, 29–​32, 34, 35, 40, 44, 45, 46, 47, 51, 53, 54, 58) were selected
by the above search criteria.

Physical activity phenotypes

In the final set of 27 adult twin studies, we encountered a large variation in measure-
ment instruments and PA measures used. By far the largest common denominator was the
use of survey-​based methods using subjective PA reporting on self or family members. In
large population-​based twin registries, surveys are often considered the only feasible strategy.
However, subjective reporting of PA is vulnerable to distortion due to recall error and reporting
biases (25). It is particularly difficult to estimate both the duration and frequency of PAs that
are light to moderate in intensity, including common activities such as walking and standing, or
household activities (5). For total PA and light to moderate PA, objective measurement strat-
egies using indirect calorimetry, the doubly labeled water method, or movement sensors (60)
are therefore superior.
When the focus is not on the detection of total PA, but when people are asked to report
moderate to vigorous PA, specifically when confined to structured activities in leisure time,
self-​reporting seems to fare much better (10). The cognitive salience of such intensive PAs is
higher than light to moderate intense activities occurring as part of daily routine. Reliability of
self-​reporting further increases when a restriction is made to the reporting of regular exercise
and sports behaviors in leisure time. In our own research, for instance, we have shown a high
short-​term test–​retest reliability (17, 55) as well as substantial tracking over longer periods of
time for the weekly volume of voluntary exercise behavior (31).
Our meta-​analysis was organized into the main domains of PA that we encountered in
the literature: total physical activity (TPA), moderate to vigorous physical activity (MVPA,
including separate measures for moderate or vigorous PA where applicable), leisure-​time phys-
ical activity (LTPA), and voluntary exercise behavior (VEB). The first two categories (TPA and
MVPA) cover activities that are only partly under the individual’s control, whereas the latter
two (LTPA and VEB) cover PA that is largely voluntary in nature. We excluded PA measures
that deliberately excluded sports and exercise activities (e.g., Baecke’s nonexercise-​related PA
index), measures exclusively reporting on PA in the occupational setting (e.g., Baecke’s Work/​
School index), and measures of daily activities of low intensity (e.g., accelerometer derived time
spent in low LPA).
To convert the measured PA behaviors into an actual summary metric for use in the genetic
analyses, again a breadth of different strategies has been used. Most studies adopt an estima-
tion of the total energy expenditure by PA as their focus. Estimation of energy expenditure
is usually based on published compendia (4) that convert each activity into equivalents of the
resting energy expenditure (1 MET, approximately 1 kcal/​kg/​hour). For VEB, most studies use
a strategy akin to what we do in the Netherlands Twin Register, where we use an open format
that allows reporting all sports and exercise activities that are performed for at least 3 months
a year and then record duration, frequency, and intensity of each reported activity. Typically,
activities are censored that do not reach a minimal threshold of intensity such as fishing or
chess. Exercise related to swimming, sailing, or skiing that are restricted to the annual holidays
is discarded. Physical education classes are also discarded as they are poorly standardized and the
activities are not voluntary. Energy expenditure in all leisure-​time VEB is then summed across

69
70

M.D. van der Zee and E. de Geus

all valid exercise activities in a weekly MET-​hour score by taking the sum of the products of
their weekly frequency, average duration, and MET score.
Another major difference across studies is the type of scale used to quantify PA. Continuous
interval scales seem most optimal in terms of statistical power but quite often either ordinal
categories (e.g., tertiles of MET-​hours weekly; frequency of LTPA as once per month, once per
week, every day) or even nominal dichotomies (YES/​NO regular exerciser;YES/​NO adheres
to PA guidelines) were used. This use of dichotomies and categories is mostly inspired by the
frequency distribution of some of the LTPA and VEB phenotypes. For instance, moderate to
vigorous PA in leisure time is very skewed. Likewise, many individuals do not engage in vol-
untary exercise behavior (zero score) and even in those who do the amount of exercise can
still show a skewed distribution. Often no transformation is available that converts such mixed
distributions to a normal distribution. For these phenotypes, therefore, the use of ordinal cat-
egories or a dichotomy can be meaningful. A liability threshold model can then be used in the
genetic analysis to recapture the normal distribution of the latent “liability to be a (vigorous)
exerciser.”

Meta-​analyses
To assess heritability and influence of common environmental factors in the four PA categories
mentioned previously, meta-​analysis on A and C estimates was conducted across twin studies.
Within the 27-​study database, the A and C estimates and their standard errors were often
reported separately for males and females. However, if the A and C estimates were reported
only for males and females combined, we assigned the same estimates to the male and female
part of the sample, adjusting the N to reflect the number of male or female twins. Most studies
used age and sex as covariates, but additional covariates were sometimes used too, including
socioeconomic status, body mass index, or fitness. When multiple estimates were generated for
different covariate compositions, we opted to use the estimates only correcting for age and sex
effects on mean PA level. In these studies, structural equation-​based variance decomposition
modeling was by far the most used analytic strategy. Unless stated otherwise, the results from
the most parsimonious structural equation model were used in the meta-​analysis. This model
was most often an AE or ACE model. If the AE model was used, C was set to zero. No studies
reported a model with nonadditivity (D).
For each of the four PA phenotypes, the estimates for A and C were computed in a sample-​
size weighted meta-​analysis across all available studies for males and females separately. Inverse
variance weighting was not possible because not all studies provided either standard errors of
the estimates or 95% CIs from which the standard errors could be approximated. Forest plots
in Figures 6.1 and 6.2 present the characteristics and the A and C estimates per study, and the
meta-​analytic results per PA phenotype. Study characteristics are as follows: country, whether
males and females were combined in A and C estimation, whether PA measurement was
through surveys (SUBJ) or experimental (OBJ), type of scale used (DIchotomy,CATegorical, or
CONtinuous), mean age, and sex-​specific sample size. The estimates for A and C are listed in
the forest plots as a function of mean sample age.

Results for total physical activity

For the meta-​analysis of TPA (see Figures 6.1 and 6.2) a total of six studies using eight
phenotypes was available, most using an objective measure to quantify PA, and all combining
males and females. The meta-​analytic heritability estimate was 48% for females and 51% for

70
71

Figure 6.1 Forest plot containing study-​specific and meta-​analytic estimates of heritability (A) of total
physical activity (TPA), moderate to vigorous activity (MVPA), leisure-​time physical activity (LTPA), and
voluntary exercise behavior (VEB), for males and females. Estimates include results from subjective (Sub),
and objective (Obj), continuous (CON), categorical (CAT), and dichotomous (DI) variables. *=studies
where an AE model was fitted. C=estimates from a combined male and female sample.
72

Figure 6.2 Forest plot containing study-​specific and meta-​analytic estimates of shared environmental
contribution (C) to the variance in total physical activity (TPA), moderate to vigorous activity (MVPA),
leisure-​time physical activity (LTPA), and voluntary exercise behavior (VEB), for males and females.
Estimates include results from subjective (Sub), and objective (Obj), continuous (CON), categorical
(CAT), and dichotomous (DI) variables. *=studies where an AE model was fitted. C=estimates from a
combined male and female sample.
73

Physical activity studies in humans

males. Largest heritability estimates were found by Joosen et al. (34) who used the most reliable
methods, namely the doubly labeled water method and more than 14 days of accelerometer
recording, albeit in the smallest sample.The two survey methods using the Baecke questionnaire
yielded results comparable to the objective methods. No evidence for common environmental
effects was found.

Results for moderate to vigorous physical activity

For the meta-​analysis of MVPA (see Figures 6.1 and 6.2) a total of five studies using seven
phenotypes was available, all combining males and females. The meta-​analytic heritability esti-
mate was 44% for females and 41% for males. Comparable heritability was found for objective
and subjective measures. Somewhat deviant results were found in the 26-​year-​old Add Health
study participants (47). Using a 7-​day recall questionnaire to obtain a dichotomy of MVPA
frequency that indicated whether participants met national PA recommendations (>5 bouts/​
week), lower heritability was paired to the only evidence for a shared household effect on
MVPA (50%).

Results for leisure-​time physical activity

We used eight twin studies for the meta-​analysis of LTPA (see Figures 6.1 and 6.2) reporting on
more than 200,000 twins from seven countries.The meta-​analytic heritability estimate was 49%
for males and 45% for females with narrow confidence intervals (95% CI upper − lower=~10%)
reflecting little heterogeneity. LTPA was only available from self-​report measures in these very
large epidemiological samples. In two countries, Norway and Australia, evidence for common
environmental impact on LTPA was found (~28%). Comparing sample sizes, the absence of a
large C in other twin studies cannot be attributed to low power.

Results for voluntary exercise behavior

VEB was also measured exclusively by survey or interview measures. A total of 15 different
studies using 14 unique samples/​measures yielded a meta-​analytic heritability estimate of 48%
for males and 51% for females (see Figure 6.1) with mild heterogeneity (95% CI upper −
lower=~20%). Some evidence for common environmental effects were found, mostly driven
by the younger samples (~10%, see Figure 6.2) with a particularly strong C effect in the young
adult Finnish twins (43% males, 49% females).

Gene-​finding studies
The above meta-​analyses show that genetic variants play a key role in adult PA, whether it is
total daily PA as measured with objective means (e.g., accelerometers), or self-​reported voluntary
sports and exercise activities in leisure time. We now turn to the gene-​finding studies aiming to
detect the actual “PA genes”. For this, we have to use a narrative approach because not enough
replication is currently available for any single locus to consider meta-​analysis. The majority of
gene-​finding studies on TPA, MVPA, LTPA, and VEB have used a candidate gene approach.The
use of this approach was predicated on the expectation that individual genetic variants would
be associated with complex behavioral traits at a magnitude that would be small but detect-
able with hundreds, or a few thousand individuals. The advent of large international consortia
performing meta-​analyses on genome-​wide association results for many complex behavioral

73
74

M.D. van der Zee and E. de Geus

traits in samples as large as hundreds of thousands of participants has found this expectation to
be untenable. The risk contributed by any single variant is tiny rather than just small, with only
an increase of ~0.05 standard deviation per risk allele at best, necessitating much larger sample
sizes to be able to detect them (59). Based on these types of concerns regarding candidate gene
studies using small samples, we focus here mainly on the data-​driven genome-​wide approach in
linkage and genome-​wide association studies (GWAS).

Linkage
The three genome-​wide linkage studies (7, 19, 52) performed to date have only produced sug-
gestive hits, in keeping with their modest sample sizes (767 <N <1120). In the Quebec family
study, Simonen et al. (52) reported linkage for TPA (13q22–​q31), MVPA (4q28.2, 7p11.2, 9q31.1,
13q22–​q31), and VEB (11p15 and 15q13.3). Interestingly, the latter 15q linkage region contains
the GABRG3 gene in which a SNP (rs8036270) was found to be significantly associated with
VEB in GWAS on a combined sample of 2622 Dutch and European American middle-​aged
adults (18). Significant association with GABRG3 was replicated for LTPA in older adults in a
GWAS in 10,684 European and 11,093 African Americans (42) although with different SNPs
(rs72707657, rs12438610, rs12902711, rs12595253). The GABRG3 gene may be involved in
the aversive effects of exercise-​induced fatigue because high expression levels of the GABRG3
gene was found after a bout of exhaustive exercise (36). However, a GWAS in 13,980 Japanese
older adults did not find significant association with any SNPs in the GABGR3 gene region
and the gene also did not surface in the other two GWAS.
In the Viva La Familia study (7), physical inactivity, recorded as the percentage time in sed-
entary activity, significantly mapped to markers D18S1102–​D18S64 in the chromosome 18q21
region where the MC4R gene resides. An additional suggestive linkage signal for TPA was
detected in the same region. MCR4 had been implicated in smaller-​scaled candidate gene
studies (9, 43) but it failed to replicate in any of the five GWAS, two of which explicitly tested
it as a candidate gene (i.e., at lower P-​value). In the Viva La Familia study, a further linkage
signal was found for MVPA in the region at 9q31.1, which harbors the RN7SK gene. A SNP
(rs7023003) near this gene produced a suggestive (P <10−5) association with TPA in 8454
Korean participants (38). However, the same RN7SK SNP was explicitly but unsuccessfully
tested for replication in Japanese participants (28) and the gene also did not surface in the other
two GWAS.
Finally, in a sample from the Netherlands Twin Register (19), a suggestive linkage with
exercise participation was found in all subjects on chromosome 19p13.3 (LOD 2.18). A SNP
(rs12462609) in this region, located in the CACNA1A gene, was associated with vigorous PA
in 261,055 older adult participants in the UK Biobank (39), and with TPA in 8454 older adult
Koreans (38) whereas another SNP (rs111901094) in GATAD2A in the 19p13 linkage region
was associated with VEB in the UK Biobank participants (39). The 19p13 region did not gen-
erate a significant or suggestive signal in the other two GWAS (28, 42). In summary, genes
suggested from linkage regions (GABRG3, MCR4, RN7SK, CACNA1A) have not held up
systematically when tested as candidate genes in GWAS.

Genome-​wide association
Five GWAS (18, 28, 38, 39, 42) conducted for exercise behavior or related PA phenotypes were
available at the time of writing. The three smallest GWAS (2632 <N <11,093) did not detect
significant associations after the required stringent correction for the multiple testing burden

74
75

Physical activity studies in humans

that is inherent in the agnostic genome-​wide approach (18, 38, 42). However, some genes
listed as receiving suggestive evidence based on more lenient thresholds (P <10−5) showed
interest-​generating patterns of replication across multiple studies. For example, a suggestive
association of VEB was found in Dutch and American participants of European descent (18) for
an eQTL (rs12612420) that influences the expression of the DNAPTP6 gene. This association
was replicated (P=0.0092) in 16,026 Japanese participants (28).
Suggestive evidence for an association between VEB and the PAPSS2 gene in Dutch and
American participants (18) was replicated in 11,903 African American participants (42) for
LTPA, although for different SNPs (rs10887741 vs. rs1819162). The PAPSS2 gene has been
linked to maximal exercise capacity (49) which may be a factor influencing the motivation to
engage in voluntary exercise (12). Explicit testing of an association between VEB and PAPSS2 in
13,980 Japanese participants, however, did not produce a significant replication (28). Suggestive
evidence for an association with LTPA was found for rs116550874 at 1p36.23, and rs3792874,
rs3792877, rs3792878, rs79173796 at 5q31.1 (P <8.61 × 10–​7) in 11093 African Americans and
for rs28524846 at 14q23.3 (P <1.30 × 10–​6) in 10,684 European Americans (42). The authors
point to the ENO1, SLC22A5, and PDLIM4 genes as potential sources of the association signal
at 1p36. ENO1 encodes a glycolytic enzyme and the integral membrane protein SLC22A5 is
associated with skeletal myopathy, whereas the PDLIM4 protein is involved in the pathway of
actin cytoskeleton remodeling, and bone and skeletal muscle development. The rs28524846
SNP at 14q23 is an eQTL for MPP5 and APT6V1D in nerve tissue, with the MPP5 protein
known to regulate myelinating Schwann cells and ATP6V1D related to synaptic vesicle cycle
and ATPase activity.
Encouragingly, the two largest GWAS to date did successfully detect genome-​wide sig-
nificant associations with PA. In a Japanese sample (N=16,016), Hara et al. (28) identified an
association (Pmeta=2.2×10−9) between LTPA and rs10252228, a SNP located in the intergenic
region between the NPSR1 and DPY19L1 genes. Although functional links of intergenic SNPs
to nearby genes should be interpreted with caution, Hara et al. (28) point out that the product
of the NPSR1 and that of the DNAPTP6 genes previously found by the GWAS of de Moor
et al. (18) are both involved in pulmonary function. Variants in NPSR1 have been shown to
be associated with asthma-​related phenotypes, and DNAPTP6 could be involved in broncho-
dilator response via the downregulation of ß2-​adrenergic receptors. Impairments of pulmonary
function could be a barrier to the adoption of (vigorous) PA.
In the UK Biobank (N=380,492), Klimentidis et al. (39) detected genome-​wide signifi-
cant genetic associations with touchscreen-​survey based measures of MVPA. After applying
corrections for work-​related PA and an indicator of socioeconomic status, associations with
MVPA were found at rs429358 (APOE), rs169504 (PBX2), rs4129572 (EXOC4), rs3094622
(RPP21), rs181220614 (ARHGEF26-​AS1), rs149943 (ZNF165), and rs2988004 (PAX5).
Intriguingly, the Alzheimer disease risk allele (E4) of the APOE gene was associated with higher
levels of MVPA. As noted by Klimentidis et al. (39), APOE e4 carriers had a more favorable
response to exercise, for instance, as suggested by a study showing larger aerobic fitness in
response to training (56). This is in keeping with the theoretical notion that individuals who
have higher exercise ability and/​or trainability will find it easier to adopt regular exercise as a
lifestyle (12, 13, 41).
From the UK Biobank self-​reports (39), two dichotomous measures of more vigorous PA
were defined: VPA by classifying participants as engaged in vigorous PA if they spend more
than 25 minutes on activities “that make you sweat or breathe hard” for 3 or more days a week
and VEB by classifying participants as regular vigorous exercisers if they spend more than 15
minutes on 2 or more days a week doing strenuous sports or other exercises. Significant genetic

75
76

M.D. van der Zee and E. de Geus

association for VPA was found at rs1248860 (CADM2), rs2764261 (FOXO3), rs3781411
(CTBP2), rs12707131 (EXOC4), and rs328919 (DPY19L1) and for VEB at rs62253088
(CADM2), rs166840 (AKAP10), rs10946808 (HIST1H1D), rs75930676 (SIPA1L1), and
rs4865656 (LOC642366). In addition, accelerometer data were available in just over 91,000
participants. From up to 7 days of accelerometer wear, overall acceleration was obtained as
a measure of TPA and the fraction of accelerations >425 mg as a measure of VPA, yielding
significant associations at rs55657917 (CRHR1) and rs185829646 (ANKRD22) for TPA and
rs743580 (PML) and rs6433478 (CIR1) for VPA.
The CADM2 gene, primarily expressed in the brain, surfaced in the UK Biobank study as a
gene influencing vigorous sports and exercise-​related PA. Previously this gene had been linked
to risk-​taking behavior and extraversion (6) as well as executive function (33). This shows a
remarkable parallel to results reported in the largest candidate study to date by van der Mee et al.
(57) where a polygenic dopaminergic risk score that summed the increaser alleles in COMT
and DAT1 for both executive function and reward sensitivity was associated with the volume
of externally paced sports and exercise activities. Possibly, CADM2, like these dopaminergic
genes operates through a “double whammy” of increased reward value of exercise and increased
sports skills.

Conclusion
Although the gradual rise in heritability of PA seen in adolescence is not continued in adulthood,
it is clear that genetic factors remain a major contributor to individual differences in adult PA
behaviors. In the reports on over 283,904 adult twins, we find that about half of the variance
in the four PA phenotypes can be explained by genetic factors (males 48%; 95% CI 44–​52%;
females 47%; 95% CI 44–​50%). The substantial contribution of the shared environment (C) to
childhood PA seems to have largely dissipated in adulthood (males and females 6%; 95% CI 3–​
10%).There is generally a good correspondence in estimates across individual studies even when
they use samples from different countries or measures. For TPA and MVPA, objective measure-
ment showed slightly higher genetic estimates than subjective measurement, but the differences
are not striking. Major sex differences in genetic architecture of LTPA and VEB seemed to be
absent; for TPA and MVPA estimates, the verdict is still out as analyses were mostly based on
combined male/​female samples.
The extant studies using a candidate gene, linkage, or whole-​genome association approach
have yielded a number of genes and variants worth careful monitoring in future gene-​finding
efforts. Previously, three biological themes have emerged as a potential source for “phys-
ical activity genes” from theory: (1) the brain circuitry related to motivational and affective
aspects of PA; (2) the brain circuitry involved in the maintenance of energy intake/​expend-
iture balance; and (3) the physiological determinants of the ability to perform (intense and/​
or prolonged) PA, ideally at an above average level (12, 13, 41). Many of the variants reviewed
above do seem to fit these theoretical notions, but we issue a note of caution that this may
partly reflect our deep desire (and uncanny ability) to make sense of data by “reasoning
towards” our theoretical models. Rigorous replication followed by experimental validation
(in animal models) is direly needed. However, the experience from many other complex
behavioral traits allows us to end upbeat: once very large samples are amassed in international
meta-​analytic consortia, we can confidently expect large progress in our understanding of
how PA is regulated by genetics.

76
77

Physical activity studies in humans

References
1. World Health Organization. Global Recommendations on Physical Activity for Health. Switzerland: World
Health Organization, 2010.
2. Aaltonen S, Ortega-​Alonso A, Kujala UM, and Kaprio J. Genetic and environmental influences
on longitudinal changes in leisure-​time physical activity from adolescence to young adulthood. Twin
Res Hum Genet 16: 535–​543, 2013.
3. Aaltonen S, Ortega-​Alonso A, Kujala UM, and Kaprio J. A longitudinal study on genetic and
environmental influences on leisure time physical activity in the Finnish Twin Cohort. Twin Res Hum
Genet 13: 475–​481, 2010.
4. Ainsworth BE, Haskell WL, Herrmann SD, Meckes N, Bassett DR, Jr., Tudor-​Locke C,
Greer JL, Vezina J, Whitt-​Glover MC, and Leon AS. 2011 Compendium of Physical Activities: a
second update of codes and MET values. Med Sci Sports Exerc 43: 1575–​1581, 2011.
5. Bassett DR, Jr, Cureton AL, and Ainsworth BE. Measurement of daily walking distance-​
questionnaire versus pedometer. Med Sci Sports Exerc 32: 1018–​1023, 2000.
6. Boutwell B, Hinds D, 23andMe Research Team, Tielbeek J, Ong KK, Day FR, and Perry
JRB. Replication and characterization of CADM2 and MSRA genes on human behavior. Heliyon
3: e00349, 2017.
7. Cai GW, Cole SA, Butte N, Bacino C, Diego V, Tan K, Goring HH, O’Rahilly S, Farooqi IS,
and Comuzzie AG. A quantitative trait locus on chromosome 18q for physical activity and dietary
intake in Hispanic children. Obesity (Silver Spring) 14: 1596–​1604, 2006.
8. Carlsson S, Andersson T, Lichtenstein P, Michaelsson K, and Ahlbom A. Genetic effects on
physical activity: Results from the Swedish twin registry. Med Sci Sports Exerc 38: 1396–​1401, 2006.
9. Cole SA, Butte NF, Voruganti VS, Cai G, Haack K, Kent JW Jr, Blangero J, Comuzzie AG,
McPherson JD, and Gibbs RA. Evidence that multiple genetic variants of MC4R play a func-
tional role in the regulation of energy expenditure and appetite in Hispanic children. Am J Clin Nutr
91: 191–​199, 2010.
10. Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE, Pratt M,
Ekelund U, Yngve A, Sallis JF, and Oja P. International physical activity questionnaire: 12-​country
reliability and validity. Med Sci Sports Exerc 35: 1381–​1395, 2003.
11. de Geus EJ, Boomsma DI, and Snieder H. Genetic correlation of exercise with heart rate and
respiratory sinus arrhythmia. Med Sci Sports Exerc 35: 1287–​1295, 2003.
12. de Geus EJ and de Moor MHM. Genes, exercise, and psychological factors. In: Genetic and Molecular
Aspects of Sport Performance, edited by Bouchard C and Hoffman EP. Chichester: Wiley, 2011, p. 294–​305.
13. de Geus EJ and de Moor MHM. A genetic perspective on the association between exercise and
mental health. Ment Health Phys Act 1: 53–​61, 2008.
14. de Moor MH, Posthuma D, Hottenga JJ, Willemsen G, Boomsma DI, and de Geus EJ.
Genome-​ wide linkage scan for exercise participation in Dutch sibling pairs. Eur J Hum Genet
15: 1252–​1259, 2007.
15. de Moor MH, Stubbe JH, Boomsma DI, and de Geus EJ. Exercise participation and self-​rated
health: do common genes explain the association? Eur J Epidemiol 22: 27–​32, 2007.
16. de Moor MH, Willemsen G, Rebollo-​Mesa I, Stubbe JH, de Geus EJ, and Boomsma DI.
Exercise participation in adolescents and their parents: Evidence for genetic and generation specific
environmental effects. Behav Genet 41: 211–​222, 2011.
17. de Moor MH and de Geus, EJ. Genetic influences on exercise behavior. In: Lifestyle Medicine,
edited by Rippe JM. Boca Raton, FL: Taylor & Francis Group, 2012, p. 1367–​1378.
18. de Moor MH, Liu YJ, Boomsma DI, Li J, Hamilton JJ, Hottenga JJ, Levy S, Liu XG, Pei YF,
Posthuma D, Recker RR, Sullivan PF, Wang L, Willemsen G, Yan H, de Geus EJ, and Deng
HW. Genome-​wide association study of exercise behavior in Dutch and American adults. Med Sci
Sports Exerc 41: 1887–​1895, 2009.
19. de Moor MH, Posthuma D, Hottenga JJ, Willemsen AHM, Boomsma DI, and de Geus
EJC. Genome-​wide linkage scan for exercise participation in Dutch sibling pairs. Eur J Hum Genet
15: 1252–​1259, 2007.
20. den Hoed M, Brage S, Zhao JH, Westgate K, Nessa A, Ekelund U, Spector TD, Wareham NJ,
and Loos RJ. Heritability of objectively assessed daily physical activity and sedentary behavior. Am J
Clin Nutr 98: 1317–​1325, 2013.

77
78

M.D. van der Zee and E. de Geus

21. Duncan GE, Cash SW, Horn EE, and Turkheimer E. Quasi-​causal associations of physical
activity and neighborhood walkability with body mass index: a twin study. Prev Med 70: 90–​95, 2015.
22. Eriksson M, Rasmussen F, and Tynelius P. Genetic factors in physical activity and the equal envir-
onment assumption –​the Swedish young male twins study. Behav Genet 36: 238–​247, 2006.
23. Frederiksen H and Christensen K. The influence of genetic factors on physical functioning and
exercise in second half of life. Scand J Med Sci Sports 13: 9–​18, 2003.
24. Gielen M, Westerterp-​Plantenga MS, Bouwman FG, Joosen AM, Vlietinck R, Derom C,
Zeegers MP, Mariman EC, and Westerterp KR. Heritability and genetic etiology of habitual
physical activity: A twin study with objective measures. Genes Nutr 9: 415, 2014.
25. Hagstromer M, Ainsworth BE, Oja P, and Sjostrom M. Comparison of a subjective and an
objective measure of physical activity in a population sample. J Phys Act Health 7: 541–​550, 2010.
26. Hagstromer M, Troiano RP, Sjostrom M, and Berrigan D. Levels and patterns of objectively
assessed physical activity –​a comparison between Sweden and the United States. Am J Epidemiol
171: 1055–​1064, 2010.
27. Hansen BH, Kolle E, Dyrstad SM, Holme I, and Anderssen SA. Accelerometer-​determined
physical activity in adults and older people. Med Sci Sports Exerc 44: 266–​272, 2012.
28. Hara M, Hachiya T, Nishida Y, Shimanoe C, Tanaka K, Sutoh Y, and Shimizu A. Genome-​
wide association study of leisure-​1 time exercise behavior in Japanese adults. Med Sci Sports Exerc, 2018.
29. Heller RF, O’Connell DL, Roberts DC, Allen JR, Knapp JC, Steele PL, and Silove D. Lifestyle
factors in monozygotic and dizygotic twins. Genet Epidemiol 5: 311–​321, 1988.
30. Huppertz C, Bartels M, de Geus EJ, van Beijsterveldt CEM, Rose RJ, Kaprio J, and
Silventoinen K. The effects of parental education on exercise behavior in childhood and youth: A
study in Dutch and Finnish twins. Scand J Med Sci Sports 27: 1143–​1156, 2017.
31. Huppertz C, Bartels M, de Zeeuw EL, van Beijsterveldt CEM, Hudziak JJ, Willemsen G,
Boomsma DI, and de Geus EJ. Individual differences in exercise behavior: Stability and change in
genetic and environmental determinants from age 7 to 18. Behav Genet 46: 665–​679, 2016.
32. Huppertz C, Bartels M, Jansen IE, Boomsma DI, Willemsen G, de Moor MH, and de Geus
EJ. A twin-​sibling study on the relationship between exercise attitudes and exercise behavior. Behav
Genet 44: 45–​55, 2014.
33. Ibrahim-​Verbaas CA, Bressler J, Debette S, Schuur M, Smith AV, Bis JC, Davies G, Trompet
S, Smith JA, Wolf C, et al. GWAS for executive function and processing speed suggests involvement
of the CADM2 gene. Mol Psychiatry 21: 189–​197, 2016.
34. Joosen AM, Gielen M, Vlietinck R, and Westerterp KR. Genetic analysis of physical activity in
twins. Am J Clin Nutr 82: 1253–​1259, 2005.
35. Karvinen S, Waller K, Silvennoinen M, Koch LG, Britton SL, Kaprio J, Kainulainen H, and
Kujala UM. Physical activity in adulthood: Genes and mortality. Sci Rep 5: 18259, 2015.
36. Kawai T, Morita K, Masuda K, Nishida K, Sekiyama A, Teshima-​Kondo S, Nakaya Y, Ohta
M, Saito T, and Rokutan K. Physical exercise-​associated gene expression signatures in peripheral
blood. Clin J Sport Med 17: 375–​383, 2007.
37. Kemper HC, Twisk JW, and van Mechelen W. Changes in aerobic fitness in boys and girls over a
period of 25 years: Data from the Amsterdam Growth And Health Longitudinal Study revisited and
extended. Pediatr Exerc Sci. 25: 524–​35, 2013.
38. Kim J, Kim J, Min H, Oh S, Kim Y, Lee AH, and Park T. Joint identification of genetic variants
for physical activity in Korean population. Int J Mol Sci 15: 12407–​12421, 2014.
39. Klimentidis YC, Raichlen DA, Bea J, Garcia DO, Mandarino LJ, Alexander GE, Chen
Z, and Going SB. Genome-​wide association study of habitual physical activity in over 377,000
UK Biobank participants identifies multiple variants including CADM2 and APOE. Int J Epidemiol,
42: 1161–​1176, 2018.
40. Kujala UM, Kaprio J, and Koskenvuo M. Modifiable risk factors as predictors of all-​cause mor-
tality: The roles of genetics and childhood environment. Am J Epidemiol 156: 985–​993, 2002.
41. Lightfoot JT, de Geus EJ, Booth FW, Bray MS, den Hoed M, Kaprio J, Kelly SA, Pomp D,
Saul MC, Thomis MA, Garland T, Jr, and Bouchard C. Biological/​genetic regulation of physical
activity level: Consensus from GenBioPAC. Med Sci Sports Exerc 50:863–​873, 2017.
42. Lin X, Chan KK, Huang Y-​T, Luo X, Liang L, Wilson J, Correa A, Levy D, and Liu S. Genetic
determinants for leisure-​time physical activity. Med Sci Sports Exerc 50: 1620–​1628, 2018.
43. Loos RJF, Rankinen T, Tremblay A, Pérusse L, Chagnon Y, and Bouchard C. Melanocortin-​
4 receptor gene and physical activity in the Québec Family Study. Int J Obes 29: 420–​428, 2004.

78
79

Physical activity studies in humans

44. McCaffery JM, Papandonatos GD, Bond DS, Lyons MJ, and Wing RR. Gene X environment
interaction of vigorous exercise and body mass index among male Vietnam-​era twins. Am J Clin Nutr
89: 1011–​1018, 2009.
45. Mustelin L, Joutsi J, Latvala A, Pietilainen KH, Rissanen A, and Kaprio J. Genetic influences
on physical activity in young adults: A twin study. Med Sci Sports Exerc 44: 1293–​1301, 2012.
46. Mustelin L, Latvala A, Pietilainen KH, Piirila P, Sovijarvi AR, Kujala UM, Rissanen A, and
Kaprio J. Associations between sports participation, cardiorespiratory fitness, and adiposity in young
adult twins. J Appl Physiol 110: 681–​686, 2011.
47. Nelson MC, Gordon-​Larsen P, North KE, and Adair LS. Body mass index gain, fast food, and
physical activity: effects of shared environments over time. Obesity (Silver Spring) 14: 701–​709, 2006.
48. Plomin R, DeFries JC, Knopik V, and Niederheiser J. Behavioral Genetics. New York: Worth, 2013.
49. Rico-​Sanz J, Rankinen T, Rice T, Leon AS, Skinner JS, Wilmore JH, Rao DC, and Bouchard
C. Quantitative trait loci for maximal exercise capacity phenotypes and their responses to training in
the HERITAGE Family Study. Physiol Genomics 16: 256–​260, 2004.
50. Schutte NM, Bartels M, and de Geus EJ. Genetics of physical activity and physical fitness.
In: Oxford Textbook of Children’s Sport and Exercise Medicine (3rd ed.), edited by Armstrong N and van
Mechelen W. New York: Oxford University Press, 2017, p. 293–​302.
51. Simonen R, Levalahti E, Kaprio J, Videman T, and Battie MC. Multivariate genetic analysis of
lifetime exercise and environmental factors. Med Sci Sports Exerc 36: 1559–​1566, 2004.
52. Simonen RL, Rankinen T, Perusse L, Rice T, Rao DC, Chagnon Y, and Bouchard C.
Genome-​wide linkage scan for physical activity levels in the Quebec family study. Med Sci Sports
Exerc 35: 1355–​1359, 2003.
53. Spinath FM, Wolf H, Angleitner A, Borkenau P, and Riemann R. Genetic and environmental
influences on objectively assessed activity in adults. Pers Individ Dif 33: 633–​645, 2002.
54. Stubbe JH, Boomsma DI, and De Geus EJ. Sports participation during adolescence: A shift from
environmental to genetic factors. Med Sci Sports Exerc 37: 563–​570, 2005.
55. Stubbe JH, de Moor MH, Boomsma D, and de Geus EJ. The association between exercise par-
ticipation and well-​being: A co-​twin study. Prev Med 44: 148–​152, 2006.
56. Thompson PD, Tsongalis GJ, Seip RL, Bilbie C, Miles M, Zoeller R, Visich P, Gordon
P, Angelopoulos TJ, Pescatello L, Bausserman L, and Moyna N. Apolipoprotein E geno-
type and changes in serum lipids and maximal oxygen uptake with exercise training. Metabolism
53: 193–​202, 2004.
57. van der Mee DJ, Fedko IO, Hottenga JJ, Ehli EA, van der Zee MD, Ligthart L, van
Beijsterveldt TCEM, Davies GE, Bartels M, Landers JG, and de Geus, EJC. Dopaminergic gen-
etic variants and voluntary externally paced exercise behavior. Med Sci Sports Exerc 50: 700–​708, 2017.
58. Vink JM, Boomsma DI, Medland SE, de Moor MH, Stubbe JH, Cornes BK, Martin NG,
Skytthea A, Kyvik KO, Rose RJ, Kujala UM, Kaprio J, Harris JR, Pedersen NL, Cherkas L,
Spector TD, and de Geus EJ. Variance components models for physical activity with age as modi-
fier: A comparative twin study in seven countries. Twin Res Hum Genet 14: 25–​34, 2011.
59. Visscher PM, Wray NR, Zhang Q, Sklar P, McCarthy MI, Brown MA, and Yang J. 10 Years of
GWAS discovery: Biology, function, and translation. Am J Hum Genet 101: 5–​22, 2017.
60. Westerterp KR. Physical activity and physical activity induced energy expenditure in
humans: Measurement, determinants, and effects. Fron Psychol 4: 90, 2013.

79
80

7
THE EVOLUTION OF GENETIC
MECHANISMS CONTROLLING
PHYSICAL ACTIVITY
J. Timothy Lightfoot, Ayland C. Letsinger, and Jorge Z. Granados

Introduction
As reviewed in earlier chapters, there is an extensive literature that suggests that physical activity
is genetically regulated, even though the mechanisms controlling this genetic influence have
not been definitively identified. While there are multiple groups working to understand how
genetically linked mechanisms may regulate physical activity, there have been few efforts
devoted to understanding how these mechanisms originally developed in humans. Two alter-
native explanations can be proposed toward this end: 1) the genetic mechanisms that control
physical activity were just “built-​in” to all mammals as part of the basic set of physiological
controllers; or 2) the genetic mechanisms that control physical activity evolved over time due
to drift, selection, or mutation (54). The former explanation does not hold if it is assumed
that humans evolved from single-​celled organisms originally, which is the so-​called universal
common ancestry (UCA) theory. The UCA theory was initially proposed by Darwin (18) and
is currently accepted by most evolutionary biologists with recent quantitative results supporting
UCA over a variety of competing, alternative hypotheses (60). If then the genetic mechanisms
regulating physical activity evolved over time in mammals, what combination of drift, selection,
and mutation resulted in genetic control of activity? Further, if the physical activity regulating
genetic mechanisms evolved, which –​if any –​potential external factors may have contributed
to this evolution?
To a large extent, these questions devolve to the following: Why would animals –​including
humans –​have moved? Why were they active? And why would humans have evolved
mechanisms that regulated how much movement was performed? There is no doubt that the
physiologies of all mammals are made for locomotion of some kind, whether it be walking,
running, galloping, trotting, and/​or swimming, but why move? There is a long history of psy-
chological research into the reason for human motivation, with some of the best-​known work
by Maslow (46) suggesting that at the root of motivation is the need to fulfill basic needs such
as hunger, safety, waste removal, and sex among others. Indeed, as related to physical activity,
early anthropological observations in hunter/​gatherer tribes suggested that with fulfillment of
nutrient demands, individuals in these tribes were not necessarily physically active (35). Further,
an argument can be made that physical activity-​induced caloric expenditure above the min-
imum necessary for collecting nutrition in an environment of limited dietary resources would

80
81

Genetic control of physical activity

be adverse to the survival of the individual and communal unit. However, if the genetic regu-
lation of physical activity developed in an inverse relationship to food availability as we have
postulated elsewhere (i.e., less food, more activity to find food; or more food, less need for
activity) (41), it is uncertain why humans still have the drive to be active given the current food
abundance (in general).While Maslow (46) specifically cautioned against applying his theory to
animals, why are some animal species (e.g., rodents) still active when they are safe and have an
ad libitum diet available? For example, we have previously highlighted several different strains of
inbred mice that run long distances on a daily basis (~7–​12 km/​day) in just such conditions (38)
with mice in the wild also choosing to run voluntarily on a running wheel (47). These types
of observations have led us to further consider why humans (and animals) would have evolved
mechanisms to control daily physical activity or if this type of activity control is just a byproduct
of the pursuit of basic needs.
The reader may wonder what use this type of inquiry would have. After all, it is extremely
difficult to experimentally test any potential selection pressure or mutation that may have caused
these activity-​regulating mechanisms to evolve (however, it should be noted that Garland and
Kelly have published methods to invoke experimental evolution in rodents that could be used
in some experimental conditions) (25). As such, at best, any answers derived from considering
how physical activity genetic regulating mechanisms evolved may be considered “educated
speculation.” To this point, Washburn and Lancaster (62) noted:

In speculations of this kind, it is well to keep the purpose of the speculation and the
limitation of the evidence in mind. Our aim is to understand human evolution. What
shaped the course of human evolution was a succession of successful adaptations, both
biological and cultural. These may be inferred in part from the direct evidence of the
archeological record. But the record is very incomplete … As we go farther back in
time, there is less evidence and the biological and cultural difference becomes progres-
sively greater.Yet it was in those remote times that the human way took shape, and it
is only through speculation that we may gain some insights into what the life of our
ancestors may have been.

In this case, insights into how genetic mechanisms controlling activity evolved may be useful in
two ways: 1) these insights may prove informative in describing and understanding the actual
genetic mechanisms that regulate activity, which at this point are fairly ambiguous; and 2) these
insights may either support or undermine the use of animal models in working to understand
human genetic regulation of activity. In regard to the former, given the lack of definitive identi-
fied genetic mechanisms that control physical activity (42), any information regarding potential
mutations or selection pressures that led to the development of activity-​regulating mechanisms
could help better direct research trying to identify these mechanisms. For example, we had
previously suggested that food availability could have been a potential selection pressure for
the development of the genetic regulation of physical activity (41). This suggestion has led us
to investigate whether food availability controls physical activity, with a multitude of studies
showing that moderate caloric restriction increases activity in a wide variety of mammals (e.g.,
27, 66) and recent studies showing that caloric overfeeding decreases physical activity signifi-
cantly (average decrease ~62%) (61).
Perhaps a more critical reason to consider the evolutionary factors for the development of
physical activity genetic regulators is to determine whether the animal models that have been
extensively used in this field of investigation are actually valid models of the genetic control
of human activity. For example, we have written extensively in other papers (38) as to why we

81
82

J.T. Lightfoot et al.

believe that a mouse wheel-​running model is a valid model of human physical activity. Briefly,
mouse wheel-​running mimics the physiological conditions that occur in humans during vol-
untary exercise, wheel-​running activates the same neural transmitters (unlike forced tread-
mill activity in rodents), and most importantly, humans and mice self-​select similar exercise
intensities when voluntarily running. If the genetic mechanisms controlling physical activity
in humans evolved before the basic mammalian species split –​for example, it is estimated that
mice and humans split on the evolutionary tree approximately 62–​100 million years ago (8,
30) –​then any insights gleaned from mice as to the genetic mechanisms responsible for physical
activity regulation would be potentially translatable to humans (and the converse would be true
as well) (39).Therefore, while perhaps not as empirically testable as other hypotheses in the field
of genetic regulation of physical activity, the exploration of potential factors that led to the evo-
lution of genetic controllers of physical activity could be used to direct research investigations.
A common question that arises is whether the evolutionary factors that drove the develop-
ment of the genetic mechanisms controlling physical activity are similar to those that drove the
development of genetic mechanisms that may control exercise endurance. While we have dir-
ectly addressed that question in other papers (e.g., 41), others have written extensively regarding
potential evolutionary selection pressures that arose approximately 50,000 years ago to drive
the development of exercise endurance (11). As such, the current literature suggests the genetic
mechanisms that control physical activity and those that control exercise endurance, while per-
haps related, are distinct mechanisms that probably arose at different times. For example, it has
been shown that the genomic quantitative trait loci (QTL) associated with exercise endurance
(e.g., 10, 17, 43, 56) and physical activity (e.g., 19, 38, 40) do not overlap or colocalize in either
human or animal models. Additionally, it has been shown repeatedly that exercise endurance and
activity levels are weakly correlated in humans (both adults and children) and animal models
(34, 36). Lastly, genetic regulation of physical activity appears to include an additional central
drive-​controlling element (42) that has not been included in discussions of genetic regulation
of exercise endurance. Thus, while there may be some phenotypic overlap, given the differing
QTL sets, the lack of a strong correlation between physical activity level and exercise endurance,
as well as the probably different regulating mechanisms, the consideration of the development
of genetic mechanisms regulating physical activity level should be treated separately from those
genetic mechanisms controlling exercise endurance (for a discussion of the evolution of exercise
endurance, see Chapter 11).
In the majority of cases, selection pressures are often erroneously considered the only factors
involved in evolution. However, it is well established that evolution occurs as a result of some
combination of selection pressures, random genetic drift, and/​or mutation events that are
dependent upon a variety of factors including population size (54). However, our previous
writings on the evolution of physical activity genetic regulation –​and those in the exercise
endurance areas –​have primarily been focused on potential selection pressures. Thus, the first
section below will consider potential selection pressures acting on physical activity requirements,
while the concepts of potential random drift and/​or mutation will be returned to in the latter
section of the chapter as we take up the consideration of when these mechanisms evolved.

Potential selection pressures driving the development of

genetic regulation of physical activity
In many cases, potential selection pressures are the first factors that are considered when inves-
tigating the reasons for evolution of a trait. While selection pressures may not be the reason for
the evolution of a trait (as discussed above), there are several potential pressures that have been

82
83

Genetic control of physical activity

previously discussed as relating directly to physical activity. Here we provide a brief overview
of these potential selection pressures and current data arguing for or against their inclusion in
future discussions.

Daily energy expenditure as a selection pressure

We had previously observed that energy expenditure (EE) may have been a potential selection
pressure in the evolution of physical activity (41). In our earlier data, there were no significant
differences in daily EEs (corrected by weight) between pre-​technology farmers and hunter-​
gatherer populations, with both populations having higher daily EE than Western populations.
This analysis was flawed, having not accounted for Kleiber’s law (31), whereas metabolic rates are
scaled to mass by a reciprocal of ¾.The application of Kleiber’s law has been illustrated recently
by Pontzer and colleagues in several articles (e.g., 50, 52) where they have shown that regardless
of population –​pre-​technology farmer, hunter-​gatherer, or Western population –​total daily
energy expenditures (TEEs) are largely similar even though daily physical activity levels are
significantly different. Similar observations had been made earlier by Westerterp and Speakman
(63), but these observations were fully empirically validated by Pontzer and colleagues (53) and
later extended to several different mammalian species (50). Therefore, given that daily EEs were
probably not different between various populations of early humans, it is likely that daily EE was
not a selection pressure for the evolution of physical activity control mechanisms.

Food accessibility as a selection pressure

Interestingly, Pontzer has suggested that daily EE, rather than being dependent on daily activity,
is homeostatically controlled (50). Pontzer and colleagues suggest that homeostatic control
of daily EE in each mammalian species is largely set by both evolutionary and ontogenetic
responses to specific influences (50, 52). One specific influence in setting EE suggested by
Pontzer and colleagues is chronic food availability (49). The hypothesis that food availability
played a role in the evolution of EE homeostasis, is congruent with our own suggestions of
food availability playing a role in the development of the genetic mechanisms that control daily
physical activity (41).
Our earlier hypothesis that food availability would have been a selection pressure contrib-
uting to the evolution of the genetic regulation of physical activity was based on the copious
literature showing an inverse relationship between physical activity and food availability. It is
well established that moderate reductions in caloric intake significantly increase physical activity
in several animal species (e.g., 27, 66), with suggestions that the same responses occur in humans
(14). Conversely, two human studies have indirectly suggested that overfeeding significantly
decreased physical activity in specific situations (37, 57). Supporting these studies, are recent
direct observations that have shown that inbred mice placed on a chronic (9–​11-​week) high-​fat,
high-​sugar diet decreased daily voluntary wheel-​running distance by 70% in males and by 57%
in female mice (61). This decreased physical activity remained as long as the mice were exposed
to the altered diet; chronic exposure to a running wheel after the chronic dietary exposure
did not correct the activity inhibition. Further, the overfeeding-​induced activity reduction
remained even with exogenous sex hormone supplementation, eliminating overfeeding-​
induced changes in sex hormones as a potential mechanism for this physical activity reduction
(61). These conclusions are supported by earlier, more limited studies that also showed that
overfeeding reduced physical activity in rodents (9, 23, 55). Thus, while the mechanisms linking
caloric intake with physical activity regulation are unclear, it is clear that food availability is

83
84

J.T. Lightfoot et al.

linked to the control of physical activity. Given that food availability has been linked with the
evolution of the homeostatic control of daily EE (49), it is also possible that food availability was
a selection pressure involved with the evolution of physical activity genetic control mechanisms.
Logic difficulties exist with the argument that food availability is inversely related to phys-
ical activity (i.e., decreased food availability=increased activity and vice versa). The logical
inconsistencies would occur if food supply were extremely limited and activity was high;
from an individual standpoint, activity would be an energy wasteful mechanism which could
not be corrected given local food availability. From a community standpoint, increased phys-
ical activity would strain the energy resources available to sustain the larger populations.
When faced with a decrease in energy resources, the individual/​community necessarily has
to increase its foraging areas in the hope of finding new energy resources. The necessity to
increase the foraging range –​the so-​called patch-​leaving problem (59) –​would require a
higher physical activity drive. A naturally occurring example of decreased food availability/​
altered physical activity in nature is seen within ungulates (such as deer) and members of
Carnivora (such as wolves) populations with larger animals having larger home ranges (24).
In a study conducted on 49 different large mammals, Garland reported that members of
carnivora also tended to have larger home ranges when compared to ungulate/​herbivore
mammals, suggesting that the diversity in home range may be due to food availability (24).
This suggestion supports observations from hunter/​gatherer populations that foraging ranges
increased as local food supplies dwindled until it was more efficient to move the community
center to a location with greater food supplies (35). Thus, availability of food may have neces-
sarily impacted physical activity, especially early in the evolution of Homo sapiens, so that indi-
viduals with greater daily activity were more successful at obtaining food and as a result were
at an advantage for reproductive success. There are several studies in other mammalian species
suggesting that food availability can be a selection pressure for physiological adaptations. For
example, Ariettaz and colleagues (3) showed that within two sibling species of mouse-​ear bats,
food availability over 13 years of observations within each species niche drove alterations in
life-​history strategies. Thus, given the linkage between food availability and the regulation of
physical activity, as well as existing literature demonstrating that food availability can drive
adaptation, it is possible that food availability was a selection pressure that influenced the
activity patterns of early mammals.
Mechanisms by which food availability could have been a selection pressure for physical activity
control mechanisms are unclear and require further investigation.The adaptations could have been
as simple as those animals that were active during times of decreased food availability found more
food and thus, were more likely to reproduce, a situation similar to what has been observed in cliff
swallows (12).Another hypothesized mechanism that could be applied to understanding how food
availability was selective for physical activity control is based on neuroeconomics and the dopa-
mine thrift theory first proposed by Beeler and colleagues (5, 6). The dopamine thrift theory (4)
and the later application of neuroeconomics suggest that availability of resources (in this case food)
modifies the dopamine response to help the animal determine whether to be active or not to find
food (7). As noted, further investigation is needed to determine potential mechanisms through
which food availability drove the regulation of physical activity.

Duration of exercise as a potential selection pressure

Recently, it has been suggested that humans evolved faster metabolisms to feed our larger brains
in conjunction with an enhanced energy-​storing capability through higher levels of adipose

84
85

Genetic control of physical activity

tissue and more efficient means of utilizing fat as a fuel source (51). However, this evolutionary
trait came with the risk of starvation if the food was not plentiful, especially when humans
began to use agriculture to supply food versus hunting and foraging.While many factors played
a role in whether a pre-​technology farmer would be able to raise adequate foodstuffs, certainly
one of the more critical success factors was the willingness of the farmer to be active for the
large amounts of time pre-​technology agriculture required.
We have previously shown (41) that on a daily basis, males and females across several pre-​
technology agricultural societies were at least three to five times more active than hunter-​
gatherers and modern western societies (Figure 7.1) even if their total daily EE (53) were the
same. Given these parameters, the farmers’ physical activity intensity would have had to be
lower to allow for the lengthy duration of physical activity. The lower intensity and longer
duration of physical activity would have required the pre-​technology farmers to favor fat
rather than carbohydrates as a fuel source to fulfill the physical activity energy requirements
of agriculture. Thus, we hypothesize that there was a selective advantage in being able to
use fat to provide energy for long periods of physical activity at relatively low intensities
in successful farmers. Therefore, with the advent of agriculture approximately 11,000 years
ago (1), this long duration/​low-​intensity physical activity requirement, coupled with the
more efficient storage of fat in humans (53), may have been a significant selection pressure
on physical activity regulating mechanisms. It is unclear how the requirement for longer
duration/​lower intensity physical activity would have driven the selection of mechanisms
regulating physical activity. However, applying the work of Beeler and colleagues regarding
dopamine drift (7), the longer duration/​lower intensity selection pressure could have exerted
a neuroeconomic stimulus such that the dopaminergic mechanisms provided a decreased cost
sensitivity resulting in a stimulus for extended activity. Those farmers who had dopamine
systems that provided the appropriate cost/​benefit ratio in dopamine signaling would have
been more likely to be successful and thus, reproduce. However, regardless of how longer
duration/​lower activity intensities may have driven selection of physical activity regulation,
given the relatively recent widespread adoption of agriculture (~11,000 years ago), variants
in the genome to regulate physical activity should have developed at approximately the same
time as the advent of agriculture.

Figure 7.1 Hours per day spent in moderate or vigorous activity per day. Pre-​technology farmers were
three to five times more active than Western and hunter-​gatherer populations. H/​G=hunter-​gatherer
populations; Pre-​tech F=pre-​technology farmers.
Source: Adapted from data in 41.

85
86

J.T. Lightfoot et al.

Age of associated alleles

Determining the mutation age of alleles associated with physical activity regulation would pro-
vide information allowing further consideration of which, if any, processes were involved in the
evolution of genetic control of physical activity. For example, if the onset of farming/​agriculture
was a selection pressure in the development of physical activity regulation, then it would be
expected that the genomic alleles associated with activity would be approximately 11,000 years
old (1). With the use of physical activity-​associated single nucleotide polymorphisms (SNPs) as
well as techniques to estimate when alleles mutated (22, 29), it is possible to approximate when
physical activity-​related SNPs became prominent.
To date, there are four known human-​based genome-​wide association studies (GWAS) and
seven candidate gene studies that have identified physical activity-​associated SNPs (13, 16, 19, 26,
32, 44, 45, 64). Within these human-​based studies, there are 104 significant SNPs identified as
being associated with physical activity. When genomically mapped, these 104 SNPs are in a var-
iety of genomic locations with differing coding outcomes: six are located in exons, 49 are located
in introns, one is in a three-​prime untranslated region (3ʹ UTR, potential RNA stabilization), two
are considered upstream of a gene (potential promoter/​enhancer), three are considered down-
stream of a gene (potential transcription unit/​terminator), and 43 are intergenic.Thus, of the 104
unique physical activity-​related SNPs, only the six exon-​located SNPs may lead to direct alter-
ations of proteins from that gene. Using an age estimation technique (22) for the average muta-
tion age of the exon-​located SNPs, they appeared approximately 478.4 ± 327.5 thousand years
ago (kya) in African Americans and 542.1 ± 369.4 kya in European Americans (Figure 7.2).With
a total range of 200–​1000 kya, these origins are at least as old as the commonly accepted time of
emergence of anatomically modern humans (15, 48, 58) and predate the speculated time-​frame
of when exercise endurance evolved (11) by 4-​to 20-​fold, suggesting that the activity-​associated
protein-​coding SNPs have been conserved throughout modern human evolution. In fact, the age
of the physical activity-​related protein-​coding SNPs (ranging from 210.5 to 785.2 kya) is consid-
erably older than the average estimated age of all 6515 exon-​located human SNPs determined
thus far (47.6 ± 1.5 kya African Americans; 34.2 ± 0.9 kya European Americans) (22). Further,
the estimated ages of the physical activity-​related SNPs would suggest that genetic control of
physical activity evolved before the onset of agriculture (~11 kya) (1). However, considering just
the age of the physical activity-​associated exon-​located SNPs, it is unclear whether these SNPs
and their relationship to physical activity occurred as the result of one evolutionary event or as
an accumulation of many different evolutionary events.
While we have only been able to date the physical activity-​associated protein-​coding SNPs,
when ages can be estimated for the nonprotein-​coding activity-​related SNPs, a more refined
age estimation of the evolution of physical activity regulation will be possible. Additionally,
age-​dating the nonprotein-​ coding activity-​related SNPs may provide additional informa-
tion as to whether the evolution of physical activity occurred as one or multiple events. As
a very limited illustration of how this nonprotein-​coding information may alter our concept
of when genomic control of physical activity evolved, consider the estimated mutation age of
the only nonprotein-​coding physical activity-​associated SNP available. Using linkage disequi-
librium decay (29), the intron variant rs7910002 of the DNAJC1 gene was predicted to have
developed 7.8 kya, similar in time to the advent of agriculture (1).Thus, including the estimated
age of this nonprotein-​coding variant suggests that the evolution of the genetic regulation of
physical activity may have occurred in multiple eras, with the basic protein-​coding genomic
variants arising early in the development of modern humans, and further genomic alterations
occurring later, possibly in response to various environmental pressures. Further refinement

86
87

Genetic control of physical activity

Figure 7.2 Estimated mutation age of SNPs correlated to physical activity. Genes listed on the x-​axis
contain SNPs related to physical activity. “All SNPs” refers to the 6517 exon-​located SNPs in the study
by Fu et al. (22).

of the estimates of when genetic control of physical activity arose will be supported by future
determination of any additional physical activity-​associated SNPs as well as the identification of
the actual mechanisms involved.
A secondary purpose of estimating the age of physical activity-​associated SNPs is to deter-
mine whether animal models are likely to have similar genetic mechanisms underlying the regu-
lation of physical activity. For example, given that the evolutionary divergence of humans and
chimpanzees has been estimated to have occurred between 6.23–​7.07 million years ago (mya;
8, 30), our estimates of when human activity-​related SNPs evolved (average of ~478–​542 kya)
suggest humans may not share the same genetic regulators of activity as do primates. Using a
similar approach, given that the evolutionary split between humans and mice has been estimated
to have occurred approximately 62–​100 mya (8, 30), suggests there may be different genetic
mechanisms controlling activity in humans and mice. However, we urge extreme caution in
interpreting these limited data to indicate the absence of similarities between genetic regula-
tion of activity in humans and other species. In particular, the extremely small set of associated
genomic SNPs, the lack of understanding regarding all the genetic factors regulating physical
activity (e.g., miRNA, epigenetics), the uncertainty of whether physical activity genetic regu-
lation arose with one event or multiple events, and whether genomic SNPs actually indicate all
genetic mechanisms involved in physical activity regulation (20, 21), significantly weaken any
cross-​species application of our SNP-​age data.

87
88

J.T. Lightfoot et al.

Table 7.1 Human, Neanderthal, and chimpanzee allele comparisons of physical activity phenotypes

rs ID Gene Effect Neanderthalb Ancestral Allele

allelea allelec frequencyd

rs1051393 IFNAR2 Not given DDDDAD:0D1A G>A/​G/​T 0.003/​38/​62

rs16933006 Closest RPL7P3 C AAADAA:0D2A A>A/​C 82/​17
rs6025590 CTCFL G DAAAA_​:0D1A G>A/​G 27/​73
rs6454672 CNR1 C AADAAA:1D0A T>C/​T 15/​85
rs8066276 ACE T DAD_​DA:0D1A C>C/​T 38/​62
rs2267668 PPARD A DDDDAA:0D1A A>A/​G 85/​15
rs1638525 AKAP10 G DADDAD:0D1A G>C/​G 61/​39
rs35622985 MMS22L G AAAADA:0D1A G>A/​G 27/​73
rs1959759 DCAF5 A ADDDAA:0D2A A>A/​G 18/​82
rs10851869 PML T D_​A_​AA:0D1A C>C/​T 43/​56
rs10145335 Closest C14ord177 G DDDDAD:0D1A A>A/​G 20/​80
rs113351744 Closest LINC01029 G AADAAA:0D1A G>A/​G 2/​98
rs12460611 Closest CCNE1 A AA_​DAA:0D3A A>A/​G 83/​17

a Allele associated with higher amount of physical activity.

b First six characters represent the allele present in Human Reference San,Yoruba, Han, Papuan, and
French populations, A=ancestral, D=derived, or _​if not known. Following the colon is the amount
of derived or ancestral alleles in Neanderthal genomes. Bold suggests a selective sweep may have
occurred in the Homo sapiens lineage.
c The first character represents the chimpanzee reference allele followed by human alleles.
d Allele frequency=UCSC allele frequencies of human genome found in Ancestral Allelew column.

Due to the overlap of the historical timelines of Homo sapiens and Homo neanderthalensis and
the suggested gene flow (i.e., interbreeding) between the two species (33), it is possible to com-
pare allele frequencies to determine if any physical activity-​associated SNPs underwent positive
selection (i.e., selective sweeps) between the two species (28). Presence of a selective sweep
would suggest that a particular SNP provided increased survivability or reproductive fitness
for Homo sapiens. Thus, we compared modern-​day human SNPs related to physical activity to
Neanderthal genotypes analyzed by Green et al. (28) using the chimpanzee reference genome
to characterize ancestral genes. Due to incomplete genetic reading and high mutation rates at
CpG sites in the Neanderthal DNA, we can only determine signs of selective sweeps for 29 of
the 104 physical activity-​related SNPs. Of these 29 SNPs, four (rs1051393 exon in IFNAR2,
rs2267668 intron in PPARD, rs1638525 intron of AKAP10, and rs10145335 intergenic region
near C14ord177) were among the top 5% strongest signals of selective sweeps among a total
of 3,202,190 substitutions and 69,029 indels sequenced (Table 7.1). The strength of these
selective sweeps suggest strong positive selection pressures in Homo sapiens’ history compared to
Neanderthals for these particular mutations, i.e., Homo sapiens were more likely to inherit these
physical activity-​related SNPs. Thus, selective sweeps that have occurred since the divergence
of Homo sapiens and Homo neanderthalensis may suggests a genetic regulation of physical activity
unique to Homo sapiens.
As noted earlier, evolution can occur through selection pressure, mutation, or genetic drift.
Albeit with the limited state of current knowledge, it appears that genomic variants that are
correlated with physical activity levels play a role in a wide variety of biological mechanisms (see
Chapters 5, 6, 8, and 9), each with low effect sizes (i.e., physical activity levels between individ-
uals with different alleles are not largely different).Thus, due to the few physical activity-​related

88
89

Genetic control of physical activity

SNPs that have hypothesized selective sweeps (4 of 104) and/​or large effect sizes (0 of 104), in
conjunction with the small community populations that existed 478–​542 kya, it is likely that
genetic drift is the major evolutionary process responsible for the genetic mechanisms that con-
trol physical activity in humans. If it is the case that the majority of physical activity regulators
do not significantly influence reproductive viability in humans, especially in modern society
(41), the lack of a mutation in genetic sequences associated with physical activity which do not
kill or lower the host’s reproductive fitness (also known as stabilizing selection) could explain
why it appears the genetic mechanisms are highly diverse across species. Because of the lack
of large effect sizes for the physical activity-​related genomic variants, we hypothesize that the
genes containing the identified SNPs have a different primary function and only affect physical
activity indirectly. Supporting this hypothesis is a recent paper by Xu and Garland (65) where
they compared the genomic structure of four mouse lines which have been independently
bred for high physical activity since 1993. Xu and Garland surprisingly found that the genomic
structure associated with physical activity in these independent lines were largely different and
thus, these independent lines had evolved differing genomic approaches to increasing physical
activity levels. Applied to the current question, it is possible that humans have evolved a var-
iety of genetic mechanisms that control physical activity and the mechanisms that have evolved
depend somewhat on the population and the environmental exposures of the population. Given
the known population differences arising in the current large-​scale genomic mapping efforts
such as the International HapMap Consortium (2), it would not be surprising to discover mul-
tiple genetic mechanisms control physical activity.

Conclusion
The literature is very clear that physical activity is critical to health; however, physical activity
levels continue to be low in most human populations. While genetics have been clearly shown
to be an important factor in determining physical activity level, the exact genetic mechanisms
that regulate physical activity are still elusive. We have continued to explore how the physical
activity genetic regulating mechanisms may have evolved in the hope of informing future
mechanistic investigations. We provide Figure 7.3 as a summary of potential evolutionary
processes and factors that may have been involved in the development of genetic mechanisms
that regulate physical activity, building from the foundation of the pyramid toward the current-​
day physical activity regulation. From the existing literature, it is unlikely that EE was a selection
pressure that drove the evolution of physical activity genetic regulation. However, given the
inverse relationship between food availability and physical activity, the hypothesized food avail-
ability selection pressure that helped evolve homeostatic control of daily EE may have also been
involved in the evolution of physical activity regulation mechanisms. Since pre-​technology
farming populations generally exhibit much higher daily levels of physical activity duration
than either Western or hunter-​gatherer populations, this suggests that longer durations of daily
activity coupled with lower intensities may have played a role in the development of the genetic
physical activity regulating mechanisms. Attractive recent theories grounded in neuroeconomic
theory regarding “dopamine thrift” suggest potential mechanisms through which longer dur-
ation/​lower activity intensities and food availability pressures could have affected genetic con-
trol mechanisms. Juxtaposed against these theories are our calculations of 478–​542 kya as the
estimated age of the few protein-​coding human genomic variants that are currently associated
with physical activity. This time frame is much older than the onset of agriculture in human
populations, arguing against a role for agriculture in the development of these protein-​coding
genomic variants. However, while extremely tentative, dating of one of the nonprotein-​coding

89
90

J.T. Lightfoot et al.

Figure 7.3 Summary of potential evolutionary processes that developed genetic control of physical
activity. We propose that the foundational physical activity (PA) genetic regulatory mechanisms (base of
the pyramid) were evolved early in the development of Homo erectus through genetic drift mechanisms.
Later evolution of physical activity regulation (moving up the pyramid) was in response to selection
pressure (e.g., finding food), leading to the current mechanisms that regulate physical activity. *These
extrapolations are based on current data and subject to change with an increase in aging data available.

SNPs associated with physical activity put the age of development of that variant near the
onset of widespread agriculture. Thus, given the available data, we would propose that the
foundational evolutionary process that initially resulted in genetic mechanisms that control
physical activity was genetic drift (Figure 7.3), as opposed to either selection or mutation. We
further propose that these initial genetic drift-​induced mechanisms were altered in particular
populations by differing environmental pressures (e.g., food availability, demands of agriculture)
at differing times (Figure 7.3), resulting in a variety of genetic solutions to physical activity.
Thus, in conclusion, we hypothesize that the genetic control of physical activity in humans may
include a combination of mechanisms that arose in different time periods.
We acknowledge that many of our conclusions are based on extrapolations from
extremely limited data sets. However, this type of investigation is critical in working toward
an understanding of the biological and environmental factors that shaped humans’ physical
activity-​regulating mechanisms. As such, our investigations present a framework by which future
genomic-​associated information can be further used to refine insights into the responsible phys-
ical activity-​regulating mechanisms as well as the appropriateness of specific animal models.

Acknowledgments: This work was supported by funding from the Omar Smith Endowment
at Texas A&M University.

References
1. Abbo S, Lev-​Yadun S and Gopher A. Plant domestication and crop evolution in the near east: On
events and processes. Crit Rev Plant Sci 31: 241–​257, 2012.
2. Altshuler DM, Gibbs RA, Peltonen L, Altshuler DM, Gibbs RA, Peltonen L, Dermitzakis
E, Schaffner SF, Yu F, Peltonen L, Dermitzakis E, Bonnen PE, Altshuler DM, Gibbs RA,

90
91

Genetic control of physical activity

de Bakker PI, Deloukas P, Gabriel SB, Gwilliam R, Hunt S, Inouye M, Jia X, Palotie A,
Parkin M, Whittaker P, Yu F, Chang K, Hawes A, Lewis LR, Ren Y, Wheeler D, Gibbs RA,
Muzny DM, Barnes C, Darvishi K, Hurles M, Korn JM, Kristiansson K, Lee C, McCarrol
SA, Nemesh J, Dermitzakis E, Keinan A, Montgomery SB, Pollack S, Price AL, Soranzo N,
Bonnen PE, Gibbs RA, Gonzaga-​Jauregui C, Keinan A, Price AL, Yu F, Anttila V, Brodeur
W, Daly MJ, Leslie S, McVean G, Moutsianas L, Nguyen H, Schaffner SF, Zhang Q, Ghori
MJ, McGinnis R, McLaren W, Pollack S, Price AL, Schaffner SF, Takeuchi F, Grossman SR,
Shlyakhter I, Hostetter EB, Sabeti PC, Adebamowo CA, Foster MW, Gordon DR, Licinio J,
Manca MC, Marshall PA, Matsuda I, Ngare D, Wang VO, Reddy D, Rotimi CN, Royal CD,
Sharp RR, Zeng C, Brooks LD and McEwen JE. Integrating common and rare genetic variation
in diverse human populations. Nature 467: 52–​58, 2010.
3. Arlettaz R, Christe P and Schaub M. Food availability as a major driver in the evolution of life-​
history strategies of sibling species. Ecol Evol 7: 4163–​4172, 2017.
4. Beeler JA. Thorndike’s law 2.0: Dopamine and the regulation of thrift. Front Neurosci 6: 116, 2012.
5. Beeler JA, Frazier CR and Zhuang X. Dopaminergic enhancement of local food-​seeking is under
global homeostatic control. Eur J Neurosci 35: 146–​159, 2012.
6. Beeler JA, McCutcheon JE, Cao ZF, Murakami M, Alexander E, Roitman MF and Zhuang
X. Taste uncoupled from nutrition fails to sustain the reinforcing properties of food. Eur J Neurosci
36: 2533–​2546, 2012.
7. Beeler JA and Mourra D. To do or not to do: Dopamine, affordability and the economics of oppor-
tunity. Front Integr Neurosci 12: 6, 2018.
8. Benton MJ and Donoghue PC. Paleontological evidence to date the tree of life. Mol Biol Evol
24: 26–​53, 2007.
9. Bjursell M, Gerdin AK, Lelliott CJ, Egecioglu E, Elmgren A, Törnell J, Oscarsson J and
Bohlooly-​Y M. Acutely reduced locomotor activity is a major contributor to Western diet-​induced
obesity in mice. Am J Physiol Endocrinol Metab 294: E251–​E260, 2008.
10. Bouchard C, Rankinen T, Chagnon YC, Rice T, Pérusse L, Gagnon J, Borecki I, An P,
Leon AS, Skinner JS, Wilmore JH, Province M and Rao DC. Genomic scan for maximal
oxygen uptake and its response to training in the HERITAGE Family Study. J Appl Physiol (1985)
88: 551–​559, 2000.
11. Bramble DM and Lieberman DE. Endurance running and the evolution of Homo. Nature
432: 345–​352, 2004.
12. Brown CR, Brown MB, Roche EA, O’Brien VA and Page CE. Fluctuating survival selection
explains variation in avian group size. Proc Natl Acad Sci U S A 113: 5113–​5118, 2016.
13. Bruneau M, Walsh S, Selinsky E, Ash G, Angelopoulos TJ, Clarkson P, Gordon P, Moyna
N, Visich P, Zoeller R, Thompson P, Gordish-​Dressman H, Hoffman E, Devaney J and
Pescatello LS. A genetic variant in IL-​15Rα correlates with physical activity among European-​
American adults. Mol Genet Genomic Med 6: 401–​408, 2018.
14. Casper RC. The “drive for activity” and “restlessness” in anorexia nervosa: Potential pathways. J
Affect Disord 92: 99–​107, 2006.
15. Chase PG, Dibble HL, Lindly J, Clark G and Guy L. On the emergence of modern humans.
Current Anthropology 31: 58–​66, 1990.
16. Comuzzie AG, Cole SA, Laston SL, Voruganti VS, Haack K, Gibbs RA and Butte NF. Novel
genetic loci identified for the pathophysiology of childhood obesity in the Hispanic population. PLoS
One 7: e51954, 2012.
17. Courtney SM and Massett MP. Identification of exercise capacity QTL using association mapping
in inbred mice. Physiol Genomics 44: 948–​955, 2012.
18. Darwin C. The preservation of favoured races in the struggle for life. In: On the Origin of Species by
Means of Natural Section. London: J. Murray, 1859, Chapt. 14.
19. De Moor MH, Liu YJ, Boomsma DI, Li J, Hamilton JJ, Hottenga JJ, Levy S, Liu XG, Pei YF,
Posthuma D, Recker RR, Sullivan PF, Wang L, Willemsen G, Yan H, DeGeus EJ and Deng
HW. Genome-​wide association study of exercise behavior in Dutch and American adults. Med Sci
Sports Exerc 41: 1887–​1895, 2009.
20. Ferguson DP, Dangott LJ, Schmitt EE, Vellers HL and Lightfoot JT. Differential skeletal
muscle proteome of high-​and low-​active mice. J Appl Physiol 116: 1057–​1067, 2014.
21. Ferguson DP, Dangott LJ, Vellers HL, Schmitt EE and Lightfoot JT. Differential protein expres-
sion in the nucleus accumbens of high and low active mice. Behav Brain Res 291: 283–​288, 2015.

91
92

J.T. Lightfoot et al.

22. Fu W, O’Connor TD, Jun G, Kang HM, Abecasis G, Leal SM, Gabriel S, Rieder MJ, Altshuler
D, Shendure J, Nickerson DA, Bamshad MJ, NHLBI ESP and Akey JM. Analysis of 6,515
exomes reveals the recent origin of most human protein-​coding variants. Nature 493: 216–​220, 2013.
23. Funkat A, Massa CM, Jovanovska V, Proietto J and Andrikopoulos S. Metabolic adaptations
of three inbred strains of mice (C57BL/​6, DBA/​2, and 129T2) in response to a high-​fat diet. J Nutr
134: 3264–​3269, 2004.
24. Garland T, Jr, Dickerman A, Janis C and Jones J. Phylogenetic analysis of covariance by computer
simulation. Syst Biol 42: 265–​292, 1993.
25. Garland T, Jr and Kelly SA. Phenotypic plasticity and experimental evolution. J Exp Biol
209: 2344–​2361, 2006.
26. Gielen M, Westerterp-​Plantenga MS, Bouwman FG, Joosen AM, Vlietinck R, Derom C,
Zeegers MP, Mariman EC and Westerterp KR. Heritability and genetic etiology of habitual
physical activity: A twin study with objective measures. Genes Nutr 9: 415, 2014.
27. Goodrick CL, Ingram DK, Reynolds MA, Freeman JR and Cider NL. Effects of intermit-
tent feeding upon growth, activity, and lifespan in rats allowed voluntary exercise. Exp Aging Res
9: 203–​209, 1983.
28. Green RE, Krause J, Briggs AW, Maricic T, Stenzel U, Kircher M, Patterson N, Li H, Zhai
W, Fritz MH, Hansen NF, Durand EY, Malaspinas AS, Jensen JD, Marques-​Bonet T, Alkan
C, Prüfer K, Meyer M, Burbano HA, Good JM, Schultz R, Aximu-​Petri A, Butthof A, Höber
B, Höffner B, Siegemund M, Weihmann A, Nusbaum C, Lander ES, Russ C, Novod N,
Affourtit J, Egholm M, Verna C, Rudan P, Brajkovic D, Kucan Ž, Gušic I, Doronichev VB,
Golovanova LV, Lalueza-​Fox C, de la Rasilla M, Fortea J, Rosas A, Schmitz RW, Johnson
PLF, Eichler EE, Falush D, Birney E, Mullikin JC, Slatkin M, Nielsen R, Kelso J, Lachmann
M, Reich D and Pääbo S. A draft sequence of the Neandertal genome. Science 328: 710–​722, 2010.
29. Hawks J, Wang ET, Cochran GM, Harpending HC and Moyzis RK. Recent acceleration of
human adaptive evolution. Proc Natl Acad Sci U S A 104: 20753–​20758, 2007.
30. Hedges SB, Dudley J and Kumar S. TimeTree: a public knowledge-​base of divergence times
among organisms. Bioinformatics 22: 2971–​2972, 2006.
31. Kleiber M. Body size and metabolic rate. Physiol Rev 27: 511–​541, 1947.
32. Klimentidis YC, Raichlen DA, Bea J, Garcia DO, Mandarino LJ, Alexander GE, Chen
Z and Going SB. Genome-​wide association study of habitual physical activity in over 277,000
UK Biobank participants identifies multiple variants including CADM2 and APOE. Int J Obesity
42: 1161–​1176, 2018.
33. Kuhlwilm M, Gronau I, Hubisz MJ, de Filippo C, Prado-​Martinez J, Kircher M, Fu Q,
Burbano HA, Lalueza-​Fox C, de la Rasilla M, Rosas A, Rudan P, Brajkovic D, Kucan
Ž, Gušic I, Marques-​Bonet T, Andrés AM, Viola B, Pääbo S, Meyer M, Siepel A and
Castellano S. Ancient gene flow from early modern humans into Eastern Neanderthals. Nature
530: 429–​433, 2016.
34. Lamb KL and Brodie DA. Leisure-​time physical activity as an estimate of physical fitness: A valid-
ation study. J Clin Epidemiol 44: 41–​52, 1991.
35. Lee RB. The Dobe Ju/​’hoansi (3rd ed.). Belmont, CA: Wadsworth Publishing, 2003.
36. Lerman I, Harrison BC, Freeman K, Hewett TE, Allen DL, Robbins J and Leinwand LA.
Genetic variability in forced and voluntary endurance exercise performance in seven inbred mouse
strains. J Appl Physiol 92: 2245–​2255, 2002.
37. Levine JA, McCrady SK, Lanningham-​Foster LM, Kane PH, Foster RC and Manohar
CU. The role of free-​ living daily walking in human weight gain and obesity. Diabetes
57: 548–​554, 2008.
38. Lightfoot JT, Leamy LJ, Pomp D, Turner MJ, Fodor AA, Knab AM, Bowen RS, Ferguson
DP, Moore-​Harrison T and Hamilton A. Strain screen and haplotype association mapping of
wheel running in inbred mouse strains. J Appl Physiol 109: 623–​634, 2010.
39. Lightfoot JT, Bamman MM and Booth FW. Translation goes both ways: The power of reverse
translation from human trials into animal models. Transl J Am Coll Sports Med 2: 1–​3, 2017.
40. Lightfoot JT, Turner MJ, Pomp D, Kleeberger SR and Leamy LJ. Quantitative trait loci (QTL)
for physical activity traits in mice. Physiol Genomics 32: 401–​408, 2008.
41. Lightfoot JT. Why control activity? Evolutionary selection pressures affecting the development of
physical activity genetic and biological regulation. Biomed Res Int 2013: 821678, 2013.

92
93

Genetic control of physical activity

42. Lightfoot JT, De Geus EJC, Booth FW, Bray MS, den Hoed M, Kaprio J, Kelly SA, Pomp
D, Saul MC, Thomis MA, Garland T and Bouchard C. Biological/​genetic regulation of physical
activity level: Consensus from GenBioPAC. Med Sci Sports Exerc In Press: 2018.
43. Lightfoot JT, Turner MJ, Knab AK, Jedlicka AE, Oshimura T, Marzec J, Gladwell W, Leamy
LJ and Kleeberger SR. Quantitative trait loci associated with maximal exercise endurance in mice.
J Appl Physiol (1985) 103: 105–​110, 2007.
44. Lin X, Chan KK, Huang YT, Luo X, Liang L, Wilson J, Correa A, Levy D and Liu S. Genetic
determinants for leisure-​time physical activity. Med Sci Sports Exerc 50: 1620–​1628, 2018.
45. Mäestu J, Lätt E, Rääsk T, Sak K, Laas K, Jürimäe J and Jürimäe T. Ace I/​D polymorphism is
associated with habitual physical activity in pubertal boys. J Physiol Sci 63: 427–​434, 2013.
46. Maslow AH. A theory of human motivation. Psychol Rev 50: 370–​396, 1943.
47. Meijer JH and Robbers Y. Wheel running in the wild. Proc Roy Soc B 281: 20140210, 2014.
48. Nitecki MH and Nitecki DV. Origins of Anatomically Modern Humans. New York: Springer, 1994.
49. Pontzer H. Constrained total energy expenditure and the evolutionary biology of energy balance.
Exerc Sport Sci Rev 43: 110–​116, 2015.
50. Pontzer H. The crown joules: Energetics, ecology, and evolution in humans and other primates. Evol
Anthropol 26: 12–​24, 2017.
51. Pontzer H, Brown MH, Raichlen DA, Dunsworth H, Hare B, Walker K, Luke A, Dugas
LR, Durazo-​Arvizu R, Schoeller D, Plange-​Rhule J, Bovet P, Forrester TE, Lambert EV,
Thompson ME, Shumaker RW and Ross SR. Metabolic acceleration and the evolution of human
brain size and life history. Nature 533: 390–​392, 2016.
52. Pontzer H, Durazo-​Arvizu R, Dugas LR, Plange-​Rhule J, Bovet P, Forrester TE, Lambert
EV, Cooper RS, Schoeller DA and Luke A. Constrained total energy expenditure and metabolic
adaptation to physical activity in adult humans. Curr Biol 26: 410–​417, 2016.
53. Pontzer H, Raichlen DA, Wood BM, Mabulla AZ, Racette SB and Marlowe FW. Hunter-​
gatherer energetics and human obesity. PLoS One 7: e40503, 2012.
54. Powell R. The future of human evolution. Brit J Phil Sci 63: 145–​175, 2012.
55. Rendeiro C, Masnik AM, Mun JG, Du K, Clark D, Dilger RN, Dilger AC and Rhodes JS.
Fructose decreases physical activity and increases body fat without affecting hippocampal neurogen-
esis and learning relative to an isocaloric glucose diet. Sci Rep 5: 9589, 2015.
56. Rico-​Sanz J, Rankinen T, Rice T, Leon AS, Skinner JS, Wilmore JH, Rao DC and Bouchard
C. Quantitative trait loci for maximal exercise capacity phenotypes and their responces to training in
the HERITAGE Family Study. Physiol Genomics 16: 256–​260, 2004.
57. Schmidt SL, Harmon KA, Sharp TA, Kealey EH and Bessesen DH. The effects of overfeeding
on spontaneous physical activity in obesity prone and obesity resistant humans. Obesity (Silver Spring)
20: 2186–​2193, 2012.
58. Smith FH, Falsetti AB and Donnelly SM. Modern human origins. Am J Phys Anthrop
32: 35–​68, 1989.
59. Stephens DW and Krebs JW. Foraging Theory. Princeton, NJ: Princeton University Press, 1986.
60. Theobald DL. A formal test of the theory of universal common ancestry. Nature 465: 219–​222,
2010.
61. Vellers HL, Letsinger AC, Walker NR, Granados JZ and Lightfoot JT. High fat high sugar diet
reduces voluntary wheel running in mice independent of sex hormone involvement. Front Physiol 8:
628, 2017.
62. Washburn SL and Lancaster CS. The evolution of hunting. In: Man The Hunter, edited by Lee RB
and Irven D. Chicago, IL: Aldine Publishing Company, 1968, p. 293–​303.
63. Westerterp KR and Speakman JR. Physical activity energy expenditure has not declined since the
1980s and matches energy expenditures of wild mammals. Int J Obesity 32: 1256–​1263, 2008.
64. Wilkinson AV, Gabriel KP, Wang J, Bondy ML, Dong Q, Wu X, Shete S and Spitz MR.
Sensation-​seeking genes and physical activity in youth. Genes Brain Behav 12: 181–​188, 2013.
65. Xu S and Garland T, Jr. A mixed model approach to genome-​wide association studies for
selection signatures, with application to mice bred for voluntary exercise behavior. Genetics
207: 785–​799, 2017.
66. Yamada Y, Colman RJ, Kemnitz JW, Baum ST, Anderson RM, Weindruch R and Schoeller
DA. Long-​term calorie restriction decreases metabolic cost of movement and prevents decrease of
physical activity during aging in rhesus monkeys. Exp Gerontol 48: 1226–​1235, 2013.

93
94

8
NEUROGENETICS OF
MOTIVATION FOR
PHYSICAL ACTIVITY
Justin S. Rhodes

Introduction
Movement is energetically costly, yet many animals travel vast distances as part of their normal
life history. From a Darwinian evolutionary perspective, this would be unexpected unless it
provided some fitness advantage for survival or reproductive success. Movement is often neces-
sary to find food, shelter, or mates. Therefore, it is generally assumed that gene frequencies have
been shaped by natural selection to support varying levels of physical activity depending on the
local ecological conditions. At least two components are necessary for an individual to perform
a voluntary behavior of any kind: they must be physically capable of performing the behavior
and they must be motivated to do it. Both processes involve biological traits that are moldable
by evolution and which, in part, shape genetic variation in physical activity levels. Physical
constraints such as aerobic capacity (the maximal amount of oxygen an organism can consume
during maximal physical exertion), muscle mass or composition, bone symmetry, and many
others limit the ability of animals to move quickly or travel long distances (30). Hence, if eco-
logical conditions demand that the animal displays challenging physical acts at the limit of their
physical capability to survive and reproduce, then that will lead to the evolution of increased
capacity for these exercise physiological features, to the extent that they are heritable.
While physical capability sets the upper limit for what an animal can possibly accomplish,
animals typically choose not to perform at their limit. A substantial portion of variation in phys-
ical activity levels is expected to arise from differences in motivation for physical exertion (79).
The exercise physiological features that animals are genetically endowed with (including the full
extent of phenotypic plasticity) are usually far in excess of what the animal requires on a daily
basis or even within a lifetime. Consider sedentary humans compared to marathon runners: the
big difference is motivation not muscles (or capability).While it is certainly true that individuals
are constrained by the exercise physiological capacities they are endowed with, and there are
real biological differences between good and average marathon runners, the point is that most
people possess the physical capability of training their bodies to accomplish a marathon, but
they choose not to. On an evolutionary time scale, the principle that motivation contributes
to behavioral variation applies equally well. Consider the long-​standing idea that behavior
evolves first, before specialized morphological and physiological adaptations arise (7, 43). There
are many interesting examples of this phenomenon in currently living species such as marine

94
95

Neurogenetics of motivation

iguanas in the Galapagos which basically are the same as mainland iguanas except for their
behavior of swimming in the ocean (20).They have no special scales, webbed feet, or fins which
other marine reptiles possess for swimming, because they haven’t existed on earth long enough
in the marine environment for these morphological features to evolve. However, one organ that
probably has evolved slightly already is the iguana’s brain that leads them to want to enter the
marine environment and swim to obtain their food. This supports the “behavior evolves first”
idea that the evolution of motivational systems in the brain precedes morphological adaptations
in the periphery.
The “behavior evolves first” idea is also supported by direct experimental evidence. Over the
last 25 years, Theodore Garland’s group has conducted a replicated artificial selection experi-
ment for increased voluntary wheel running in house mice (88, 94). If you look at the timeline
of the discoveries, overall results suggest that the big changes initially were in neurological traits
related to motivation (77–​80), and then later changes in exercise physiological traits such as
muscle phenotypes (41), aerobic capacity (76), and bone symmetry (31) became more evident.
The interpretation is that in initial generations, variation in running was mostly attributed to
differences in motivation to run, i.e., the reason why one mouse ran 5 km per day and another
1 km per day was not because of any differences in physical capability between the mice
but rather because of differences in their intrinsic motivation for running. Several generations
of selection increased motivation to a local maximum during which time variation in exer-
cise physiological features became more and more significant contributors to the individual
variation. As a result, changes in exercise physiological traits became more apparent in later
generations. Taken together, these data strongly support the notion that motivation for physical
activity can evolve, and is an important component of individual and species-​level differences
in activity levels. These data further imply that if the goal is to find genetic factors involved in
the evolution of physical activity levels, then understanding how genes influence motivation to
engage in the behavior is of paramount importance.
The purpose of this chapter is to provide an updated review on what is known about gen-
etic variation underlying motivation for physical activity levels across individuals and species.
The material is relevant in the broad sense, for understanding how behavior evolves, but also
important in the narrower sense for human health. Arguably the single most important factor
that has contributed to the decline of health in Western society is our increasingly sedentary
lifestyle, which contributes to metabolic syndrome (28), obesity (65), diabetes (42), heart disease
(67), stroke (55), and cancer (63). Not only is physical activity crucial for maintaining physical
health, but evidence accumulated over the past 30 years has established that there are profound
benefits of exercise on mental health as well (21, 39). Exercise is a useful strategy for coping
with stress (37), treating depression (89), preventing relapse in drugs abuse (11), and delaying
cognitive aging (16). Randomized human clinical trials have established that walking quickly
for 40 minutes a day, three times a week for 6 months results in changes in brain functional
and anatomical measures which support the pro-​cognitive outcomes (24). An enormous litera-
ture in rodent models has established profound effects of exercise on brain anatomy, chemistry,
and physiology providing mechanistic support for pro-​cognitive outcomes (as reviewed in 18,
19, 97). One of the highlights of this literature is the landmark discovery that several weeks of
voluntary wheel running increases the formation of entire new neurons in the dentate gyrus, a
region of the brain crucial for learning and memory (96). Taken together, what this means for
medicine is that we already know how to prevent the bulk of our health problems in a pretty
broad sweeping way, and we usually have the physical capability to do it, yet we generally lack
the motivation.Therefore, understanding genetic pathways which increase motivation for phys-
ical activity could have broad implications for preventative medicine.

95
96

J.S. Rhodes

The natural reward circuit as a molding block for evolution of behavior

The natural reward circuit is defined as a collection of interconnected brain regions involved
in the perception of pleasure. In the 1950s, Olds and Milner conducted experiments with rats
placed in operant boxes in which a lever could be pressed by the rat on a voluntary basis that
would deliver mild electric shocks to various parts of the brain through differential placement
of the electrode (69). They discovered that the rats would learn to self-​administer shocks when
the electrode was placed in numerous places throughout the brain. However, when they placed
the electrode in certain regions, results were particularly striking. If the electrode was placed
anywhere along the medial forebrain bundle, the axonal projections from the ventral tegmental
area to the nucleus accumbens, or passing through the lateral hypothalamus, the rats would
display an extraordinary motivation for the shock to the point of pressing so often as to forgo
eating, drinking, and sleeping (Figure 8.1) (69). It is interesting to note from the original work
that rats self-​administered shocks to many different regions throughout the brain, suggesting the
existence of an extended reward circuit that includes most of the brain (64, 100). This is con-
sistent with the idea that a lot of real estate in the brain is involved in the perception of reward,
and that a fundamental function of the brain, the organ which shapes behavior, is reward pro-
cessing, in a manner that informs decision-​making.
If one extends the idea of the stimulating electrode to extrinsic stimuli which are naturally
capable of electrically activating the same circuit (albeit at a physiological level, and much less
intensely than the electrode), then one way to coerce animals to behave a certain way is to make
the behavior result in the electrical activation of the reward circuit and thereby perception of
reward. In other words, if genes regulated the development of neural circuits in the brain in such
a way that high levels of physical activity were perceived as rewarding because they activated
the reward circuit and served as a positive reinforcer, then individuals would choose to display
heightened activity. Therefore, the next logical question to ask is how physical activity stimulates

Prefrontal
cortex Hippocampus

Ventral
tegmental area
Nucleus
accumbens Dopamine neurons
Lateral
hypothalamus

Figure 8.1 The natural reward circuit. A simplified illustration of interactions between the hippocampus
and the natural reward circuit in the rodent brain. Forceful movement activates the hippocampus in
proportion to the intensity or effort. The ventral midbrain and the nucleus accumbens, core nodes of the
natural reward circuit, receive input from the hippocampus (32, 51). The highlighted areas indicate places
where the neurophysiology may differ between individuals highly motivated for activity as compared
to those less motivated. If the hippocampal input results in stronger activation of the ascending limb of
the reward circuit through dopamine innervation of the nucleus accumbens, and prefrontal cortex, and
subsequent activation of the lateral hypothalamus, physical activity would be reinforced. The precise
cellular and molecular signatures and underlying sequence variants responsible for motivating physical
activity as distinguished from other natural or artificial rewards (such as drugs of abuse) are not known
and are currently actively being investigated.

96
97

Neurogenetics of motivation

the natural reward circuit, and what the difference is between the reward circuit of an animal
that is unmotivated for physical activity versus an animal that is highly motivated for physical
activity and which seeks out physical activity despite its high energetic costs and generation
of fatigue. Here we come to a point where it starts getting complicated, because the old view
that it will be one molecule or even a few has proven false (23). The reality is more likely a
story about billions of neurons interacting with each other and with glial cells, with complex
chemical signaling pathways involving hundreds if not thousands of molecules. These signaling
pathways transmit information from one cell to the next, but there are numerous receptors
and neurochemicals implicated at any one synapse. Realistically, it is a combination of multiple
mixes that influence reward processing and that distinguish physical activity reward from other
natural rewards or drugs of abuse.

Beyond dopamine
Dopamine has historically been considered the substrate for reward, and there is abundant evi-
dence in the literature supporting this idea (91, 103). For example, dopamine levels increase in
extracellular spaces in the nucleus accumbens in response to inherently rewarding stimuli such
as sweet flavor (61), orgasms (72), and drugs of abuse (102). Moreover, the dopamine release
depends critically on the perceived rewarding value of the stimuli which can change depending
on the circumstances. In one study, rats were fed either a diet deficient or replete with salt, and
then administered a saline solution directly into the mouth and extracellular dopamine in the
nucleus accumbens was measured using carbon fiber voltammetry methods. This showed that
the saline solution elicited a dopamine response only when the rats were deprived of salt, not
when the diet contained sufficient salt. Moreover, the dopamine response was severely reduced
in the salt-​deprived rats if the salt taste receptors were pharmacologically blocked (17). In add-
ition, the same group in a different study found clear evidence, as others have, that a bitter solu-
tion that elicits an aversive taste reaction reduces dopamine release (62). Taken together these
data suggest that dopamine release in response to a stimulus is context dependent and matches
the perceived rewarding value of the stimulus, rather than a direct response to the perception of
the stimulus itself. Furthermore, these data constitute strong recent evidence in support of the
dopamine reward hypothesis, demonstrating that stimuli that are perceived as rewarding increase
dopamine in the nucleus accumbens (103).
On the other hand, serious questions have been raised about the validity of the hypothesis
that dopamine is a substrate for reward as opposed to a more general role for dopamine as a sali-
ence detector, or in learning (5, 87). First, dopamine-​deficient mice appear to find sweet flavor
rewarding, and display the normal preference for sweet solutions, suggesting dopamine is not
necessary for reward (14). More recent studies have established that dopamine-​deficient mice
can also perceive opioid reward (40). Second, stress and pain have also been found to increase
dopamine in the nucleus accumbens, suggesting dopamine release is not specific to reward (1,
48, 85). In light of the conflicting evidence, it has been suggested that dopamine is released into
the nucleus accumbens in response to an emotionally important event but does not distinguish
between aversive and rewarding qualities, at least at this level of biological organization (nucleus
accumbens dopamine as a whole) (85, 87). The large literature suggesting a role for dopamine
in regulating neuronal plasticity, and especially reinforcement learning, supports this idea (4, 73,
86, 90, 101).
In light of the fact that physical activity involves rewarding and aversive emotional reactions,
and that dopamine also functions in voluntary control of movement through the nigrostriatal
pathway (35, 58), it is not surprising that the dopamine signaling pathway has been implicated in

97
98

J.S. Rhodes

genetics of increased physical activity in several model systems now (10, 27, 53, 60, 79, 81, 83, 84,
99). However, the simple hypothesis that the difference would be in the amount of dopamine
released or in the number of receptors has turned out to be false. For example, in the replicated
selective breeding experiment for increased voluntary wheel running referenced previously, we
discovered that the high-​runner mice responded completely differently to drugs which block
the dopamine reuptake transporter protein (77, 78). Similarly in a selective breeding experiment
for increased and decreased voluntary wheel running in rats, the lines responded differently to a
dopamine reuptake transporter blocker (10). The main cellular mechanism of action attributed
to the psychoactive effects of the drugs which were used, cocaine, methylphenidate (Ritalin),
and GBR 12909, was thought to be through increased dopamine in extracellular spaces. These
drugs decreased wheel running in high-​runner mice in a dose-​dependent fashion whereas they
had no effect or the opposite effect, and increased wheel running, in control mice (77, 78). The
results demonstrated that some feature of the dopamine system had been altered by selection,
but follow-​up experiments found no differences in dopamine levels, or dopamine turnover in
the nucleus accumbens (unpublished observations). Dopamine levels appear to have decreased
in the dorsal raphe, the region of the brain containing the greatest density of serotonin neuron
cell bodies, and DOPAC, which is a metabolite of dopamine, appears to have been reduced in
the substantia nigra (98). Further, genes for two dopamine receptors displayed 20% increased
expression in the hippocampus of the high-​runner mice (9), but dopamine has diverse roles in
different brain regions and the direct implications of these results for reward processing remain
unclear. Additional experiments found no differences in dopamine receptors, or transporters in
the nucleus accumbens or frontal cortical brain regions (unpublished data) using radiolabeled
ligand binding assays established by Aaron Janowsky (45, 46). As described and characterized so
well by the Nobel Prize laureate, Paul Greengard, dopamine receptors are only at the tip of an
iceberg in terms of activing an entire intracellular signaling cascade involving many enzymes,
interacting proteins, and molecular complexes (33). Any one of these could be altered at the
genetic level to cause a change in the dopamine signal in response to a stimulus such as phys-
ical activity. Furthermore, dopamine neurons interact with neurons of many other types, so
changes in other neurotransmitter systems could influence dopamine function through these
interactions. Other neurotransmitter systems implicated in reward processing and motivation for
physical activity include serotonin (38, 88, 98), glutamate (34, 92), gamma aminobutyric acid
(GABA) (54), opioids (8, 71), and cannabinoids (22, 49, 50, 59, 74, 75, 93, 95). Each of these
systems comes with their own degree of complexity in terms of signaling molecules down-
stream of receptors. Some of these receptors can be on presynaptic terminals and postsynaptic
terminals to elicit a diverse array of effects on the reward circuit. Taken together, this paints a
complex picture, and one that is only meaningful if considered at a systems level. The dopa-
mine signaling system is clearly involved in the motivation for physical activity, but the specific
molecular underpinnings are only beginning to become realized, and the answer is not one
or two molecules, but the way in which systems-​level gene expression is coordinated through
development of a nervous system and alterations in chromatin structure in a cell type-​specific
manner (88).

Specificity of reward circuitry for promoting physical

activity remains elusive
Unfortunately, determining the reward circuitry that promotes physical activity is an area in
its infancy with little progress. Our understanding of how circuits in the brain are involved in
motivation for specific behaviors is still remarkably rudimentary. Only a few reports illustrate

98
99

Neurogenetics of motivation

specificity in the neural physiology underlying the perception of different types of rewards.
Regina Carelli has done work showing that specific neurons in the nucleus accumbens are
activated in response to sugar but are not activated in response to cocaine reward (13, 15).
Other studies, attempting to find differential patterns of brain activation associated with
expecting a food reward versus cocaine, have been only moderately successful in finding dif-
ferential patterns of brain activation (106), and even then it is difficult to know whether these
differences can be attributed to the magnitude of the perceived reward rather than the reward
type. Even finding differential patterns of brain activation in response to aversive stimuli versus
a reward is difficult as the majority of the brain activation response appears to reflect relative
emotional potency rather than valence (47). In other words, when a subject is presented with a
stimulus that is either very rewarding or very aversive, many of the same brain regions become
activated. One possible explanation is that the majority of the brain response is a reflection of
attention or arousal to the stimulus or learning about the stimulus so it can be approached or
avoided in the future, and the learning, attention, and arousal are consistent between the two
valences.
While some relevant regions and molecules have been identified for many different types
of rewards, the question of specificity remains a major issue. When specificity is explicitly
evaluated, often it is not found (29, 47, 68, 106). Therefore, how the different regions and
molecules that have been identified mix differently to create different versions of motiv-
ation, particularly when it is directed at one or another reward such as a drug or exercise, the
understanding becomes increasingly cloudy. The regions including the limbic system (amyg-
dala, hippocampus, frontal cortex), natural reward circuit (ventral tegmental area to nucleus
accumbens), and hypothalamus are activated for all sorts of natural and artificial rewards as
well as for aversive stimuli such as in response to pain and stress. This variety of activation
triggers suggest that in order to find specificity one will have to look at a more detailed
aspect of the physiology, such as examining specific neurons and their firing rates similar to
the work by Carelli mentioned briefly above. It appears that no one has tried to determine
whether neurons can be found that are selectively sensitive to physical activity reward over
other rewards.This area of investigation, finding the specificity of motivation for different types
of natural rewards, will be important before we can really understand the motivation for phys-
ical activity at a neurobiological level.
An important clue about how the reward circuit may be modified to specifically promote
one behavior over another comes not from studies exploring motivation for physical activity
but rather from studies exploring motivation for pair-​bonding behavior in voles. The leading
hypothesis is that the evolution of monogamy in voles involves genetic changes in the pro-
motor regions of genes which affect the way oxytocin and arginine vasopressin receptors are
distributed and expressed on different nodes of the reward circuit (105). So, the idea is that
the reward circuit itself remains unchanged (e.g., in terms of dopamine, opioids, cannabinoids,
medial forebrain bundle, etc.). However, whenever two voles come into contact with each other,
the peptides oxytocin in females and arginine vasopressin in males are released. The receptors
for these two peptides in the brain are situated in such a way that they cause a differential level
of electrical activation of the reward circuit and perception of reward. For example, monog-
amous female Prairie voles have more oxytocin receptors in the nucleus accumbens than pol-
ygynous female Montane voles (44, 82, 105). In males, the arginine vasopressin V1a receptor is
concentrated in regions of the reward circuit in Prairie voles as compared to Montane voles (57,
104, 105). Moreover, individual differences in affiliative behavior in Prairie voles are associated
with differential expression of the peptide receptors in the nodes of the natural reward circuit
in expected ways based on their behavior (3, 36). Incredibly, the genetic regulatory mechanism

99
100

J.S. Rhodes

for this differential spatial distribution and expression of the V1a receptors has been fairly well
worked out in the voles, and appears to involve tandem repeated DNA sequences in the pro-
motor region (57). Although the tandem repeat sequences cannot underlie variation in mon-
ogamy/​polygyny in Microtus voles more generally because many polygynous voles also have
the repeated sequences (26), strong, direct molecular evidence supports the hypothesis that at
least the Prairie vole variant of the promotor and resulting change in neuroanatomical distri-
bution and expression of arginine vasopressin V1a receptors facilitates pair bonding. Both mice
and meadow voles, which are polygynous species, can be coerced into displaying affiliative
behavior if the appropriate neurons are engineered to express the Prairie vole variant of the
V1a receptor (57, 104). Moreover, recent evidence demonstrates that this differential distribu-
tion of receptors may increase the synchronicity of electrical activation between two key nodes
of the reward circuit, the nucleus accumbens and prefrontal cortex, when the voles come in
contact with each other (2). The degree of future affiliative behavior is greatly predicted by a
measure of coherence between the electrical rhythms in each reward region during initial con-
tact. Taken together, the vole work suggests that the evolution of a voluntary behavior results
from changes in the way signaling systems involved in the behavior interact with the different
nodes of the reward circuit. More specifically, the circuit appears to have to function in such a
way that it facilitates coherence between different regions of the reward circuit at the time when
the animals are performing the behavior. To the best of our knowledge, no study has explored
coherence in electrical activation of the different nodes of the reward circuit during high levels
of physical activity. It is possible that the evolution of increased motivation for physical activity
required a change in the physiology of the circuit in such a way that coherence occurs between
prefrontal cortex and nucleus accumbens during exhaustive physical exercise.

Physical activity activates the hippocampus, but the functional

significance remains a mystery
Although no study has examined coherence between nodes of the reward circuit during phys-
ical activity as was performed in the vole affiliative study, several studies have examined neuronal
activation of different regions of the brain during physical activity. As it turns out, the region
most activated from voluntary wheel-​running behavior and forced treadmill running is the
hippocampus (80). This is somewhat unexpected because the hippocampus is more known for
its role in learning and memory than movement. However, a large literature has established that
the hippocampal formation is instantaneously engaged at the onset and duration of strenuous
physical movements, displaying a rhythmic synchronous activity of large numbers of neurons,
which is proportional to the speed (or force) of the movement and persists for as long as phys-
ical activity persists (6, 12, 56).
In addition to the literature establishing activation of the hippocampus in proportion to the
intensity of physical activity, the hippocampus has an extensive literature for its role in motiv-
ation, and it is often included as a peripheral reward region well interconnected with the central
nodes of the reward circuit such as the ventral midbrain, nucleus accumbens, and prefrontal
cortex (Figure 8.1) (70). It is possible that the physical activity-​related electrical activation of the
hippocampus stimulates the reward circuit to a greater degree in mice that are highly motivated
for physical activity, though this would require additional testing to confirm. Finding sequence
variants which affect the resulting development and function of hippocampus–​reward circuit
interactions could be a fruitful avenue for future research on the neurogenetics of motivation
for physical activity.

100
101

Neurogenetics of motivation

Genetic regulatory mechanisms for increasing motivation

for physical activity
It was thought that science would eventually find genetic differences in dopamine receptors
and transporters in cocaine addiction, opioid receptors for heroin addiction, GABA and glu-
tamate for alcoholism, and cannabinoid receptors for marijuana. However, results from human
genetics over the past 20 years suggest instead that the heritability of drug addiction is highly
overlapping with personality traits such as novelty seeking, and usually not specific to any one
type of reward (52). Furthermore, the contribution of single candidate genes such as a dopamine
receptor to a complex trait accounts for only a tiny fraction of the heritability, if any at all (66).
Therefore, it is exceedingly unlikely that an approach that examines single candidate genes such
as dopamine receptors will make headway into understanding the genetic mechanisms which
regulate motivation for physical activity, at least with any degree of specificity distinguishing it
from other rewards (25).
As described previously, genetic changes that lead to the development and function of a
reward circuit that reinforces physical activity is a more realistic mechanism (25, 81, 88). Systems
genetic analysis of gene expression changes in the striatum (both dorsal and ventral nucleus
accumbens, main dopamine innervation region) of the high-​runner mice implicated a network
of genes that affect chromatin and/​or transcriptional states (88). Chromatin can exist in multiple
functional states, either in a condensed state that prohibits gene expression or a relaxed state
that facilitates gene expression. Thus, sequence variants that affect chromatin structure could
explain gene expression patterns in crucial regions of the reward circuit which could affect
the way the cells respond to physical activity. Specifically, for the evolution of voluntary wheel
running a structural polymorphism in the SMARCA4 protein is suggested to cause differential
interaction with the BAZ1A protein, altering the distribution of the H1F0 histone subunit pro-
tein and thereby creating a chromatin state conducive for increased expression of the serotonin
receptor and an orphan G protein-​coupled receptor. Future work is needed to rigorously test
this hypothesis by assessing the chromatin state through direct measurement.

Conclusions
In conclusion, heritable variation in motivation for physical activity is hypothesized to result
from differences in the way neural signals involved in regulating the intensity of the physical
activity interact with the natural reward circuit. Therefore, genetics of motivation for increased
physical activity likely involve genetic programs for building a nervous system in which the
act of intense physical activity produces a strong reward signal that leads to reinforcement of
the behavior, and subsequent drive for the activity. We propose that a crucial location for the
occurrence of neurological differences is at the interface between the hippocampus and various
nodes of the reward circuit (Figure 8.1, shaded regions). This is because the hippocampus
becomes electrically activated like no other region in the brain in proportion to exercise inten-
sity. If the brain were wired in such a way that physical activity-​induced neuronal activation
stimulated the reward circuit, that would lead to reinforcement. However, the precise molecules
and developmental trajectories that result in a brain predisposed for high or low motivation for
physical activity are far from known. Models such as long-​term selective breeding experiments
for increased voluntary wheel running in mice and rats provide promising avenues for research
and discovery in this area. Finding neurogenetic pathways that increase motivation for physical
activity has broad implications for evolution of behavior. It also has important applications in
preventative medicine. The health benefits of exercise are profound on all dimensions. Getting

101
102

J.S. Rhodes

people to exercise is the hard part. Understanding the pathways which increase intrinsic drive
for physical activity therefore holds the key to prolonged health and prosperity in the 21st
century.

References
1. Abercrombie ED, Keefe KA, DiFrischia DS, and Zigmond MJ. Differential effect of stress on
in vivo dopamine release in striatum, nucleus accumbens, and medial frontal cortex. Journal of neuro-
chemistry 52: 1655–​1658, 1989.
2. Amadei EA, Johnson ZV, Kwon YJ, Shpiner AC, Saravanan V, Mays WD, Ryan SJ, Walum
H, Rainnie DG, and Young LJ. Dynamic corticostriatal activity biases social bonding in monog-
amous female prairie voles. Nature 546: 297, 2017.
3. Barrett CE, Keebaugh AC, Ahern TH, Bass CE, Terwilliger EF, and Young LJ. Variation in
vasopressin receptor (Avpr1a) expression creates diversity in behaviors related to monogamy in prairie
voles. Hormones and behavior 63: 518–​526, 2013.
4. Beeler JA and Mourra D. To do or not to do: Dopamine, affordability and the economics of oppor-
tunity. Frontiers in integrative neuroscience 12: 6, 2018.
5. Berridge KC and Robinson TE. What is the role of dopamine in reward: Hedonic impact, reward
learning, or incentive salience? Brain research reviews 28: 309–​369, 1998.
6. Bland BH and Oddie SD. Theta band oscillation and synchrony in the hippocampal formation
and associated structures: the case for its role in sensorimotor integration. Behavioural brain research
127: 119–​136, 2001.
7. Blomberg SP, Garland Jr T, and Ives AR. Testing for phylogenetic signal in comparative
data: Behavioral traits are more labile. Evolution 57: 717–​745, 2003.
8. Boecker H, Sprenger T, Spilker ME, Henriksen G, Koppenhoefer M, Wagner KJ, Valet M,
Berthele A, and Tolle TR. The runner’s high: Opioidergic mechanisms in the human brain. Cerebral
cortex 18: 2523–​2531, 2008.
9. Bronikowski A, Rhodes J, Garland Jr T, Prolla T, Awad T, and Gammie S. The evolution of
gene expression in mouse hippocampus in response to selective breeding for increased locomotor
activity. Evolution 58: 2079–​2086, 2004.
10. Brown JD, Green CL, Arthur IM, Booth FW, and Miller DK. Cocaine-​induced locomotor
activity in rats selectively bred for low and high voluntary running behavior. Psychopharmacology
232: 673–​681, 2015.
11. Brown RA, Abrantes AM, Read JP, Marcus BH, Jakicic J, Strong DR, Oakley JR, Ramsey
SE, Kahler CW, and Stuart GL. A pilot study of aerobic exercise as an adjunctive treatment for
drug dependence. Mental health and physical activity 3: 27–​34, 2010.
12. Buzsaki G. Theta oscillations in the hippocampus. Neuron 33: 325–​340, 2002.
13. Cameron CM, Wightman RM, and Carelli RM. Dynamics of rapid dopamine release in the nucleus
accumbens during goal-​directed behaviors for cocaine versus natural rewards. Neuropharmacology
86: 319–​328, 2014.
14. Cannon CM and Palmiter RD. Reward without dopamine. Journal of neuroscience
23: 10827–​10831, 2003.
15. Carelli RM, Ijames SG, and Crumling AJ. Evidence that separate neural circuits in the
nucleus accumbens encode cocaine versus “natural” (water and food) reward. Journal of neuroscience
20: 4255–​4266, 2000.
16. Colcombe S and Kramer AF. Fitness effects on the cognitive function of older adults: A meta-​
analytic study. Psychological science 14: 125–​130, 2003.
17. Cone JJ, Fortin SM, McHenry JA, Stuber GD, McCutcheon JE, and Roitman MF. Physiological
state gates acquisition and expression of mesolimbic reward prediction signals. Proceedings of the
National Academy of Sciences 113: 1943–​1948, 2016.
18. Cotman CW and Berchtold NC. Exercise: A behavioral intervention to enhance brain health and
plasticity. Trends in neurosciences 25: 295–​301, 2002.
19. Cotman CW, Berchtold NC, and Christie L-​A. Exercise builds brain health: Key roles of growth
factor cascades and inflammation. Trends in neurosciences 30: 464–​472, 2007.
20. Dawson WR, Bartholomew GA, and Bennett AF. A reappraisal of the aquatic specializations of
the Galapagos marine iguana (Amblyrhynchus cristatus). Evolution 31: 891–​897, 1977.

102
103

Neurogenetics of motivation

21. Dishman RK, Berthoud HR, Booth FW, Cotman CW, Edgerton VR, Fleshner MR,
Gandevia SC, Gomez-​Pinilla F, Greenwood BN, and Hillman CH. Neurobiology of exercise.
Obesity 14: 345–​356, 2006.
22. Dubreucq S, Koehl M, Abrous DN, Marsicano G, and Chaouloff F. CB1 receptor deficiency
decreases wheel-​running activity: Consequences on emotional behaviours and hippocampal neuro-
genesis. Experimental neurology 224: 106–​113, 2010.
23. Eichler EE, Flint J, Gibson G, Kong A, Leal SM, Moore JH, and Nadeau JH. Missing her-
itability and strategies for finding the underlying causes of complex disease. Nature reviews genetics
11: 446, 2010.
24. Erickson KI, Voss MW, Prakash RS, Basak C, Szabo A, Chaddock L, Kim JS, Heo S, Alves
H, and White SM. Exercise training increases size of hippocampus and improves memory. Proceedings
of the National Academy of Sciences 108: 3017–​3022, 2011.
25. Ferguson DP, Dangott LJ, Vellers HL, Schmitt EE, and Lightfoot JT. Differential pro-
tein expression in the nucleus accumbens of high and low active mice. Behavioural brain research
291: 283–​288, 2015.
26. Fink S, Excoffier L, and Heckel G. Mammalian monogamy is not controlled by a single gene.
Proceedings of the National Academy of Sciences 103: 10956–​10960, 2006.
27. Foley TE, Greenwood BN, Day HE, Koch LG, Britton SL, and Fleshner M. Elevated cen-
tral monoamine receptor mRNA in rats bred for high endurance capacity: Implications for central
fatigue. Behavioural brain research 174: 132–​142, 2006.
28. Frugé AD, Byrd SH, Fountain BJ, Cossman JS, Schilling MW, and Gerard P. Increased phys-
ical activity may be more protective for metabolic syndrome than reduced caloric intake. An analysis
of estimated energy balance in US adults: 2007–​2010 NHANES. Nutrition, metabolism and cardiovas-
cular diseases 25: 535–​540, 2015.
29. Garavan H, Pankiewicz J, Bloom A, Cho J-​K, Sperry L, Ross TJ, Salmeron BJ, Risinger R,
Kelley D, and Stein EA. Cue-​induced cocaine craving: neuroanatomical specificity for drug users
and drug stimuli. American journal of psychiatry 157: 1789–​1798, 2000.
30. Garland Jr T and Carter P. Evolutionary physiology. Annual review of physiology 56: 579–​621, 1994.
31. Garland Jr T and Freeman PW. Selective breeding for high endurance running increases hindlimb
symmetry. Evolution 59: 1851–​1854, 2005.
32. Gasbarri A, Packard MG, Campana E, and Pacitti C. Anterograde and retrograde tracing of
projections from the ventral tegmental area to the hippocampal formation in the rat. Brain research
bulletin 33: 445–​452, 1994.
33. Greengard P, Allen PB, and Nairn AC. Beyond the dopamine receptor. Neuron 23: 435–​447, 1999.
34. Grigsby KB, Kovarik CM, Rottinghaus GE, and Booth FW. High and low nightly running
behavior associates with nucleus accumbens N-​Methyl-​d-​aspartate receptor (NMDAR) NR1 sub-
unit expression and NMDAR functional differences. Neuroscience letters 671: 50–​55, 2018.
35. Groves PM. A theory of the functional organization of the neostriatum and the neostriatal control
of voluntary movement. Brain research reviews 5: 109–​132, 1983.
36. Hammock EA and Young LJ. Microsatellite instability generates diversity in brain and
sociobehavioral traits. Science 308: 1630–​1634, 2005.
37. Harris AH, Cronkite R, and Moos R. Physical activity, exercise coping, and depression in a 10-​year
cohort study of depressed patients. Journal of affective disorders 93: 79–​85, 2006.
38. Hayes DJ and Greenshaw AJ. 5-​HT receptors and reward-​related behaviour: a review. Neuroscience
& biobehavioral reviews 35: 1419–​1449, 2011.
39. Hillman CH, Erickson KI, and Kramer AF. Be smart, exercise your heart: exercise effects on brain
and cognition. Nature reviews neuroscience 9: 58, 2008.
40. Hnasko TS, Sotak BN, and Palmiter RD. Morphine reward in dopamine-​deficient mice. Nature
438: 854, 2005.
41. Houle-​Leroy P, Guderley H, Swallow JG, and Garland Jr T. Artificial selection for high activity
favors mighty mini-​muscles in house mice. American journal of physiology –​regulatory, integrative and
comparative physiology 284: R433–​R443, 2003.
42. Hu FB, Sigal RJ, Rich-​Edwards JW, Colditz GA, Solomon CG, Willett WC, Speizer FE,
and Manson JE. Walking compared with vigorous physical activity and risk of type 2 diabetes in
women: a prospective study. JAMA 282: 1433–​1439, 1999.
43. Huey RB, Hertz PE, and Sinervo B. Behavioral drive versus behavioral inertia in evolution: A null
model approach. The American naturalist 161: 357–​366, 2003.

103
104

J.S. Rhodes

44. Insel TR and Shapiro LE. Oxytocin receptor distribution reflects social organization in monog-
amous and polygamous voles. Proceedings of the National Academy of Sciences 89: 5981–​5985, 1992.
45. Janowsky A, Mah C, Johnson RA, Cunningham CL, Phillips TJ, Crabbe JC, Eshleman AJ,
and Belknap JK. Mapping genes that regulate density of dopamine transporters and correlated
behaviors in recombinant inbred mice. Journal of pharmacology and experimental therapeutics
298: 634–​643, 2001.
46. Janowsky A, Neve KA, Kinzie JM, Taylor B, de Paulis T, and Belknap JK. Extrastriatal dopa-
mine D2 receptors: distribution, pharmacological characterization and region-​specific regulation by
clozapine. Journal of pharmacology and experimental therapeutics 261: 1282–​1290, 1992.
47. Johnson ZV, Revis AA, Burdick MA, and Rhodes JS. A similar pattern of neuronal Fos activa-
tion in 10 brain regions following exposure to reward-​or aversion-​associated contextual cues in mice.
Physiology & behavior 99: 412–​418, 2010.
48. Kalivas PW and Duffy P. Selective activation of dopamine transmission in the shell of the nucleus
accumbens by stress. Brain research 675: 325–​328, 1995.
49. Keeney BK, Meek TH, Middleton KM, Holness LF, and Garland Jr T. Sex differences in can-
nabinoid receptor-​1 (CB1) pharmacology in mice selectively bred for high voluntary wheel-​running
behavior. Pharmacology biochemistry and behavior 101: 528–​537, 2012.
50. Keeney BK,Raichlen DA,Meek TH,Wijeratne RS,Middleton KM,Gerdeman GL,and Garland
Jr T. Differential response to a selective cannabinoid receptor antagonist (SR141716: rimonabant)
in female mice from lines selectively bred for high voluntary wheel-​running behaviour. Behavioural
pharmacology 19: 812–​820, 2008.
51. Kelley A and Domesick V. The distribution of the projection from the hippocampal formation
to the nucleus accumbens in the rat: An anterograde and retrograde-​horseradish peroxidase study.
Neuroscience 7: 2321–​2335, 1982.
52. Kendler KS, Jacobson KC, Prescott CA, and Neale MC. Specificity of genetic and environmental
risk factors for use and abuse/​dependence of cannabis, cocaine, hallucinogens, sedatives, stimulants,
and opiates in male twins. American Journal of Psychiatry 160: 687–​695, 2003.
53. Knab AM and Lightfoot JT. Does the difference between physically active and couch potato lie in
the dopamine system? International journal of biological sciences 6: 133, 2010.
54. Koob GF. Drugs of abuse: Anatomy, pharmacology and function of reward pathways. Trends in
pharmacological sciences 13: 177–​184, 1992.
55. Kyu HH, Bachman VF, Alexander LT, Mumford JE, Afshin A, Estep K, Veerman JL,
Delwiche K, Iannarone ML, and Moyer ML. Physical activity and risk of breast cancer, colon
cancer, diabetes, ischemic heart disease, and ischemic stroke events: Systematic review and dose-​
response meta-​analysis for the Global Burden of Disease Study 2013. BMJ 354: i3857, 2016.
56. Li JY, Kuo TB, Hsieh IT, and Yang CC. Changes in hippocampal theta rhythm and their
correlations with speed during different phases of voluntary wheel running in rats. Neuroscience
213: 54–​61, 2012.
57. Lim MM, Wang Z, Olazábal DE, Ren X, Terwilliger EF, and Young LJ. Enhanced partner pref-
erence in a promiscuous species by manipulating the expression of a single gene. Nature 429: 754, 2004.
58. Lotharius J and Brundin P. Pathogenesis of Parkinson’s disease: Dopamine, vesicles and α-​synuclein.
Nature reviews neuroscience 3: 932, 2002.
59. Lupica CR, Riegel AC, and Hoffman AF. Marijuana and cannabinoid regulation of brain reward
circuits. British journal of pharmacology 143: 227–​234, 2004.
60. Mathes WF, Nehrenberg DL, Gordon R, Hua K, Garland Jr T, and Pomp D. Dopaminergic
dysregulation in mice selectively bred for excessive exercise or obesity. Behavioural brain research
210: 155–​163, 2010.
61. McCutcheon JE, Beeler JA, and Roitman MF. Sucrose-​predictive cues evoke greater phasic dopa-
mine release than saccharin-​predictive cues. Synapse 66: 346–​351, 2012.
62. Mccutcheon JE, Ebner SR, Loriaux AL, and Roitman MF. Encoding of aversion by dopamine
and the nucleus accumbens. Frontiers in neuroscience 6: 137, 2012.
63. Michaud DS, Giovannucci E, Willett WC, Colditz GA, Stampfer MJ, and Fuchs CS. Physical
activity, obesity, height, and the risk of pancreatic cancer. JAMA 286: 921–​929, 2001.
64. Milner PM. Brain-​stimulation reward: a review. Canadian journal of psychology/​Revue canadienne de
psychologie 45: 1, 1991.
65. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, and Marks JS. Prevalence
of obesity, diabetes, and obesity-​related health risk factors, 2001. JAMA 289: 76–​79, 2003.

104
105

Neurogenetics of motivation

66. Munafo M, Matheson I, and Flint J. Association of the DRD2 gene Taq1A polymorphism and
alcoholism: A meta-​analysis of case–​control studies and evidence of publication bias. Molecular psych-
iatry 12: 454, 2007.
67. Myers J, McAuley P, Lavie CJ, Despres J-​P, Arena R, and Kokkinos P. Physical activity and
cardiorespiratory fitness as major markers of cardiovascular risk: Their independent and interwoven
importance to health status. Progress in cardiovascular diseases 57: 306–​314, 2015.
68. Noori HR, Linan AC, and Spanagel R. Largely overlapping neuronal substrates of reactivity to drug,
gambling, food and sexual cues: A comprehensive meta-​analysis. European neuropsychopharmacology
26: 1419–​1430, 2016.
69. Olds J and Milner P. Positive reinforcement produced by electrical stimulation of septal area and
other regions of rat brain. Journal of comparative and physiological psychology 47: 419, 1954.
70. Otmakhova N, Duzel E, Deutch AY, and Lisman J. The hippocampal-​VTA loop: the role of nov-
elty and motivation in controlling the entry of information into long-​term memory. In: Intrinsically
motivated learning in natural and artificial systems. New York: Springer, 2013, p. 235–​254.
71. Peciña S. Opioid reward “liking” and “wanting” in the nucleus accumbens. Physiology & behavior
94: 675–​680, 2008.
72. Pfaus J, Damsma G, Nomikos GG, Wenkstern D, Blaha C, Phillips A, and Fibiger H. Sexual
behavior enhances central dopamine transmission in the male rat. Brain research 530: 345–​348, 1990.
73. Pignatelli M and Bonci A. Role of dopamine neurons in reward and aversion: A synaptic plasticity
perspective. Neuron 86: 1145–​1157, 2015.
74. Raichlen DA, Foster AD, Gerdeman GL, Seillier A, and Giuffrida A. Wired to run: Exercise-​
induced endocannabinoid signaling in humans and cursorial mammals with implications for the
“runner’s high”. Journal of experimental biology 215: 1331–​1336, 2012.
75. Raichlen DA, Foster AD, Seillier A, Giuffrida A, and Gerdeman GL. Exercise-​induced
endocannabinoid signaling is modulated by intensity. European journal of applied physiology
113: 869–​875, 2013.
76. Rezende EL, Garland T, Chappell MA, Malisch JL, and Gomes FR. Maximum aerobic per-
formance in lines of Mus selected for high wheel-​running activity: Effects of selection, oxygen avail-
ability and the mini-​muscle phenotype. Journal of experimental biology 209: 115–​127, 2006.
77. Rhodes J and Garland T. Differential sensitivity to acute administration of Ritalin, apormorphine,
SCH 23390, but not raclopride in mice selectively bred for hyperactive wheel-​running behavior.
Psychopharmacology 167: 242–​250, 2003.
78. Rhodes J, Hosack G, Girard I, Kelley A, Mitchell G, and Garland T. Differential sensitivity to
acute administration of cocaine, GBR 12909, and fluoxetine in mice selectively bred for hyperactive
wheel-​running behavior. Psychopharmacology 158: 120–​131, 2001.
79. Rhodes JS, Gammie SC, and Garland Jr T. Neurobiology of mice selected for high voluntary
wheel-​running activity. Integrative and comparative biology 45: 438–​455, 2005.
80. Rhodes JS, Garland Jr T, and Gammie SC. Patterns of brain activity associated with variation in
voluntary wheel-​running behavior. Behavioral neuroscience 117: 1243, 2003.
81. Roberts MD, Gilpin L, Parker KE, Childs TE, Will MJ, and Booth FW. Dopamine D1 receptor
modulation in nucleus accumbens lowers voluntary wheel running in rats bred to run high distances.
Physiology & behavior 105: 661–​668, 2012.
82. Ross HE, Freeman SM, Spiegel LL, Ren X, Terwilliger EF, and Young LJ. Variation in oxytocin
receptor density in the nucleus accumbens has differential effects on affiliative behaviors in monog-
amous and polygamous voles. Journal of neuroscience 29: 1312–​1318, 2009.
83. Ruegsegger GN and Booth FW. Running from disease: molecular mechanisms associating dopa-
mine and leptin signaling in the brain with physical inactivity, obesity, and type 2 diabetes. Frontiers in
endocrinology 8: 109, 2017.
84. Ruegsegger GN, Brown JD, Kovarik MC, Miller DK, and Booth FW. Mu-​opioid receptor
inhibition decreases voluntary wheel running in a dopamine-​dependent manner in rats bred for high
voluntary running. Neuroscience 339: 525–​537, 2016.
85. Salamone JD. The involvement of nucleus accumbens dopamine in appetitive and aversive motiv-
ation. Behavioural brain research 61: 117–​133, 1994.
86. Salamone JD and Correa M. The mysterious motivational functions of mesolimbic dopamine.
Neuron 76: 470–​485, 2012.
87. Salamone JD, Correa M, Mingote SM, and Weber SM. Beyond the reward hypothesis: alterna-
tive functions of nucleus accumbens dopamine. Current opinion in pharmacology 5: 34–​41, 2005.

105
106

J.S. Rhodes

88. Saul M, Majdak P, Perez S, Reilly M, Garland T, and Rhodes JS. High motivation for exercise is
associated with altered chromatin regulators of monoamine receptor gene expression in the striatum
of selectively bred mice. Genes, brain and behavior 16: 328–​341, 2017.
89. Schuch FB, Vancampfort D, Richards J, Rosenbaum S, Ward PB, and Stubbs B. Exercise as
a treatment for depression: a meta-​analysis adjusting for publication bias. Journal of psychiatric research
77: 42–​51, 2016.
90. Schultz W. Dopamine reward prediction-​error signalling: a two-​component response. Nature reviews
neuroscience 17: 183, 2016.
91. Schultz W, Dayan P, and Montague PR. A neural substrate of prediction and reward. Science
275: 1593–​1599, 1997.
92. Smith-​Roe SL and Kelley AE. Coincident activation of NMDA and dopamine D1Receptors
within the nucleus accumbens core is required for appetitive instrumental learning. Journal of neurosci-
ence 20: 7737–​7742, 2000.
93. Sparling P, Giuffrida A, Piomelli D, Rosskopf L, and Dietrich A. Exercise activates the
endocannabinoid system. Neuroreport 14: 2209–​2211, 2003.
94. Swallow JG, Carter PA, and Garland T. Artificial selection for increased wheel-​running behavior
in house mice. Behavior genetics 28: 227–​237, 1998.
95. Thompson Z, Argueta D, Garland Jr T, and DiPatrizio N. Circulating levels of endocannabinoids
respond acutely to voluntary exercise, are altered in mice selectively bred for high voluntary wheel
running, and differ between the sexes. Physiology & behavior 170: 141–​150, 2017.
96. Van Praag H, Kempermann G, and Gage FH. Running increases cell proliferation and neurogen-
esis in the adult mouse dentate gyrus. Nature neuroscience 2: 266, 1999.
97. Voss MW, Nagamatsu LS, Liu-​Ambrose T, and Kramer AF. Exercise, brain, and cognition across
the life span. Journal of applied physiology 111: 1505–​1513, 2011.
98. Waters RP, Pringle R, Forster G, Renner K, Malisch J, Garland Jr T, and Swallow J. Selection
for increased voluntary wheel-​running affects behavior and brain monoamines in mice. Brain research
1508: 9–​22, 2013.
99. Waters RP, Renner K, Pringle R, Summers CH, Britton S, Koch L, and Swallow J. Selection
for aerobic capacity affects corticosterone, monoamines and wheel-​running activity. Physiology &
behavior 93: 1044–​1054, 2008.
100. Wise RA. Addictive drugs and brain stimulation reward. Annual review of neuroscience 19: 319–​340, 1996.
101. Wise RA. Dopamine, learning and motivation. Nature reviews neuroscience 5: 483, 2004.
102. Wise RA. Neurobiology of addiction. Current opinion in neurobiology 6: 243–​251, 1996.
103. Wise RA and Rompre P-​P. Brain dopamine and reward. Annual review of psychology 40: 191–​225, 1989.
104. Young LJ, Nilsen R, Waymire KG, MacGregor GR, and Insel TR. Increased affiliative response
to vasopressin in mice expressing the V 1a receptor from a monogamous vole. Nature 400: 766, 1999.
105. Young LJ and Wang Z. The neurobiology of pair bonding. Nature neuroscience 7: 1048, 2004.
106. Zombeck JA, Chen G-​T, Johnson ZV, Rosenberg DM, Craig AB, and Rhodes JS.
Neuroanatomical specificity of conditioned responses to cocaine versus food in mice. Physiology &
behavior 93: 637–​650, 2008.

106
107

9
PERIPHERAL MECHANISMS
ARISING FROM GENETICS
THAT REGULATE ACTIVITY
David P. Ferguson

Introduction
Behavior traits, such as physical activity, are generally accepted to be a result of environmental
and genetic/​biological factors. A large amount of literature (7, 18, 19, 38, 39, 43, 45, 47, 50,
52–​56) has examined the environmental influence on physical activity. While the literature has
focused on factors such as culture, peer influence, and the “built environment” (i.e., access to
sidewalks) (5, 54, 56), it has been shown that the main factors contributing to the control of
voluntary physical activity are genetic and biological influences (27).Voluntary physical activity
is an important trait to study because it has been positively correlated with decreases in cardio-
vascular disease, obesity, type 2 diabetes, and some types of cancer (27).With only 3.5% of adults
meeting the recommended physical activity guidelines (51), physical inactivity is the second
actual leading cause of death (~250,000 cases/​year) in the US with an estimated $507 billion
a year in healthcare costs (34). Thus, the identification of biological mechanisms that regulate
voluntary physical activity could improve the quality of life of individuals and potentially reduce
healthcare costs.
As described in Chapter 5, the use of animal models is a reliable approach to examine gen-
etic mechanisms of physical activity and has several benefits over the human model. Specifically,
mice can be bred to limit heterozygosity of the population, environmental factors can be con-
trolled by the investigator, and the short lifespan of the animal provides the opportunity to
measure physical activity across multiple generations (27). Inbred mice are produced using at
least 20 consecutive generations of parent × offspring or sister × brother mating, allowing for
mice to be genetically homozygous at all loci (30). Thus, the use of inbred mice allows for the
rigorous control of heterozygosity of the genome as well as control of the environment allowing
for investigation into the genetic regulation of physical activity. Furthermore, the mouse is
an excellent model of the human genome given that there is approximately 75% homology
between the mouse and human genome (8). While there is some discrepancy in homology
between mice and humans, this difference appears to be a result of differences in noncoding
regions of the genome (8).Thus, the demonstrated human–​mouse genomic homology increases
the probability that results from genetic studies using the mouse model can be directly translated
to humans.

107
108

D.P. Ferguson

To evaluate voluntary physical activity in the mouse model, several methods have been
utilized. The most frequently used methods in the literature have been wheel running, home
cage activity, and/​or maze activity. However, it has been suggested that studies using home
cage activity and/​or maze activity evaluate both physical activity level and fear/​anxiety level
(26). Therefore, in the mouse model, the method most often used to assess voluntary physical
activity level is wheel running (31).Wheel running as a measure of physical activity has a similar
physiological response as does partaking in physical activity by humans, such as self-​selection
of intensity, heart rate response, and neurological changes. Specifically it has been shown that
humans self-​select an average running intensity of 70% (12). Interestingly it has been shown that
mice self-​select a wheel running intensity of 65–​70% (28).The self-​selection of physical activity
level results in similar cardiovascular and neurological adaptations between mice and humans.
Adlam et al. (2) showed that during acute wheel running there was a significant increase in heart
rate and blood pressure, with 5 weeks of wheel running resulting in a decreased resting heart
rate and blood pressure (1). Similarly, it has been well established that regular bouts of moderate
intensity exercise (60–​70% of maximum heart rate) in humans result in a decrease of resting
heart rate and systolic blood pressure over time (36). Additionally, mouse wheel running and
human physical activity engagement result in similar neurological responses. Dishman et al. (10)
characterized the norepinephrine, 5-​hydroxytryptamine, dopamine, and gamma aminobutyric
acid responses to wheel running and showed that there was an increase in norepinephrine and
dopamine following mouse wheel running which led the authors to conclude that physical
activity had a protective effect on brain monoamine depletion and thus an antidepressant affect
(10). The human literature has documented similar findings of antidepressant effects following
bouts of physical activity (11). However, due to the nature of the study, brain monoamines could
not be directly measured in humans (10) for comparison to the mouse model. Given the simi-
larities between human and mouse physiological parameters during voluntary exercise and vol-
untary wheel running, the use of the mouse wheel running model appears to be appropriate to
determine the regulation of physical activity level and further allows the use of genetic linkage,
transcriptome, and proteome analyses.
The physiological site of activity regulation is an evolving question, with current litera-
ture suggesting that multiple genes and environmental factors influence one’s physical activity
level (29). It has been proposed by Kelly et al. that the genetic control of voluntary physical
activity is a result of both central drive and peripheral capability (22). Central drive is the result
of mechanisms in the brain that influence the “want” to be active while the “capability” to be
active is a result of peripheral factors that reduce pain/​fatigue allowing for more sustained phys-
ical activity level (35). While Chapter 8 focuses on potential central pathways that influence
physical activity, the remainder of this chapter focuses on peripheral factors that can be genet-
ically controlled to regulate physical activity.

Peripheral factors affecting the capability to be physically activity

There is limited literature addressing the role of potential peripheral “capability to be active”
factors in regulating physical activity. The available studies in this area have focused on skeletal
muscle with potential peripheral factors that may be associated with physical activity regulation
being components of substrate utilization and skeletal muscle contraction. Both these factors
influence fatigue resistance, which interestingly is also affected by the dopaminergic systems
and/​or endocannabinoid systems (located centrally) (24, 25). Given that both the dopaminergic
and endocannabinoid systems are involved with pain perception and since pain associated with
activity has been correlated to skeletal muscle fatigue (16), it is possible that the fatigability

108
109

Peripheral mechanisms regulating activity

Substrate Utilization
Glucose Transport to working Tissues Fatty Acid Availability
(Tsao et al.) (Meek et al.)

Skeletal Muscle Rigidity Physical Activity Engagement

(Ferguson et al.)

Skeletal Muscle Fatigue Resistance

IL-15Rα and SERCA2 Calsquestrin 1 Annexin VI

(Pescatello and Pistilli et al.) (Ferguson et al.and Pistilli et al.) (Ferguson et al.)

Calcium Homeostasis

Figure 9.1 Schematic indicating the key processes involved with the peripheral regulation of physical
activity along with associated journal citations. Specifically, substrate utilization and calcium homeostasis
are the primary determinants to physical activity engagement. An increase in glucose transport and
increasing fatty acid ability will increase physical activity. Optimal calcium homeostasis is achieved by an
increase in annexin VI, calsequestrin 1, and the expression of the IL-​15αAA and the associated increase
in SERCA2. These calcium regulatory factors influence skeletal muscle fatigue resistance which will
regulate skeletal muscle fatigue resistance. Lastly, an increase in proteins associated with skeletal muscle
rigidity will decrease physical activity engagement.

of skeletal muscle –​a peripheral factor influencing capability –​could also influence physical
activity level. Regardless of where these effects are exerted, there are at least two pathways
through which peripheral factors may influence physical activity (Figure 9.1).
Inherent mechanisms by which skeletal muscle can regulate physical activity may include
mechanisms associated with metabolism, contraction, and calcium flux (Figure 9.1). Specifically,
the ability to transport glucose and fatty acids to the working tissue may increase physical
activity levels. Furthermore, improved calcium homeostasis via IL-​ 15Rα, annexin VI, and
calsequestrin 1 may reduce skeletal muscle fatigue allowing for increased physical activity. Lastly,
it should be noted that there is a unique set of proteins that increase skeletal muscle rigidity
and their overexpression –​as seen in low-​active animals (14) –​that may lead to a decrease in
physical activity levels.

Substrate utilization may influence physical activity

Tsao et al. showed that an overexpression of glucose transport 4 (Glut4) in skeletal muscle led
to a fourfold increase in wheel running as compared to control animals (57). The methodology
used by Tsao et al. utilized an overexpression of Glut4 in muscle and heart tissue of the 129/​
SV mouse strain (48) by splicing the MLCl promoter upstream of the Glut4 gene. The authors
observed a 45% increase in food intake and a fourfold increase in wheel running as compared
to the control mice (~1 km ran for control mice and ~4 km ran for Glut4 overexpressed mice)
(57). The authors hypothesized that glucose transport was the rate-​limiting step in skeletal
muscle glucose uptake and therefore, an increase in glucose transporters could increase glucose
uptake allowing for an increased physical activity response (57).
In addition to Glut4, there have been several studies that have shown nutritional factors may
alter physical activity through alteration of substrate utilization. Researchers from the Garland

109
110

D.P. Ferguson

lab have shown that mice bred for high activity for over 60 generations –​mice whose activity
had plateaued at approximately generation 25 –​when treated with a high-​fat diet, signifi-
cantly increased their daily wheel running by 160% (33). To date, this is the only treatment that
Garland’s research group has used that has increased the physical activity level of their high-​
active mice (32).Therefore, the authors suggested that treatment with a high-​fat diet allowed for
increased fat utilization as a primary fuel source, allowing for an increased capability resulting
in a higher physical activity level in these mice whose activity had previously plateaued (33).

Skeletal muscle force of contraction and muscle fatigue

may influence physical activity
During muscle contraction, the axon terminal of a motor neuron releases a neurotransmitter
into the synaptic cleft. The neurotransmitters bind to receptors on the sarcolemma, which
transmit an action potential to the transverse tubule (T tubule) of the sarcolemma. Once the
action potential reaches the T tubule, there is a release of calcium (Ca2+) from the sarcoplasmic
reticulum (SR). Ca2+ binds to troponin C which causes a conformational change in tropomy-
osin, thereby exposing actin binding sites allowing for binding of the myosin head (41) and sub-
sequent skeletal muscle contraction (41). Taking a broad view of skeletal muscle contraction, it
is possible that biological factors that influence force production or fatigue at the motor neuron
or fiber level could influence physical activity level. In fact, in 2015 Pistilli et al. showed that by
knocking out IL-​15Rα there was a shift in skeletal muscle fibers (with no change in the motor
unit) to a fatigue-​resistant phenotype, which also led to a 6.3-​fold increase in wheel running
(40). The authors concluded that the fatigue resistance, and increased physical activity, was due
to an improved calcium homeostasis. Specifically, knocking out IL-​15Rα altered the expres-
sion of key calcium regulating proteins, SERCA2 and calsequestrin 1, which prevented fatigue.
The authors hypothesized that by preventing skeletal muscle fatigue, the muscle could maintain
repeated bouts of contraction and therefore increase wheel running. Supporting this hypoth-
esis is promising new research from the Pescatello lab that has demonstrated that men who are
chronically active have the IL-​15Rα 1775AA genotype as opposed to the IL-​15Rα 1775CC
genotype (6), confirming that the expression of IL-​15Rα can influence physical activity levels
in humans.
In 2014, we were the first to use the two-​dimensional differential in gel electrophoresis
(2D-​DIGE) method to elucidate peripheral candidate proteins that were associated with phys-
ical activity regulation (14). The proteins that were differentially expressed between high-​and
low-​active animals were primarily associated with metabolism, skeletal muscle structure, and
calcium regulation. Using previous quantitative trait loci (QTL) mapping results, we selected
key proteins of interest based on their expression in the 2D-​DIGE results and if the protein’s
genetic code existed in a QTL region previously identified as being involved in the biological
regulation of physical activity (14). With these criteria, annexin 6 and calsequestrin 1 were two
highly likely candidate proteins involved in the peripheral regulation of physical activity.
The annexin family of proteins, which consists of 17 different proteins in animals, is involved
primarily in Ca2+ homeostasis (46). Annexins are expressed in all tissues and share structural
and functional similarities with other members of the annexin family, with the exception of
annexin VI which contains a repeat sequence resulting in eight domains as opposed to the
standard four (21).The combination of the similarities in structure and the expression in various
tissues has led to difficulty in identifying the functions of individual members of the annexin
family (13). In fact, functions of annexin VI have been hypothesized to include inhibition of
blood coagulation, inhibition of protein kinase C, intracellular trafficking of endosomal vesicles,

110
111

Peripheral mechanisms regulating activity

regulation of voltage-​gated calcium cannels, structural modification of actin, and organization

of lipid components of cell membranes (9, 13, 46). The function of annexin VI in muscle is still
ambiguous partly because the role of annexin VI varies depending on the type of muscle. What
is known is that annexin VI binds acidic phospholipids of membranes in a Ca2+-​dependent
manner (3). Thus, the more calcium that is present, the more binding there is of annexin VI to
lipid membranes.
The function of annexin VI in skeletal muscle is less studied as compared to cardiac and
smooth muscle. Initially it was shown by Diaz-​Munoz et al. that annexin VI bound calcium
channels allowing for an increase in Ca2+ release (9). Additionally, Hazarika et al. demonstrated
that in skeletal muscle annexin VI is associated with the SR and that during the initial phase
of muscle contraction with the release of Ca2+ from channels of the sarcolemma, annexin VI
binds to the T tubule and propagates the further release of Ca2+ (17). Thus, Hazarika et al. (17)
proposed that annexin VI acts to stimulate more Ca2+ release from the SR, hence increasing
force of contraction (17).
Another regulator of calcium homeostasis in skeletal muscle involved in regulation of phys-
ical activity based on proteomic overexpression is calsequestrin 1 (Casq1). Unlike annexin VI,
Casq1 has been widely investigated for its function in skeletal muscle (4, 44). Casq1 is a Ca2+
buffer that helps to regulate the release of and reuptake of Ca2+ in skeletal muscle (42). Casq1
has a highly acidic C-​terminal tail with a high binding affinity for Ca2+. As Ca2+ binds to Casq1,
a conformational change in the structure occurs allowing for a further increase in Ca2+ binding
(4). It was initially thought that Casq1’s only function was to remove Ca2+ from the myoplasm
following muscle contraction; however, recent evidence has shown that Casq1 functions in
the release of Ca2+ from the ryanodine receptor (RyR) (42). It has been suggested that Casq1
acts on the RyR to inhibit Ca2+ release through triadin and junctin (4). As a result of these
functions, Casq1 acts as a Ca2+ sensor ensuring that Ca2+ levels are appropriately maintained in
the muscle cell.
There have been several studies that evaluated Casq1 expression following regular bouts
of exercise (20, 23, 37, 49). The studies by Jiao et al. (20) and Sugizaki et al. (49) were
conducted in cardiac tissue and showed that with an increase in endurance training there
was an increase in Casq1 protein expression, suggesting that Casq1 may play a role in the
increased endurance. Kinnunen and Manttari (23) examined Casq1 protein expression in
various skeletal muscle tissues following endurance or sprint training in a mouse model.
Interestingly, the authors observed that depending on the muscle fiber composition and the
type of training, there was a differential expression of Casq1 (23). Specifically, with endur-
ance training there was a decrease in Casq1 in the soleus (slow-​twitch fiber) while there
was an increase in Casq1 in the extensor digitorum longus (EDL, fast twitch fiber) (23).
Following sprint training, Casq1 was increased in the gastrocnemius (mixed muscle fiber)
and the EDL, with no change in the soleus (23). The authors noted that in fast-​twitch fiber
muscle the Ca2+ storage capacity is three to four times higher and the release of and uptake
of Ca2+ is more efficient compared to slow-​twitch fibers (23). Thus, the authors propose that
modulation of Casq1 following training allows for skeletal muscle to have increased cap-
ability to engage in the trained activity (23).
In summary, by showing that high-​ active animals have mechanisms that alter calcium
homeostasis –​increased expression of Anxn6 and Casq1 –​our 2D-​DIGE analysis (14) support
Pistilli’s earlier work (40) that showed through modification of IL-​15α there was an alteration
in calcium homeostasis in muscle which influenced physical activity level.
Adding to the complexity in the understanding of the peripheral regulation of physical
activity is that low-​active mice expressed their own unique protein signature in our 2D-​DIGE

111
112

D.P. Ferguson

results (14). This finding contradicted our original hypothesis which was that physical activity
level was determined by a unique set of proteins and if an individual expressed more of those
proteins they would be more active. Compared to the high-​active mice, low-​active mice had
a reduction in glycolysis enzymes which we hypothesized slowed glycolysis and limited the
generation of acetyl-​CoA with a resultant slowing of the Krebs’s cycle. Also, we observed an
upregulation in electron transport chain enzymes in the low-​active mice which might serve as
compensation for the slowing of glycolysis and the Krebs’s cycle, thus leading to no change in
metabolism or physical activity (14). Additionally, the low-​active mice had an increased expres-
sion in structural proteins of skeletal muscle (radixin, vinculin, vimentin, and the tubulins) that
have been shown to increase rigidity and impair force production during skeletal muscle con-
traction. This overexpression of these structural proteins and potential increase in rigidity could
increase fatigue and potentially impair physical activity engagement (14).
Combined, the above referenced studies encompassing both targeted and proteomic
approaches, have provided tentative evidence that substrate utilization and skeletal muscle con-
tractile performance may be primary peripheral factors that directly affect the capability to be
active. However, these conclusions have been primarily based on correlative results highlighting
the need for cause–​effect research designs to provide mechanisms of activity regulation.

Confirming candidate genes’ involvement in physical activity

An issue in many targeted genetics studies is the difficulty in providing cause–​effect relationships
between targeted genes/​proteins and the phenotype being studied. In order to confirm the
peripheral factors that influence physical activity, we utilized Vivo-​Morpholinos to transi-
ently silence annexin 6 and calsequestrin 1 genes in high-​active mice. Vivo-​Morpholinos are
oligonucleotides bound to a morpholino ring that will bind to the target mRNA and block
translation (15). The benefit of using Vivo-​Morpholinos is that their use allows for partial silen-
cing which is transient (15); thus we were able to knockdown annexin 6 and calsequestrin 1
in the high-​active C57L/​J mice to the levels of the C3H/​HeJ low-​active mice for 1 week. In
effect, we were able to transiently convert the skeletal muscle of high-​active mice to the pheno-
type of low-​active mice (in regard to annexin 6 and calsequestrin 1) in order to quantitatively
assess their role in physical activity regulation (15).
As shown in Figure 9.2, after baseline exposure to running wheels, mice treated with the
annexin VI Vivo-​Morpholino or the calsequestrin 1 Vivo-​Morpholino had a 40% decrease in
daily wheel running as compared to the control mice which received a Vivo-​Morpholino that
did not block the translation of any proteins. Once treatment with Vivo-​Morpholinos ceased
and the Vivo-​Morpholinos were metabolized (~1 week), wheel running returned to control
levels (14). From this data, we concluded that knocking down annexin VI and calsequestrin 1
to the level of a low-​active mouse impaired the skeletal muscles’ ability to maintain calcium
homeostasis, resulting in an increased rate of fatigue which ultimately led to a decrease in wheel
running.
While these results are exciting, this work could be greatly expanded. Specifically, since
we only assessed wheel running as a measure of physical activity, there are a variety of further
experiments that could be completed to further illustrate the balance between central (see
Chapter 8) and peripheral factors regulating physical activity especially with the advancement
of digital monitoring of home cages. For example, since the central drive to be active was not
altered during Vivo-​Morpholino treatments (14), quantifying the number of times a mouse gets
on (and off) the running wheel would be an interesting approach to determine if the transiently
silenced mice compensated for their reduced physical activity by engaging with the wheel more

112
113

Peripheral mechanisms regulating activity

Figure 9.2 Voluntary wheel running in mice that had skeletal muscle annexin VI or calsequestrin 1
transiently knocked down by Vivo-​Morpholinos. Wheel running was reduced when both annexin VI
and calsequestrin was reduced and recovered once the Vivo-​Morpholino was metabolized.Values are
expressed as means ± standard deviations.

than the control mice. In short, the use of technologies such as Vivo-​Morpholinos will allow
other investigators to further delineate peripheral factors regulating physical activity.

Summary
The literature shows there are inherent mechanisms by which skeletal muscle can regulate
physical activity with mechanisms associated with metabolism, contraction, and calcium flux
(Figure 9.1). Specifically, the ability to transport glucose and fatty acids to the working tissue
will increase physical activity levels. Furthermore, improved calcium homeostasis via IL-​15Rα,
annexin VI, and calsequestrin 1 will reduce skeletal muscle fatigue allowing for increased phys-
ical activity. Lastly, it should be noted that there is a unique set of proteins that increase skeletal
muscle rigidity and their overexpression –​as seen in low-​active animals –​may lead to a decrease
in physical activity levels.

References
1. Adlam D and Channon KM. Radial artery graft string sign due to lumen obliteration by
neointima: Insight from optical coherence tomography. JACC Cardiovasc Interv 4: 586–​587, 2011.
2. Adlam D, De Bono JP, Danson EJ, Zhang MH, Casadei B, Paterson DJ, and Channon KM.
Telemetric analysis of haemodynamic regulation during voluntary exercise training in mouse models.
Exp Physiol 96: 1118–​1128, 2011.

113
114

D.P. Ferguson

3. Babiychuk EB and Draeger A. Annexins in cell membrane dynamics. Ca(2+)-​regulated associ-

ation of lipid microdomains. J Cell Biol 150: 1113–​1124, 2000.
4. Beard NA, Wei L, and Dulhunty AF. Ca(2+) signaling in striated muscle: The elusive roles of
triadin, junctin, and calsequestrin. Eur Biophys J 39: 27–​36, 2009.
5. Ben-​Joseph E, Lee JS, Cromley EK, Laden F, and Troped PJ. Virtual and actual: Relative
accuracy of on-​site and web-​based instruments in auditing the environment for physical activity.
Health Place 19: 138–​150, 2013.
6. Bruneau M, Jr., Walsh S, Selinsky E, Ash G, Angelopoulos TJ, Clarkson P, Gordon P, Moyna
N, Visich P, Zoeller R, Thompson P, Gordish-​Dressman H, Hoffman E, Devaney J, and
Pescatello LS. A genetic variant in IL-​15Ralpha correlates with physical activity among European-​
American adults. Mol Genet Genomic Med, 6: 401–​408, 2018.
7. Cheng Y, Macera CA, Davis DR, Ainsworth BE, Troped PJ, and Blair SN. Physical activity
and self-​reported, physician-​diagnosed osteoarthritis: Is physical activity a risk factor? J Clin Epidemiol
53: 315–​322, 2000.
8. Church DM, Goodstadt L, Hillier LW, Zody MC, Goldstein S, She X, Bult CJ, Agarwala R,
Cherry JL, DiCuccio M, Hlavina W, Kapustin Y, Meric P, Maglott D, Birtle Z, Marques AC,
Graves T, Zhou S, Teague B, Potamousis K, Churas C, Place M, Herschleb J, Runnheim R,
Forrest D, Amos-​Landgraf J, Schwartz DC, Cheng Z, Lindblad-​Toh K, Eichler EE, Ponting
CP, and Mouse Genome Sequencing Consortium. Lineage-​specific biology revealed by a
finished genome assembly of the mouse. PLoS Biol 7: e1000112, 2009.
9. Diaz-​Munoz M, Hamilton SL, Kaetzel MA, Hazarika P, and Dedman JR. Modulation of
Ca2+ release channel activity from sarcoplasmic reticulum by annexin VI (67-​kDa calcimedin). J Biol
Chem 265: 15894–​15899, 1990.
10. Dishman RK, Renner KJ, Youngstedt SD, Reigle TG, Bunnell BN, Burke KA, Yoo HS,
Mougey EH, and Meyerhoff JL. Activity wheel running reduces escape latency and alters brain
monoamine levels after footshock. Brain Res Bull 42: 399–​406, 1997.
11. Eikelboom R. Human parallel to voluntary wheel running: Exercise. Anim Behav 57: F11–​F12, 1999.
12. Ekkekakis P, Hall EE, and Petruzzello SJ. Variation and homogeneity in affective responses to
physical activity of varying intensities: An alternative perspective on dose-​response based on evolu-
tionary considerations. J Sports Sci 23: 477–​500, 2005.
13. Enrich C, Rentero C, de Muga SV, Reverter M, Mulay V, Wood P, Koese M, and
Grewal T. Annexin A6-​linking Ca(2+) signaling with cholesterol transport. Biochim Biophys Acta
1813: 935–​947, 2011.
14. Ferguson DP, Dangott LJ, Schmitt EE, Vellers HL, and Lightfoot JT. Differential skeletal
muscle proteome of high-​and low-​active mice. J Appl Physiol (1985) 116: 1057–​1067, 2014.
15. Ferguson DP, Schmitt EE, and Lightfoot JT. Vivo-​morpholinos induced transient knockdown of
physical activity related proteins. PloS One 8: e61472, 2013.
16. Ge HY, Nie H, Graven-​Nielsen T, Danneskiold-​Samsoe B, and Arendt-​Nielsen L.
Descending pain modulation and its interaction with peripheral sensitization following sustained
isometric muscle contraction in fibromyalgia. Eur J Pain 16: 196–​203, 2012.
17. Hazarika P, Kaetzel MA, Sheldon A, Karin NJ, Fleischer S, Nelson TE, and Dedman JR.
Annexin VI is associated with calcium-​sequestering organelles. J Cell Biochem 46: 78–​85, 1991.
18. Heath GW and Troped PJ. The role of the built environment in shaping the health behaviors of
physical activity and healthy eating for cardiovascular health. Future Cardiol 8: 677–​679, 2012.
19. James P, Troped PJ, Hart JE, Joshu CE, Colditz GA, Brownson RC, Ewing R, and Laden F.
Urban sprawl, physical activity, and body mass index: Nurses’ Health Study and Nurses’ Health Study
II. Am J Public Health 103: 369–​375, 2013.
20. Jiao Q, Bai Y, Akaike T, Takeshima H, Ishikawa Y, and Minamisawa S. Sarcalumenin is essen-
tial for maintaining cardiac function during endurance exercise training. Am J Physiol Heart Circ
Physiol 297: H576–​582, 2009.
21. Kaetzel MA and Dedman JR. Annexin VI regulation of cardiac function. Biochem Biophys Res
Commun 322: 1171–​1177, 2004.
22. Kelly SA, Nehrenberg DL, Peirce JL, Hua K, Steffy BM, Wiltshire T, Pardo-​Manuel de
Villena F, Garland T, Jr, and Pomp D. Genetic architecture of voluntary exercise in an advanced
intercross line of mice. Physiological genomics 42: 190–​200, 2010.
23. Kinnunen S and Manttari S. Specific effects of endurance and spring training on protein expression
of calsequestrin and SERCA in mouse skeletal muscle. J Muscle Res Cell Motil 33: 123–​130, 2012.

114
115

Peripheral mechanisms regulating activity

24. Knab AM, Bowen RS, Hamilton AT, Gulledge AA, and Lightfoot JT. Altered dopamin-
ergic profiles: Implications for the regulation of voluntary physical activity. Behav Brain Res
204: 147–​152, 2009.
25. Knab AM, Bowen RS, Hamilton AT, and Lightfoot JT. Pharmacological manipulation of the
dopaminergic system affects wheel-​running activity in differentially active mice. J Biol Regul Homeost
Agents 26: 119–​129, 2012.
26. Knab AM and Lightfoot JT. Does the difference between physically active and couch potato lie in
the dopamine system? Int J Biol Sci 6: 133–​150, 2010.
27. Lightfoot JT.Current understanding of the genetic basis for physical activity.J Nutr 141: 526–​530, 2011.
28. Lightfoot JT. Why control activity? Evolutionary selection pressures affecting the development of
physical activity genetic and biological regulation. Biomed Res Int 2013: 821678, 2013.
29. Lightfoot JT, EJC DEG, Booth FW, Bray MS, M DENH, Kaprio J, Kelly SA, Pomp D, Saul
MC, Thomis MA, Garland T, Jr, and Bouchard C. Biological/​genetic regulation of physical
activity level: Consensus from GenBioPAC. Med Sci Sports Exerc 50: 863–​873, 2018.
30. Lightfoot JT, Leamy L, Pomp D, Turner MJ, Fodor AA, Knab A, Bowen RS, Ferguson D,
Moore-​Harrison T, and Hamilton A. Strain screen and haplotype association mapping of wheel
running in inbred mouse strains. J Appl Physiol (1985) 109: 623–​634, 2010.
31. Lightfoot JT, Leamy L, Pomp D, Turner MJ, Fodor AA, Knab A, Bowen RS, Ferguson D,
Moore-​Harrison T, and Hamilton A. Strain screen and haplotype association mapping of wheel
running in inbred mouse strains. J Appl Physiol 109: 623–​634, 2010.
32. Meek TH, Dlugosz EM, Vu KT, and Garland T, Jr. Effects of leptin treatment and Western diet
on wheel running in selectively bred high runner mice. Physiol Behav 106: 252–​258, 2012.
33. Meek TH, Eisenmann JC, and Garland T, Jr.Western diet increases wheel running in mice select-
ively bred for high voluntary wheel running. Int J Obes (Lond) 34: 960–​969, 2010.
34. Mokdad AH, Marks JS, Stroup DF, and Gerberding JL. Actual causes of death in the United
States, 2000. JAMA 291: 1238–​1245, 2004.
35. Moore-​Harrison T and Lightfoot JT. Driven to be inactive? The genetics of physical activity. Prog
Mol Biol Transl Sci 94: 271–​290, 2010.
36. Morey SS. ACSM revises guidelines for exercise to maintain fitness. Am Fam Physician 59: 473, 1999.
37. Myllymaki T, Rusko H, Syvaoja H, Juuti T, Kinnunen ML, and Kyrolainen H. Effects of exer-
cise intensity and duration on nocturnal heart rate variability and sleep quality. Eur J Appl Physiol
112: 801–​809, 2012.
38. Orstad SL, McDonough MH, Klenosky DB, Mattson M, and Troped PJ. The observed and
perceived neighborhood environment and physical activity among urban-​dwelling adults: The mod-
erating role of depressive symptoms. Soc Sci Med 190: 57–​66, 2017.
39. Peterson KE, Dubowitz T, Stoddard AM, Troped PJ, Sorensen G, and Emmons KM.
Social context of physical activity and weight status in working-​class populations. J Phys Act Health
4: 381–​396, 2007.
40. Pistilli EE, Bogdanovich S, Garton F, Yang N, Gulbin JP, Conner JD, Anderson BG, Quinn
LS, North K, Ahima RS, and Khurana TS. Loss of IL-​15 receptor alpha alters the endurance, fatig-
ability, and metabolic characteristics of mouse fast skeletal muscles. J Clin Invest 121: 3120–​3132, 2011.
41. Powers SK and Howley ET. Exercise Physiology. Theory and Application to Fitness and Performance.
New York: McGraw Hill, 2004.
42. Protasi F, Paolini C, Canato M, Reggiani C, and Quarta M. Lessons from calsequestrin-​1 abla-
tion in vivo: much more than a Ca(2+) buffer after all. J Muscle Res Cell Motil 32: 257–​270, 2011.
43. Rodriguez DA, Brown AL, and Troped PJ. Portable global positioning units to complement
accelerometry-​based physical activity monitors. Med Sci Sports Exerc 37: S572–​581, 2005.
44. Rossi AE and Dirksen RT. Sarcoplasmic reticulum: the dynamic calcium governor of muscle.
Muscle Nerve 33: 715–​731, 2006.
45. Smith AL, Troped PJ, McDonough MH, and DeFreese JD. Youth perceptions of how neigh-
borhood physical environment and peers affect physical activity: a focus group study. Int J Behav Nutr
Phys Act 12: 80, 2015.
46. Song G, Harding SE, Duchen MR, Tunwell R, O’Gara P, Hawkins TE, and Moss SE. Altered
mechanical properties and intracellular calcium signaling in cardiomyocytes from annexin 6 null-​
mutant mice. FASEB J 16: 622–​624, 2002.
47. Starnes HA, Troped PJ, Klenosky DB, and Doehring AM. Trails and physical activity: A review.
J Phys Act Health 8: 1160–​1174, 2011.

115
116

D.P. Ferguson

48. Stenbit AE, Burcelin R, Katz EB, Tsao TS, Gautier N, Charron MJ, and Le Marchand-​
Brustel Y. Diverse effects of Glut 4 ablation on glucose uptake and glycogen synthesis in red and
white skeletal muscle. J Clin Invest 98: 629–​634, 1996.
49. Sugizaki MM, Leopoldo AP, Conde SJ, Campos DS, Damato R, Leopoldo AS, Nascimento
AF, Oliveira Junior Sde A, and Cicogna AC. Upregulation of mRNA myocardium calcium
handling in rats submitted to exercise and food restriction. Arq Bras Cardiol 97: 46–​52, 2011.
50. Tamura K, Puett RC, Hart JE, Starnes HA, Laden F, and Troped PJ. Spatial clustering of
physical activity and obesity in relation to built environment factors among older women in three
U.S. states. BMC Public Health 14: 1322, 2014.
51. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, and McDowell M. Physical activity
in the United States measured by accelerometer. Med Sci Sports Exerc 40: 181–​188, 2008.
52. Troped PJ and Saunders RP. Gender differences in social influence on physical activity at different
stages of exercise adoption. Am J Health Promot 13: 112–​115, 1998.
53. Troped PJ, Saunders RP, Pate RR, Reininger B, and Addy CL. Correlates of recreational
and transportation physical activity among adults in a New England community. Prev Med
37: 304–​310, 2003.
54. Troped PJ, Tamura K, Whitcomb HA, and Laden F. Perceived built environment and physical
activity in U.S. women by sprawl and region. Am J Prev Med 41: 473–​479, 2011.
55. Troped PJ, Wiecha JL, Fragala MS, Matthews CE, Finkelstein DM, Kim J, and Peterson KE.
Reliability and validity of YRBS physical activity items among middle school students. Med Sci Sports
Exerc 39: 416–​425, 2007.
56. Troped PJ, Wilson JS, Matthews CE, Cromley EK, and Melly SJ. The built environment and
location-​based physical activity. Am J Prev Med 38: 429–​438, 2010.
57. Tsao TS, Li J, Chang KS, Stenbit AE, Galuska D, Anderson JE, Zierath JR, McCarter RJ, and
Charron MJ. Metabolic adaptations in skeletal muscle overexpressing GLUT4: Effects on muscle
and physical activity. FASEB J 15: 958–​969, 2001.

116
117

10
TOXICANT AND DIETARY
EXPOSURES AS UNIQUE
ENVIRONMENTAL FACTORS
THAT AFFECT THE GENETIC
REGULATION OF ACTIVITY
Emily E. Schmitt and Heather L. Vellers

Introduction
As indicated in previous chapters, genetic background plays a crucial role in determining an
individual’s daily physical activity patterns; however, genetics is not the only factor driving physical
activity. Like other behavioral phenotypes, physical activity is a complex behavior and is driven by
various physiological variables not only influenced by genetics, but also, combined effects of gen-
etics and environmental factors. When considering environmental influences on physical activity,
the general thought is that common environmental factors, an example of which is our ‘built
environment” or presence and ease of access to sidewalks, parks, trails, and recreational facilities, is
an important factor in determining individual activity engagement.This thought, however, has yet
to be supported by evidence in adults where physical activity has been measured objectively (13,
18, 49). Thus, ease of access to a gym or parks does not necessarily prompt an increase in phys-
ical activity among individuals. Less commonly considered are unique environmental factors that
affect physical activity, defined as all individual-​specific experiences (see Chapter 6). Such factors
have been calculated to account for 8–​52% of the observed interindividual variation in daily
activity patterns (33, 62, 66, 71). The precise unique environmental factors that directly influence
an individual’s physical activity are beginning to be studied, with work isolating factors including
specific types of environmental chemicals and diet demonstrating a significant and detrimental
effect of these factors on physical activity patterns. In this chapter, the purpose is to discuss two
unique environmental influencers of the genetic regulation of activity and potential mechanisms
(32) through which each are hypothesized to regulate activity.

How might unique environmental factors influence an

individual’s drive to be active?
Prior to genetics being considered a regulator of physical activity, work from the early 1920s
and 1930s suggested that daily activity was controlled through biological mechanisms (58, 74).
Wang, Richter, and colleagues (58, 65, 74) investigated the relationship between the female

117
118

E.E. Schmitt and H.L. Vellers

reproductive cycle and wheel-​running behavior in rodents. In 1923, Wang et al. (74) reported
a cyclical relationship between content inside the vagina of female rats and activity levels, and
noted female mice were most active when they were in “heat” (74). Shortly thereafter, Slonaker
(65) followed up on Wang’s findings and concluded physical activity in females was driven by
“copulation and successful stimulation of the cervis uteri,” or what we presently refer to as the
estrous cycle (65). In males, a role of the reproductive system on physical activity regulation was
also apparent, when in 1925, Hoskins (30) demonstrated through castrating male rats that daily
wheel running was significantly reduced and remained significantly lower than their control
counterparts throughout their lifetime.Taken together, these initial studies were among the first
to suggest a strong relationship between physical activity and the estrous cycle in females and
between physical activity and testicular function in males.
Since these early studies, multiple other studies have shown a strong influence of sex hormones
in regulating physical activity. For example, Gorzek et al. (25) found female ovariectomized
mice displayed significant reductions in wheel-​running activity by up to 80%, while physio-
logical replacement of estrogen increased activity in these mice to approximately 54% of base-
line. Additionally, Bowen and colleagues (9) found that removal of sex hormones resulted
in approximately 90% decreases in the daily distances ran in male and female mice, while
exogenous replacement of these hormones resulted in varying levels (35–​110%) of recovered
baseline activity. Further, Bowen, et al.’s (9) results revealed specific influences of each hormone
on activity patterns, where recovered amounts of distance and duration ran were primarily
influenced by testosterone administration while the recovery of speed was primarily influenced
by 17β-​estradiol.
While many studies have shown that sex hormones regulate physical activity, the mechanisms
through which sex hormones work are somewhat unclear. Roy and Wade (59) examined the
effect of nonaromatizable (dihydrotestosterone proprionate) and aromatizable (testosterone
proprionate) androgen administration in male rats that underwent castration to determine
whether the conversion of testosterone to estrogen through the aromatase complex was a
pathway regulating male rat activity. The results showed the aromatizable form of testosterone
had the greatest effect on physical activity, while the nonaromatizable form of testosterone had
no effect on activity suggesting that activity regulation was dependent on testosterone con-
version to estradiol via the aromatase complex (59). Conversely, more recent work by Bowen
et al. (8), showed no effect on male mouse physical activity when the aromatase complex
was inhibited through administration of reversible and irreversible aromatase pharmacological
interventions, calling into question the applicability of Roy and Wade’s results in a mouse model.
Lastly, in male mice only, Jardi et al. (31) investigated the role of testosterone, and identi-
fied a pathway through which testosterone elicits its effect on physical activity. These authors
showed the free fraction of testosterone stimulated activity by acting on central (brain) dopa-
minergic pathways via the androgen receptor, and indirectly through the aromatase complex.
Furthermore, using a muscle-​specific androgen receptor knockout mouse model and then pro-
viding exogenous testosterone, the authors demonstrated that testosterone-​induced increases
in activity do not occur through muscle or peripheral mechanism(s) (31). Thus, in total, work
demonstrates direct effects of testosterone and 17β-​estradiol on physical activity regulation
in male and female rodents (25, 31, 59). However, additional mechanistic work is needed in
females to determine whether similar sex-​dependent pathways as in males participate in physical
activity regulation through these hormones. Additionally and perhaps most critically, the work
by Jardi et al. (31) merged two separate bodies of literature involving the role sex hormones and
dopamine play in regulating activity regulation with other studies previously linking wheel-​
running behavior in mice to genomic-​dependent dopaminergic pathways (38, 39). Thus, the

118
119

Toxicant and dietary exposures

specific dopaminergic pathways stimulated by sex hormones and resulting physical activity may
be dependent on genetic background, especially in the dopaminergic pathways (see Chapter 8).
As previously mentioned, unique environmental factors have been suggested to play signifi-
cant direct roles in regulating physical activity. One potential pathway through which unique
environmental factors may influence physical activity is via alteration in sex hormone levels
given that alterations in sex hormone levels can have dramatic effects on physical activity levels.
Thus, the remaining sections of this chapter will discuss specific, unique environmental factors
that have been suggested to have direct effects on sex hormones, and thus, may indirectly affect
physical activity.

Endocrine-​disrupting chemicals
The harmful consequences of endocrine-​disrupting chemicals (EDCs) on human health have
been widely studied over the past few decades and growing evidence has supported that man-​
made chemicals contribute to adverse health effects on humans and wildlife (3). Not only do
EDCs lead to harmful developmental, reproductive, neurological, and immune effects through
the alteration of endocrine systems, they also cost the US healthcare system hundreds of billions
of dollars (3). EDCs are defined as man-​made synthetic substances that alter an organism’s
internal milieu which presents a risk of death, disease, or birth defect through absorption, inges-
tion, and inhalation (23). EDCs are found in, but not limited to, metals, plastics, personal care
products, pesticides, pharmaceutical drugs, and industrial chemicals. Endocrine disruptors affect
hormone response by mimicking hormones and binding agonistically or antagonistically to hor-
mone receptors. EDCs may turn off, turn on, or modify hormonal signals, which affects normal
functioning of tissues and organs (14). For example, EDCs can mimic estrogens, androgens, and
thyroid hormones by competitively binding to the appropriate receptor and potentially produ-
cing over-​or understimulation of these receptors (5). Exposures to EDCs are more dangerous
if exposure occurs during “critical periods” of life when organisms are still developing and are
more sensitive to hormonal disruption (23). The developing brain relies strongly on proper
functioning hormones and disruption in hormone receptor function can lead to altered neural
development of the brain (reviewed in 56). Specifically, it has recently been shown that EDCs
can have a profound effect on early life brain development due to nervous system alterations in
the hypothalamus–​pituitary–​thyroid gland axis (reviewed in 34). These alterations of the brain’s
connectivity and development have further been elucidated when mice exposed to low doses of
a single EDC resulted in disruptions in midbrain dopaminergic nuclei of mice, where physical
activity is thought to be regulated (67). Given that sex hormones regulate physical activity and
that EDCs have a profound effect on hormone activity, it is possible that ECDs can cause lower
physical activity levels through sex hormone disruption as illustrated in Figure 10.1.
While there are literally thousands of EDCs, there are some common EDCs that alter regu-
lation of physical activity including bisphenol A (BPA), several different categories of phthalates,
and a common fungicide, vinclozolin. BPA is a chemical produced in large quantities and
is primarily used in production of plastics, PVC, and epoxy resins. BPA has been studied in
a variety of in vivo and in vitro models, is estrogenic, and has been found to have significant
adverse effects on mammary gland development in female mice (47) and sperm counts in males
(42). Phthalates or phthalate esters are considered EDCs (35) and encompass a large group of
compounds that share similar chemical structures. Common phthalates include diethylhexyl
phthalate (DEHP), dibutyl phthalate (DBP), butyl benzyl phthalate (BBP), di-​isononyl phthalate
(DINP, DiBP, DIDP), and dipentyl phthalate (DPP); all have been shown to have adverse effects
on human health primarily as antiandrogenic compounds (37, 55). Phthalate compounds have

119
120

E.E. Schmitt and H.L. Vellers

Figure 10.1 Schematic outlining the unique environmental factor of endocrine-​disrupting chemicals
(EDCs) and the effect on physical activity regulation. Research has shown that EDCs influence sex
hormones, but the mechanisms through which the EDCs act to control activity are not fully understood
(indicated by the dotted line). a) Vinclozolin exposure decreases activity in females. b) BPA exposure
decreases physical activity in females. c) BBP exposure decreases physical activity in males and females.

been in the environment since the 1930s and are often found in plastics (most concerning in
toys for infants) (35), and personal care products, such as perfumes, lubricants, aerosols, and nail
polish products (75). Food is also a major source of phthalate exposure, specifically DEHP, with
humans having diets high in meat and dairy showing a twofold increases in DEHP exposure
(64). Finally, vinclozolin is a common fungicide used to control diseases in crops and has been
characterized as an endocrine disruptor (antiandrogenic) and is a possible human carcinogen
with exposure causing disease and abnormalities for up to four generations after exposure (2),
with significant major morphological alterations observed in exposed male rats (27). In short,
EDCs are ubiquitous in our environment with exposure ranging from personal care products
to food packaging and processing. Disruption in hormone levels from EDC exposure can cause
a variety of physiological problems that vary in severity and differ by sex, yet few studies have
examined the direct relationship between EDC exposure and physical activity.

Endocrine disruptors’ effect on physical activity

Given that sex hormone levels alter physical activity and EDC exposure alters sex hormone
physiology, it is possible that EDCs affect physical activity level through their effects on sex
hormones. The number of studies considering EDC influences on activity is few, with only
three studies tackling the question of the effect of EDCs on physical activity (Table 10.1). The
available studies have used animal models as they allow investigators to directly assess exposure
to EDCs and resulting activity levels.
One of the earliest studies to investigate if a toxicant led to changes in activity levels was
from Flynn et al. (21) who studied the fungicide vinclozolin. In 2001, Flynn et al. (21) showed
that female rats, but not male rats, exposed to long-​term dietary vinclozolin both pre-​and
postnatally had alterations in locomotor activity compared to control animals. The researchers
fed pregnant rats soy-​free diets containing vinclozolin (approximately 0, 0.8, 12, and 60 mg/​
kg/​day for an adult) beginning on gestational day 7, and the diets were continuously fed to the

120
 newgenrtpdf
121
Table 10.1 Summary of studies demonstrating effects of endocrine-​disrupting chemicals (EDCs) on physical activity in rodents

EDC Study Species Species characteristics Study design Effect of EDC on physical
activity

Vinclozolin Flynn et al. Rat Sprague Dawley male 2 weeks prior to mating, females were shifted from standard In the high-​dosed vinclozolin
(21) and female rats diet to irradiated soy-​free diet. The vinclozolin-​dosed diets group, females (but not
began on gestation day 7 and continued through weaning males) demonstrated
on day 22. Weaned pups continued their respective diet until decreased running-​wheel
sacrifice on day 77. Resulting pups were randomly assigned activity
to behavioral measures ranging from wheel-​running activity
to play behavior
Bisphenol Johnson Mouse Founder outbred adult 2 weeks prior to mating, females were assigned to a diet BPA-​exposed females
A (BPA) et al. (32) California mice group (control or BPA) and maintained their respective (but not males) were
diets throughout gestation and lactation. Weaned males significantly less
and females were then placed on the control diet for the active than control
remainder of the study counterparts as
indicated by calorimetry
measurements (cage
activity), but no
differences in voluntary
wheel running were
examined in either sex
Benzyl butyl Schmitt Mouse C57BL/​6J inbred mice Pregnant damns were exposed to BBP via oral gavage on Male and female pups
phthalate et al. (62) days 9–​16 of gestation. Resulting male and female pups that were exposed to
(BBP) were weaned and analyzed for the study BBP prenatally showed
significant decreases in
voluntary wheel running
compared to controls
122

E.E. Schmitt and H.L. Vellers

offspring until sacrifice on postnatal day 77. It was found that females in the high-​dose group
were significantly less active than control animals, while treated males did not show significant
alterations in wheel activity (21). Given that vinclozolin is antiandrogenic, Flynn et al.’s (21)
results suggest that the decreased activity due to vinclozolin exposure was due to alterations in
the androgenic system.
Johnson et al. (32) examined the effect of the antiestrogenic toxicant BPA on physical activity
in founder outbred adult California mice (32). Two weeks before breeding, virgin female mice
were randomly assigned to different diets and remained on the same diets throughout gestation
and lactation. Diet #1 was a low-​phytoestrogen diet supplemented with corn oil, diet #2 was
the same as the first diet with 50 mg of BPA/​kg feed weight, and diet #3 was the same phyto-
estrogen diet supplemented with 0.1 parts per billion of ethinyl estradiol. Importantly, the authors
found BPA-​exposed females were significantly less active, as evident by several indirect calorim-
etry measurements. In males, no significant differences were observed. Thus, while these results
suggest that BPA exposure can affect physical activity, there are indications that males and females
may be differentially affected due to differential mechanisms of physical activity regulation.
Schmitt and colleagues (62) found prenatal exposure to the antiandrogenic phthalate BBP
resulted in decreased anogenital distances (indicates puberty development in male mice) and
delayed vaginal openings (indicates puberty development in female mice). Additionally, serum
testosterone concentrations were significantly lower in prenatally exposed male mice when
measured during adolescent and adult years, indicating a lasting effect of in utero exposure to
BBP on sex hormone function throughout life. In females, a similar decrease in testosterone
and significant reductions in estrogen concentrations were found in prenatally exposed adults
when compared to controls. Concurrent with these observations, wheel-​running activity from
8 to 20 weeks of age in the BBP-​exposed offspring was decreased (on average) by 20% in male
and 15% in female mice compared to control counterparts. This study was the first to suggest
maternal endocrine disruption can cause an altered hormone response, with a corresponding
decreased activity throughout early adulthood in all offspring. Overall, these results suggest in
utero exposure to a common antiandrogenic EDC can reduce sex hormone concentrations and
physical activity in mice, regardless of sex (62).
To date, only three studies exist that directly assessed the impact of prenatal exposure of
EDCs on physical activity in offspring (see Table 10.1). The results of these three studies con-
firm that environmental toxicants can decrease physical activity and need to be considered as
a possible factor causing physical inactivity. In addition, it appears that EDC effects on physical
activity are mediated through sex hormone systems, and males and females appear to be differ-
entially affected, perhaps indicating different mechanisms of action between sexes. Finally, even
though the link has been made between the biological pathway interacting with the unique
environment to control physical activity, it is also important to recognize that the dopaminergic
system could play a factor as well (see Chapter 8). In a comprehensive review on neurotoxicity
of EDCs, Masuo and Ishido (48) suggest that EDCs disrupt the development of the central
dopamine system neurons in the brain by affecting monoaminergic neurons in the central ner-
vous system. These neural effects of EDCs may play a role in causing the behavioral alterations
observed when EDC effects on activity are study and are certainly a potential avenue for further
investigation.

Effects of diet on physical activity regulation

Like EDCs, dietary intake is another unique environmental influence that potentially affects
pathways of physical activity regulation. Observations that diet directly alters an individual’s

122
123

Toxicant and dietary exposures

physical activity patterns dates back to studies in the early 1800s (52, 73) –​as reviewed by
Casper (12) –​that documented “heightened physical activity” as a hallmark behavioral charac-
teristic in individuals diagnosed with anorexia nervosa (AN; an eating disorder characterized
by varying degrees of caloric deprivation and malnourishment). Casper (12) highlighted that
heightened levels of activity (e.g., “restlessness” and “increased drive” to move), were typically
found only after following a period of self-​starvation. From this, Casper hypothesized that the
caloric restriction related to AN led to increased physical activity, as opposed to the prevailing
belief that increased physical activity led to AN. While Casper addressed the hypocaloric-​
induced hyperactivity response phenomenon as it relates to a diseased condition, this response
falls in line with the theory that human locomotor activity has evolved partially to a selective
pressure of food availability (discussed in 44). Briefly, throughout evolution, a foraging mech-
anism (e.g., foraging gene) is hypothesized to have been stimulated by food availability, where
low food availability led to increased locomotor activity in search for food, while food abun-
dance inhibited, or simply did not trigger this mechanism. In the current state of extremely high
food abundance and availability in Western societies, it brings to question whether this shift of
increased food availability has contributed to the low daily activity levels among US residents
(69). Research specifically addressing the effect of dietary patterns on physical activity is in
its infancy, however, evidence from both caloric restriction or overfeeding demonstrate links
between feeding and physical activity. For the purposes of this chapter, this section will focus
on evidence linking caloric restriction with increased activity with associated pathways through
which this response occurs and evidence from work that has begun investigating the effect of
caloric excess on activity.

Caloric restriction
The caloric restriction-​ induced increases in physical activity have been linked to various
neurochemicals central to the brain, including leptin (4, 16, 26, 72), serotonin (28), and dopa-
mine (36). The adipose tissue-​derived hormone leptin, is one of the most widely known
and investigated mechanisms associated with dietary restriction and increased activity (4, 19,
26, 29, 43, 72). The secretion of leptin is correlated with fat mass, where leptin secretion is
increased with increases in fat mass, and decreased with decreases in fat mass (1). The involve-
ment of leptin in diet and activity regulation has been tested through leptin administration in
AN patients and activity-​based anorexia (ABA) rodent models, where increased leptin reduced
hyperactive responses (17, 43). A study by Verhagen et al. (72) demonstrated that low levels of
leptin associated with increased wheel running was mediated through the ventral tegmental area
of the brain, and speculated that the suppressed leptin levels induced an increased firing rate of
the dopaminergic neurons. Interestingly, the dopaminergic neurons contain highly expressed
leptin receptors (20), and may provide one mechanism through which leptin participates in
caloric restriction-​induced hyperactivity.
While plasma leptin has been suggested as a key player between caloric restriction and hyper-
activity, such work is correlational and has not been directly tested. Some evidence suggests the
hypocaloric-​induced increases in physical activity are independent of leptin levels (28, 54).
For example, one study by Morton et al. (54) demonstrated a hypocaloric-​induced increased
activity response independent of leptin, through fasting (fasted for 24 hours) wild-​type and
ob/​ob (leptin-​deficient) mice. Fasting in the leptin-​deficient mice led to significant increases
in both ambulatory and wheel-​running activity. Hillebrand et al. (28) furthered this work by
evaluating the effect of the atypical antipsychotic drug, olanzapine (Zyprexa), on wheel running
in an ABA rat model, as well as measuring physical activity in a cohort of AN patients who

123
124

E.E. Schmitt and H.L. Vellers

were classified as “hyperactive” upon hospital admittance. Olanzapine is a thienobenzodiazepine

compound, which has a high affinity for 5-​HT2A/​AC receptors, histamine (H1) receptors, adren-
ergic (α1 receptors), and a moderate affinity for dopamine (D1–​D4) receptors (10, 11, 53, 63). In
Hillebrand, et al.’s study (28), olanzapine was found to inhibit the heightened wheel running
in ABA rats as well as the hyperactive responses in patients with AN. However, olanzapine
inhibited the hyperactive responses without altering plasma leptin levels suggesting that other
nonleptin mechanisms were involved in the hyperactive responses. Hillebrand et al. (28)
hypothesized that hypocaloric hyperactivity was mediated through serotonin activation (5-​
HT2A/​AC receptors), histamine receptors, and/​or dopamine given the high–​moderate affinity of
olanzapine on these systems. In rats, Klenotich et al. (36) expanded upon this thought by select-
ively blocking 5-​HT2A/​AC receptors and dopamine receptors (DR) DR2 and DR3, and found
that selective antagonism of DR2 and DR3 reduced ABA, whereas antagonism of 5-​HT2A/​AC
did not.Therefore, this study concluded that the D2 and D3 receptors were mechanisms through
which olanzapine reduced activity in ABA rats. As noted previously, earlier work has established
the dopaminergic system as a potential regulator of physical activity where biological (31) and
genetic predisposition (38) link to influence physical activity. While the dopaminergic system
presents one pathway through which caloric restriction could increase physical activity, it would
be interesting to further investigate how caloric restriction, as a unique environmental factor,
interacts with other biological/​genetic factors to then influence physical activity.

Overfeeding
Compared to literature describing links between caloric restriction and increased activity, less
work (Table 10.2) has described the effect of caloric excess on activity, especially in terms of
potential mechanism(s) mediating how this response might occur. Of the studies in Table 10.2,
only one has directly investigated a link between chronic overfeeding and physical activity.
Similar to EDCs, chronic overfeeding has been observed to alter sex hormones levels and
functions (7). In males, a direct effect of chronic overfeeding on sex hormones is known (7),
where, in general, androgen levels are decreased and estrogens are increased. In females, how-
ever, direct ties between chronic overfeeding and sex hormone levels are unclear. For example,
overfeeding-​induced metabolic diseases in women (e.g., polycystic ovarian syndrome, type 2
diabetes, and obesity) show an opposing effect of chronic overfeeding on sex hormone levels as
compared to men: increased androgens and decreases in estrogens (15). Given that it is established
that sex hormones regulate activity, and that overfeeding potentially alters sex hormones (7),
we hypothesized that alterations in sex hormones after chronic overfeeding were significant
mediators of a reduced wheel-​running activity in mice (71). With the chronic overfeeding of
a high-​fat, high-​sugar diet, there was a significant reduction in acute wheel-​running activity in
males (~70%) and females (~57%) (71). However, this decreased activity did not appear to be
mediated through alterations in sex hormones given that sex hormone levels between control
and intervention animals did not differ and exogenous supplementation of sex hormones in
overfed animals did not alter physical activity.Thus, chronic overfeeding decreased daily activity,
but apparently through a mechanism(s) not involving sex hormones.
Another possible mediator of overfeeding-​induced reductions in physical activity is the dopa-
minergic system, which as noted above, has been linked to hypocaloric-​induced increases in phys-
ical activity through D2 and D3 receptors (36). The dopaminergic system was previously linked
to genetic differences in wheel-​running activity (38), has direct effects on physical activity via sex
hormones (31), and is directly affected by overfeeding (68). Briefly, dopamine, as a neurotrans-
mitter, functions via neural circuits to regulate pleasure, and through its production, reinforces

124
 newgenrtpdf
125
Table 10.2 Summary of studies demonstrating direct and indirect links between overfeeding and reduced activity in humans and rodents

Diet type Study Species Subject characteristics Study design Effect of diet on physical activity

Excess calories Levine et al. Human 22 healthy men and women, Weight maintenance feeding occurred Overfeeding reduced free-​living
(41) lean and obese sedentary during the first 10 experimental days, walking to a similar degree in lean
individuals aged 39 ± and baseline locomotor activity was and obese subjects (~1.5 miles/​day)
8 years also assessed during this time. These
measurements were repeated again after
8 weeks of being overfed by 1000 kcal/​
day
Excess calories Schmidt Human 55 healthy men and Spontaneous physical activity was 3 days following overfeeding, OP
et al. (61) women aged 25–​ measured using a physical activity subjects significantly decreased the
35 years classified as monitoring system, either in a amount of time they spent walking
either obesity resistant controlled eucaloric condition or (−2.0% of time), whereas OR
(OR) or obesity during 3 days of overfeeding (1.4 × subjects maintained their walking
prone (OP) defined basal energy) and for the subsequent (+0.2%)
by body mass index 3 days (ad libitum recovery period)
of 16.9–​25.5 kg/​m2
and 19.6–​30.6 kg/​m2,
respectively
“Very high-​fat Funkat Mice C57BL/​6, DBA/​2, and Mice were fed either a control or a 129T2 mice fed either diet exhibited
diet” et al. (24) 129T2 aged 14 weeks high-​fat (60% fat) diet for 6 weeks. less voluntary activity compared
During week 6, voluntary wheel-​ with both C57BL/​6 and DBA/​2
running activity was assessed for mice. The high-​fat diet decreased
7 days and readings were taken on voluntary activity in both C57BL/​
days 5–​7, then averaged, and expressed 6 and DBA/​2 mice, but there was
as cycles/​day no further decrease in activity in
the 129T2 mice
(continued )
 newgenrtpdf
126
Table 10.2 (Cont.)

Diet type Study Species Subject characteristics Study design Effect of diet on physical activity
“Westernized Meek et al. Mice Male selectively bred From 3 to 12 weeks of age, mice Western diet initially caused increased
diet” (51) high-​runner (HR) received either a standard chow diet caloric intake in both HR and C
mouse line, and or Western diet and wheel-​running mice, but this effect was reversed
corresponding control activity was assessed during the last 4 weeks of the
(C) line aged 12 weeks study. Western diet had little or no
effect on wheel running in C mice,
but increased revolutions per day
by as much as 75% in HR mice,
mainly through increased time
spent running
“High fructose Rendeiro Mice Male C57BL/​6J mice Mice received either a diet containing Mice given the fructose diet traveled
diet” et al. (57) aged 14 weeks fructose or glucose at 18% for 77 days. significantly less (~20% less) in
Wheel-​running activity was assessed comparison to animals fed the
on days 76 and 77 glucose diet
High-​fat/​high-​ Vellers et al. Mice Male and female C57BL/​ At 3 weeks of age, mice were assigned HFHS diet decreased acute wheel-​
sugar diet (71) 6J mice aged 12 weeks to either a control (CFD) or high-​ running activity in male (~70%)
fat/​high-​sugar (HFHS) diet for 9–​11 and female mice (~57%)
weeks. Acute wheel-​running activity
was assessed for 3 days at 12 weeks
of age
“Westernized Bjursell Mice Male C57Bl/​6J mice aged Mice were given Westernized diet or Mice that switched from chow to
diet” et al. (6) 8 weeks chow diet for 21 days. At weeks 3 Westernized diet had significant
and 7, mice were switched to the reductions in locomotor activity
opposing diet and locomotor activity after only 3–​5 hours following the
was analyzed for 24 hours diet switch
127

Toxicant and dietary exposures

rewarding behaviors. In mice, overfeeding via a high-​fat diet suppresses dopamine production,
with this suppression shown in areas of the brain known to house mechanisms regulating physical
activity (e.g., striatum/​nucleus accumbens) (60). Hence, it is possible that suppression of dopamine
production, specific to dopamine receptors in areas of brain associated with activity regulation,
could partially explain why motivation to be physically active decreases because of overfeeding.
In this context, it is also important to consider genetic predisposition, where different phys-
ical activity phenotypes are influenced by divergent dopamine signaling pathways (high-​active
strain=dopamine 1-​like receptor; low-​active strain=dopamine reuptake) (38). Thus, it is possible
that there is a potential interactive effect between overfeeding-​induced alterations to dopamine sig-
naling and a subsequent influence on physical activity. Even though it remains unknown whether
the dopaminergic system links chronic overfeeding and reduced activity, the separate literature
on overfeeding and physical activity and direct associations with dopaminergic pathways suggests
there could be a possible link between overfeeding and dopaminergic control of physical activity.
When exploring potential mechanisms of an overfeeding-​induced reduction in physical
activity, another factor that is frequently suggested is body weight. Some work suggests that
weight alone may influence physical activity primarily because the increased energy cost of
movement with increased body weight correlates with declines in physical activity (e.g., 40, 70).
However, this correlation does not arise in all cases, as has been demonstrated in numerous other
studies (e.g., 21, 45, 46, 71). An important concept to bear in mind is that with overfeeding-​
induced increased body weight comes a myriad of other physiological and body compos-
ition changes. Thus, studies simply demonstrating a correlation with increased body weight
and lowered activity cannot rule out other potential factors linking overfeeding with decreased
physical activity. As noted above, other work that has assessed the influence of weight on activity
primarily in mouse models has repeatedly shown an insignificant relationship between weight
and activity (22, 45, 46, 71). Most recently, work by Friend et al. (22), ruled out an effect of
body weight on activity, when restoration of DR2 binding of the basal ganglia that was altered
by the diet, recovered baseline activity in spite of no change in body weight. While the current
understanding of whether body weight directly influences activity is limited to the mouse
model, the available evidence suggests that weight is not an independent factor inhibiting phys-
ical activity. As such, body weight, among the host of other chronic overfeeding-​induced body
composition and physiological changes, must be further investigated to delineate which factor(s)
are involved in pathways causing decreased physical activity.
While overfeeding in most studies has been shown (and associated) with reduced physical
activity, there is evidence showing that genetic predisposition for specific physical phenotypes
can interact with dietary intake and influence differential activity responses. Specifically, Meek
et al. (51) utilized a specialized “Western diet” (42% fat) and found this diet actually increased
wheel-​running activity in mice selectively bred for high running-​wheel activity when compared
to a selected control line.The authors postulated the high-​fat diet stimulated further increases in
the already high-​running activity-​selected mouse line because these mice normally expended
their available energy stores daily. The authors hypothesized that the extra fat in the Western
diet provided an additional fuel source for these high-​active mice (51). Meek et al. (50) further
demonstrated the Western diet-​induced increased wheel running in the high-​active mouse
line was linked to leptin sensitivity. When the authors gave leptin to the high-​active mice
that were on the standard diet, their activity levels increased and were similar to those animals
on the Western diet (50). Lastly, another study by Funkat et al. (24), provided the specialized
“very high-​fat” diet to three different inbred mouse strains (C57BL/​6, DBA/​2, and 129T2),
and found compared to baseline activity, wheel-​running activity was significantly reduced in
strains characterized with moderate activity (C57BL/​6 and DBA/​2; 46), while unchanged in

127
128

E.E. Schmitt and H.L. Vellers

Figure 10.2 The effect of chronic overfeeding on physical activity regulation and potential mechanisms.
Indirect evidence in humans (41, 61), and in mice (71) show chronic overfeeding inhibits physical
activity. Of these studies,Vellers et al. (71) tested whether the primary sex hormones in males
(testosterone) and females (17β-​estradiol) were mechanisms linking chronic overfeeding with reduced
physical activity; however, the findings demonstrated chronic overfeeding reduces activity independent of
sex hormone involvement as depicted in the diagram.

a low-​active strain (129T2) (45). Overall, the two studies by Meek et al. and Funkat et al. (24,
51) show a gene by “unique” environment interaction, where calorically dense diets appear
to enhance physical activity in mice with a genetic predisposition to be highly active, with
a decrease in activity in mice that do not have the selectively bred highly active genome. If
extended to humans, these findings would likely suggest that only elite athletes would have
enhanced physical activity with overfeeding and all other humans (i.e., majority of the popula-
tion) would incur an inhibited physical activity level with overfeeding.
In summary, the collective evidence discussed in this section clearly presents an effect of diet
on physical activity regulation. As depicted in Figure 10.2, the sex hormones are the only factors
that have been directly assessed as potential mediators between overfeeding and reduced activity.
Though work indicates sex hormones do not regulate diet-​induced changes in activity, there
are a myriad of other chronic overfeeding-​induced changes that warrant future investigations
(e.g., dopaminergic system). Despite the lack of clarity regarding how these mechanism(s) are
linked, the knowledge that diet influences whether an individual is active or sedentary has sub-
stantial applications for exercise physiologists seeking to find modifiable factors to encourage
regular daily physical activity in their clients.Therefore, as basic research continues to investigate
mechanistic links between diet modifications and physical activity regulation, applied research
is also needed to determine appropriate individual-​based diets that effectively alter regular daily
physical activity in clinical and healthy populations.

Conclusion and future directions

Interactions between genetics and the environment determine regulation of physical activity,
with environmental influences, such as EDCs and diet, being modifiers of an individual’s
phenotype to be physically active or not. That there are known agents, such as diet and EDCs,
which disrupt biological and/​or genetic factors regulating an individual’s drive to be active is a
relatively new concept. To date, the focus on such environmental agents has primarily been on
how each independently contributes to chronic diseases including cardiovascular and metabolic

128
129

Toxicant and dietary exposures

diseases with only a few studies considering how they alter human behaviors such as physical
activity. Existing literature supports that the environment can play a unique role in regulating
physical activity mechanisms; however, the mechanistic pathways through which activity is
regulated are not fully delineated or understood. Additionally, the combined effects of exposure
to poor diet and endocrine disruptors are unknown, but we would hypothesize that they
would likely cause an even greater inhibition of physical activity. Therefore, as scientists, exer-
cise physiologists, and other healthcare providers seek ways to promote increased daily physical
activity, reducing environmental exposures and altering diet should be integral components to
consider as interventions to increase physical activity.

References
1. Ahima RS, and Flier JS. Adipose tissue as an endocrine organ. Trends in Endocrinology & Metabolism
11: 327–​332, 2000.
2. Anway MD, Leathers C, and Skinner MK. Endocrine disruptor vinclozolin induced epigenetic
transgenerational adult-​onset disease. Endocrinology 147: 5515–​5523, 2006.
3. Attina TM, Hauser R, Sathyanarayana S, Hunt PA, Bourguignon J-​P, Myers JP, DiGangi J,
Zoeller RT, and Trasande L. Exposure to endocrine-​disrupting chemicals in the USA: A population-​
based disease burden and cost analysis. The Lancet Diabetes & Endocrinology 4: 996–​1003, 2016.
4. Baranowska B, Baranowska-​Bik A, Bik W, and Martynska L. The role of leptin and orexins in
the dysfunction of hypothalamo-​pituitary-​gonadal regulation and in the mechanism of hyperactivity
in patients with anorexia nervosa. Neuro Endocrinology Letters 29: 37–​40, 2008.
5. Bigsby R, Chapin RE, Daston GP, Davis BJ, Gorski J, Gray LE, Howdeshell KL, Zoeller
RT, and vom Saal FS. Evaluating the effects of endocrine disruptors on endocrine function during
development. Environmental Health Perspectives 107 Suppl. 4: 613–​618, 1999.
6. Bjursell M, Gerdin A-​K, Lelliott CJ, Egecioglu E, Elmgren A, Törnell J, Oscarsson J, and
Bohlooly-​Y M. Acutely reduced locomotor activity is a major contributor to western diet-​induced
obesity in mice. American Journal of Physiology –​Endocrinology and Metabolism 294: E251–​E260, 2008.
7. Bouchard C, Tchernof A, and Tremblay A. Predictors of body composition and body energy
changes in response to chronic overfeeding. International Journal of Obesity 38: 236–​242, 2014.
8. Bowen RS, Ferguson DP, and Lightfoot JT. Effects of aromatase inhibition on the physical
activity levels of male mice. Journal of Steroids & Hormonal Science 1: 1, 2011.
9. Bowen RS, Knab AM, Hamilton AT, McCall JR, Moore-​Harrison TL, and Lightfoot JT.
Effects of supraphysiological doses of sex steroids on wheel running activity in mice. Journal of Steroids
& Hormonal Science 3: 110, 2012.
10. Bymaster FP, Hemrick-​Luecke SK, Perry K, and Fuller R. Neurochemical evidence for antag-
onism by olanzapine of dopamine, serotonin, α1-​adrenergic and muscarinic receptors in vivo in rats.
Psychopharmacology 124: 87–​94, 1996.
11. Bymaster FP, Rasmussen K, Calligaro DO, Nelson DL, DeLapp NW, Wong DT, and Moore
NA. In vitro and in vivo biochemistry of olanzapine: A novel, atypical antipsychotic drug. The Journal
of Clinical Psychiatry 58: 1,478–​436, 1997.
12. Casper RC.The “drive for activity” and “restlessness” in anorexia nervosa: Potential pathways. Journal
of Affective Disorders 92: 99–​107, 2006.
13. Creatore MI, Glazier RH, Moineddin R, Fazli GS, Johns A, Gozdyra P, Matheson FI,
Kaufman-​Shriqui V, Rosella LC, and Manuel DG. Association of neighborhood walkability with
change in overweight, obesity, and diabetes. JAMA 315: 2211–​2220, 2016.
14. Diamanti-​Kandarakis E, Bourguignon JP, Giudice LC, Hauser R, Prins GS, Soto AM,
Zoeller RT, and Gore AC. Endocrine-​ disrupting chemicals: An endocrine society scientific
statement. Endocrine Reviews 30: 293–​342, 2009.
15. Escobar-​Morreale HF, Álvarez-​Blasco F, Botella-​Carretero JI, and Luque-​Ramírez M. The
striking similarities in the metabolic associations of female androgen excess and male androgen defi-
ciency. Human Reproduction 2014.
16. Exner C, Hebebrand J, Remschmidt H, Wewetzer C, Ziegler A, Herpertz S, Schweiger U,
Blum W, Preibisch G, and Heldmaier G. Leptin suppresses semi-​starvation induced hyperactivity
in rats: Implications for anorexia nervosa. Molecular Psychiatry 5: 476–​481, 2000.

129
130

E.E. Schmitt and H.L. Vellers

17. Farooqi IS, Jebb SA, Langmack G, Lawrence E, Cheetham CH, Prentice AM, Hughes IA,
McCamish MA, and O’Rahilly S. Effects of recombinant leptin therapy in a child with congenital
leptin deficiency. New England Journal of Medicine 341: 879–​884, 1999.
18. Ferdinand A, Bisakha S, Rahurkar S, Engler S, and Menachemi N. The relationship between
built environments and physical activity: A systematic review. American Journal of Public Health
102: e7-​e13, 2012.
19. Fernandes Maria Fernanda A, Matthys D, Hryhorczuk C, Sharma S, Mogra S, Alquier T,
and Fulton S. Leptin suppresses the rewarding effects of running via stat3 signaling in dopamine
neurons. Cell Metabolism 22: 741–​749, 2015.
20. Figlewicz D, Evans S, Murphy J, Hoen M, and Baskin D. Expression of receptors for insulin
and leptin in the ventral tegmental area/​ substantia nigra (Vta/​ Sn) of the rat. Brain Research
964: 107–​115, 2003.
21. Flynn KM, Delclos KB, Newbold RR, and Ferguson SA. Behavioral responses of rats exposed to
long-​term dietary vinclozolin. Journal of Agricultural and Food Chemistry 49: 1658–​1665, 2001.
22. Friend DM, Devarakonda K, O’Neal TJ, Skirzewski M, Papazoglou I, Kaplan AR, Liow J-​S,
Guo J, Rane SG, and Rubinstein M. Basal ganglia dysfunction contributes to physical inactivity in
obesity. Cell Metabolism 25: 312–​321, 2017.
23. Frye CA, Bo E, Calamandrei G, Calza L, Dessi-​Fulgheri F, Fernandez M, Fusani L,
Kah O, Kajta M, Le Page Y, Patisaul HB, Venerosi A, Wojtowicz AK, and Panzica GC.
Endocrine disrupters: A review of some sources, effects, and mechanisms of actions on behaviour and
neuroendocrine systems. Journal of Neuroendocrinology 24: 144–​159, 2012.
24. Funkat A, Massa CM, Jovanovska V, Proietto J, and Andrikopoulos S. Metabolic adaptations of
three inbred strains of mice (C57bl/​6, Dba/​2, and 129t2) in response to a high-​fat diet. The Journal of
Nutrition 134: 3264–​3269, 2004.
25. Gorzek JF, Hendrickson KC, Forstner JP, Rixen JL, Moran AL, and Lowe DA. Estradiol and
tamoxifen reverse ovariectomy-​induced physical inactivity in mice. Medicine and Science in Sports and
Exercise 39: 248–​256, 2007.
26. Hebebrand J, Exner C, Hebebrand K, Holtkamp C, Casper R, Remschmidt H,
Herpertz-​Dahlmann B, and Klingenspor M. Hyperactivity in patients with anorexia ner-
vosa and in semistarved rats: evidence for a pivotal role of hypoleptinemia. Physiology & Behavior
79: 25–​37, 2003.
27. Hellwig J, van Ravenzwaay B, Mayer M, and Gembardt C. Pre-​and post-​natal oral toxicity of
vinclozolin in Wistar and Long-​Evans rats. Regulatory Toxicology and Pharmacology 32: 42–​50, 2000.
28. Hillebrand JJ, van Elburg AA, Kas MJ, van Engeland H, and Adan RA. Olanzapine reduces
physical activity in rats exposed to activity-​based anorexia: Possible implications for treatment of
anorexia nervosa? Biological Psychiatry 58: 651–​657, 2005.
29. Holtkamp K, Herpertz-​Dahlmann B, Mika C, Heer M, Heussen N, Fichter M, Herpertz
S, Senf W, Blum WF, and Schweiger U. Elevated physical activity and low leptin levels co-​
occur in patients with anorexia nervosa. The Journal of Clinical Endocrinology & Metabolism
88: 5169–​5174, 2003.
30. Hoskins RG. The effect of castration on voluntary activity. American Journal of Physiology
324–​330, 1925.
31. Jardí F, Laurent MR, Kim N, Khalil R, Bundel D, Eeckhaut A, Helleputte L, Deboel L,
Dubois V, and Schollaert D. Testosterone boosts physical activity in male mice via dopaminergic
pathways. Scientific Reports 8: 957, 2018.
32. Johnson SA, Painter MS, Javurek AB, Ellersieck MR, Wiedmeyer CE, Thyfault JP, and
Rosenfeld CS. Sex-​dependent effects of developmental exposure to bisphenol A and ethinyl estra-
diol on metabolic parameters and voluntary physical activity. Journal of Developmental Origins of Health
and Disease 6: 539–​552, 2015.
33. Joosen AM, Gielen M, Vlietinck R, and Westerterp KR. Genetic analysis of physical activity in
twins. The American Journal of Clinical Nutrition 82: 1253–​1259, 2005.
34. Kajta M, and Wojtowicz AK. Impact of endocrine-​disrupting chemicals on neural development
and the onset of neurological disorders. Pharmacological Reports 65: 1632–​1639, 2013.
35. Kamrin MA. Phthalate risks, phthalate regulation, and public health: A review. Journal of Toxicology
and Environmental Health 12: 157–​174, 2009.
36. Klenotich SJ, Ho EV, McMurray MS, Server CH, and Dulawa SC. Dopamine D2/​3 receptor
antagonism reduces activity-​based anorexia. Transl Psychiatry 5: e613, 2015.

130
131

Toxicant and dietary exposures

37. Kluwe WM, Haseman JK, Douglas JF, and Huff JE.The carcinogenicity of dietary di(2-​ethylhexyl)
phthalate (DEHP) in Fischer 344 Rats and B6c3f1 mice. Journal of Toxicology and Environmental Health
10: 797–​815, 1982.
38. Knab A, Bowen R, Hamilton A, and Lightfoot J. Pharmacological manipulation of the dopamin-
ergic system affects wheel-​running activity in differentially active mice. Journal of Biological Regulators
and Homeostatic Agents 26: 119, 2012.
39. Knab AM, Bowen RS, Hamilton AT, Gulledge AA, and Lightfoot JT. Altered dopaminergic
profiles: Implications for the regulation of voluntary physical activity. Behavioural Brain Research
204: 147–​152, 2009.
40. LaRoche DP, Marques NR, Shumila HN, Logan CR, St Laurent R, and Gonçalves M. Excess
body weight and gait influence energy cost of walking in older adults. Medicine and Science in Sports
and Exercise 47: 1017, 2015.
41. Levine JA, McCrady SK, Lanningham-​Foster LM, Kane PH, Foster RC, and Manohar CU.
The role of free-​living daily walking in human weight gain and obesity. Diabetes 57: 548–​554, 2008.
42. Li DK, Zhou Z, Miao M, He Y, Wang J, Ferber J, Herrinton LJ, Gao E, and Yuan
W. Urine bisphenol-​ A (BPA) level in relation to semen quality. Fertility and Sterility
95: 625–​630.e621–​624, 2011.
43. Licinio J, Caglayan S, Ozata M, Yildiz BO, de Miranda PB, O’Kirwan F, Whitby R, Liang
L, Cohen P, and Bhasin S. Phenotypic effects of leptin replacement on morbid obesity, diabetes
mellitus, hypogonadism, and behavior in leptin-​deficient adults. Proceedings of the National Academy of
Sciences of the United States of America 101: 4531–​4536, 2004.
44. Lightfoot JT. Why control activity? Evolutionary selection pressures affecting the development of
physical activity genetic and biological regulation. BioMed Research International 2013: 821678, 2013.
45. Lightfoot JT, Leamy L, Pomp D, Turner MJ, Fodor AA, Knab A, Bowen RS, Ferguson D,
Moore-​Harrison T, and Hamilton A. Strain screen and haplotype association mapping of wheel
running in inbred mouse strains. Journal of Applied Physiology 109: 623–​634, 2010.
46. Lightfoot JT, Turner MJ, Daves M, Vordermark A, and Kleeberger SR. Genetic influence on
daily wheel running activity level. Physiological Genomics 19: 270–​276, 2004.
47. Markey CM, Luque EH, Munoz De Toro M, Sonnenschein C, and Soto AM. In utero exposure
to bisphenol A alters the development and tissue organization of the mouse mammary gland. Biology
of Reproduction 65: 1215–​1223, 2001.
48. Masuo Y, and Ishido M. Neurotoxicity of endocrine disruptors: Possible involvement in
brain development and neurodegeneration. Journal of Toxicology and Environmental Health, Part B
14: 346–​369, 2011.
49. McGrath LJ, Hopkins WG, and Hinckson EA. Associations of objectively measured built-​
environment attributes with youth moderate–​vigorous physical activity: A systematic review and
meta-​analysis. Sports Medicine 45: 841–​865, 2015.
50. Meek TH, Dlugosz EM, Vu KT, and Garland Jr T. Effects of leptin treatment and western diet on
wheel running in selectively bred high runner mice. Physiology & Behavior 106: 252–​258, 2012.
51. Meek TH, Eisenmann JC, and Garland T. Western diet increases wheel running in mice select-
ively bred for high voluntary wheel running. International Journal of Obesity 34: 960–​969, 2010.
52. Meyer BC, and Weinroth LA. Observations on psychological aspects of anorexia nervosa: Report
of a case. Psychosomatic Medicine 19: 389–​398, 1957.
53. Moore N. Olanzapine: Preclinical pharmacology and recent findings. The British Journal of Psychiatry
41–​44, 1998.
54. Morton GJ, Kaiyala KJ, Fisher JD, Ogimoto K, Schwartz MW, and Wisse BE. Identification
of a physiological role for leptin in the regulation of ambulatory activity and wheel running in mice.
American Journal of Physiology –​Endocrinology and Metabolism 300: E392–​E401, 2011.
55. Parks LG, Ostby JS, Lambright CR, Abbott BD, Klinefelter GR, Barlow NJ, and Gray LE,
Jr. The plasticizer diethylhexyl phthalate induces malformations by decreasing fetal testosterone syn-
thesis during sexual differentiation in the male rat. Toxicological Sciences 58: 339–​349, 2000.
56. Pinson A, Bourguignon JP, and Parent AS. Exposure to endocrine disrupting chemicals and
neurodevelopmental alterations. Andrology 4: 706–​722, 2016.
57. Rendeiro C, Masnik AM, Mun JG, Du K, Clark D, Dilger RN, Dilger AC, and Rhodes JS.
Fructose decreases physical activity and increases body fat without affecting hippocampal neurogen-
esis and learning relative to an isocaloric glucose diet. Scientific Reports 5: 2015.
58. Richter CP. A behavioristic study of the activity of the rat. Comparative Psychology Monographs 1922.

131
132

E.E. Schmitt and H.L. Vellers

59. Roy EJ and Wade GN. Role of estrogens in androgen-​induced spontaneous activity in male rats.
Journal of Comparative Physiological Psychology 89: 573–​579, 1975.
60. Salamone JD. Complex motor and sensorimotor functions of striatal and accumbens dopa-
mine: Involvement in instrumental behavior processes. Psychopharmacology (Berlin) 107: 160–​174, 1992.
61. Schmidt SL, Harmon KA, Sharp TA, Kealey EH, and Bessesen DH. The effects of
overfeeding on spontaneous physical activity in obesity prone and obesity resistant humans. Obesity
20: 2186–​2193, 2012.
62. Schmitt EE, Vellers HL, Porter WW, and Lightfoot JT. Environmental endocrine disruptor affects
voluntary physical activity in mice. Medicine and Science in Sports and Exercise 48: 1251–​1258, 2016.
63. Schotte A, Janssen P, Gommeren W, Luyten W, Van Gompel P, Lesage A, De Loore K, and
Leysen J. Risperidone compared with new and reference antipsychotic drugs: in vitro and in vivo
receptor binding. Psychopharmacology 124: 57–​73, 1996.
64. Serrano SE, Braun J, Trasande L, Dills R, and Sathyanarayana S. Phthalates and diet: A review
of the food monitoring and epidemiology data. Environmental Health 13: 43, 2014.
65. Slonaker JR. The effect of pubescence, oestruation and menopause on the voluntary activity in the
albino rat. American Journal of Physiology 362–​394, 1925.
66. Stubbe JH, Boomsma DI, Vink JM, Cornes BK, Martin NG, Skytthe A, Kyvik KO, Rose RJ,
Kujala UM, and Kaprio J. Genetic influences on exercise participation in 37,051 twin pairs from
seven countries. PloS One 1: e22, 2006.
67. Tanida T, Warita K, Ishihara K, Fukui S, Mitsuhashi T, Sugawara T, Tabuchi Y, Nanmori
T, Qi WM, Inamoto T, Yokoyama T, Kitagawa H, and Hoshi N. Fetal and neonatal exposure
to three typical environmental chemicals with different mechanisms of action: Mixed exposure to
phenol, phthalate, and dioxin cancels the effects of sole exposure on mouse midbrain dopaminergic
nuclei. Toxicology Letters 189: 40–​47, 2009.
68. Tellez LA, Medina S, Han W, Ferreira JG, Licona-​Limon P, Ren X, Lam TT, Schwartz GJ,
and de Araujo IE. A gut lipid messenger links excess dietary fat to dopamine deficiency. Science
341: 800–​802, 2013.
69. Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, and McDowell M. Physical activity in
the United States measured by accelerometer. Medicine and Science in Sports and Exercise 40: 181, 2008.
70. Tucker JM, Tucker LA, LeCheminant J, and Bailey B. Obesity increases risk of declining physical
activity over time in women: A prospective cohort Study. Obesity 21: E715–​E720, 2013.
71. Vellers HL, Letsinger AC, Walker NR, Granados JZ, and Lightfoot JT. High fat high sugar
diet reduces voluntary wheel running in mice independent of sex hormone involvement. Frontiers in
Physiology 8: 2017.
72. Verhagen LA, Luijendijk MC, and Adan RA. Leptin reduces hyperactivity in an animal model for
anorexia nervosa via the ventral tegmental area. European Neuropsychopharmacology 21: 274–​281, 2011.
73. Waller JV, Kaufman RM, and Deutsch F. Anorexia nervosa: A psychosomatic entity. Psychosomatic
Medicine 2: 3–​16, 1940.
74. Wang GH. The relationship between spontaneous activity and oestrous cycle in the rat. Comparative
Psychology Monographs 2: 1–​27, 1923.
75. Witorsch RJ, and Thomas JA. Personal care products and endocrine disruption: A critical review
of the literature. Critical Reviews in Toxicology 40 Suppl 3: 1–​30, 2010.

132
133

SECTION 3

Systems genetics of exercise

endurance and trainability

The previous section explored the systems genetics underlying the trait of physical activity, and
its component parts. This (and the following) section will further explore genetic and molecular
mechanisms driving variation in traits associated with performance –​first for endurance-​related tasks
(Section 3) and then for strength-​related tasks (Section 4). For each of the sections, we have experts
provide some background context and primers associated with the key traits that underlie perform-
ance. Each author was tasked particularly to review traits at baseline (untrained state) and following
acute and chronic exercise training –​which highlights the similarities and differences between basic
human variation and variation in the response to training. As you will have noted from the first part
of the book, there is a lot yet to be learned in sport and exercise systems genetics, so our authors also
highlight gaps in our knowledge and provide directions for future potential studies.
Chapter 11 starts off Section 3 with Dr. David Raichlen and colleagues providing a fan-
tastic overview of human evolution related to the endurance trait. As many readers will have
exercise science backgrounds, having the perspective on the endurance trait from an evolu-
tionary perspective is unique and quite fascinating. Given that endurance and endurance-​related
traits are complex, Drs. Quindry and Roberts provide a primer in Chapter 12, reviewing the
physiological traits that comprise the complex “endurance” phenotype and discuss underlying
molecular drivers of these traits. In Chapters 13 (Dr. Michael Massett and colleagues) and 14
(Drs. Jacob Barber and Mark Sarzynski), our experts discuss the heritability of endurance-​related
phenotypes that have been gleaned so far from animal (Chapter 13) and human (Chapter 14)
research models. Specific genetic contributions to cardiorespiratory fitness are then discussed by
Dr. Reuben Howden and colleagues in Chapter 15. Given that mitochondria play such a crucial
role in energy production, Dr. Mark Tarnopolsky reviews mitochondrial biology in Chapter 16,
and discusses the unique aspects that mitochondrial genetics contribute to endurance-​related
traits, and to overall health and disease.While each chapter in Sections 3 and 4 will explore mul-
tiple pathways and genes, there are a few genetic variations that have been extensively studied
in relation to sport and exercise-​related traits. One of these variations is the common insertion/​
deletion variation in the angiotensin-​converting enzyme (ACE) gene, which is reviewed by
Dr. Linda Pescatello and colleagues in Chapter 17.
Together, the chapters in Section 3 provide a comprehensive review of traits underlying
endurance and endurance-​related traits both at baseline and following training, highlight gaps
in our current knowledge, and provide a solid foundation for future work to complete the
systems models.
133
134
135

11
THE EVOLUTION OF THE
HUMAN ENDURANCE
PHENOTYPE
David A. Raichlen, James T. Webber, and Herman Pontzer

Humans are exceptional endurance athletes. Compared to other quadrupedal mammals, our
ability to engage in anaerobic activity such as sprinting is highly constrained. However, humans
compare favorably to other species in athletic performance at moderate aerobic intensities
(6, 7, 39). Understanding the evolutionary history of our endurance capabilities, along with
adaptations that support human aerobic activity, may provide a foundation for exploring the
underlying systems genetics of endurance sports and aerobic exercise.
In a sense, detailing the genetics of human performance is an evolutionary exploration, and
here we discuss the evidence for how and why human endurance adaptations, and their under-
lying genetic associations, evolved. By doing so, we suggest human evolutionary physiology
can play a key role in studies of exercise, health, and performance. In addition, as described in
more detail below, a systems genetics approach is needed to fully understand the evolution of
the endurance phenotype. Since many endurance adaptations in humans are confined to soft
tissue, which does not fossilize, we must reconstruct our evolutionary history using comparative
physiology. Genomic and genetic approaches can help us detail the specific changes, and timing
of these changes, that have led humans to become the endurance primate.Thus, an evolutionary
approach to human exercise physiology can both aid, and benefit from, genetic and genomic
studies of exercise performance.
To build an evolutionary framework for exercise physiology, we review data from a combin-
ation of methodological approaches including analyses of the human fossil record, comparative
physiology of humans and nonhuman primates, and studies of humans living more traditional
lifestyles (e.g., hunter-​gatherers). By drawing on data from these approaches, we will develop
a model of the human endurance phenotype, a timeline for how and when it evolved, and a
blueprint for how to use this evolutionary framework in future work.

Defining the endurance phenotype

Aerobic activities are those which can be sustained entirely by oxygen-​based metabolism.
The rate of oxygen consumption during an activity, called VO2, is often used as a measure of
energy for aerobic activities. Activity costs tend to increase with body size, and so to account
for differences in body size, energy expenditure during an activity is often expressed as meta-
bolic equivalents (METs), the ratio of activity energy expenditure to basal metabolic rate

135
136

D.A. Raichlen, J.T. Webber, and H. Pontzer

Orangutan

Gorilla
Chimpanzee

LCA Bonobo

Early
Australopithecines Homo

Hu
hominins

m an
Time (millions of years)

3
4

0
7

5
6

1
H. sapiens
Au. sediba
Orrorin tugenensis
Au. afarensis
H. heidelbergensis
Ar.ramisus
H. neandertalensis
Au. africanus
Sahelanthropus
tchadensis H. erectus
Ar. kadaba H. habilis
H. naledi

P.aethiopecus P. boisei H. floresiensis

P. robustus

Figure 11.1 Phylogeny of Hominidae. This figure shows the relationships between extant great apes
including modern humans and bipedal human ancestors. Individual hominin species (ancestors of
modern humans following the divergence from living chimpanzees and bonobos) are shown along with
their known dates of existence. Hominin species are shaded by their major grouping (early hominins,
australopithecines, Homo).

(BMR; the rate of energy expenditure while completely at rest and fasted in a thermoneutral
environment) (15). Thus, energy expenditure during sleep is 1 MET. The metabolic ceiling
for aerobic activities –​VO2max, the maximum rate of oxygen consumption –​is approxi-
mately 20 METs (54) for trained human athletes and is considerably lower (~5–​10 METs)
for untrained or elderly individuals (63). Walking at 5 km per hour is an approximately 3
MET activity, while running at 10 km per hour (~10 minutes per mile) is an approximately
10 MET activity (35).
We can define the human endurance phenotype by traits that improve our ability to sustain
moderate to vigorous physical activity (MVPA; aerobic intensities between 3 and 10 METs) for
periods greater than 20 minutes (see Chapter 12 in this volume for more detailed discussion
of the human endurance phenotype). While endurance activities in traditional human societies
often include endeavors other than walking and running (e.g., digging to extract wild tubers
for food) (32), locomotion has been the focus of most research into the evolution of the human
endurance phenotype (6, 7). For example, in humans, running distances greater than about
5 km requires sustained aerobic energy expenditure and is considered an endurance activity
(26). Among extant primates, humans are the only taxon that regularly engages in endurance
locomotor activities, suggesting selection for the endurance phenotype occurred sometime
after the divergence of the human lineage from nonhuman apes (6) (Figure 11.1). Tracking an
evolutionary increase in endurance capabilities allows us to identify the specific pressures that
generated humans’ unique athletic abilities.

136
137

Human endurance phenotype evolution

Here, we focus on the evolution of locomotor-​related traits that we can track through human
evolution (i.e., in fossil taxa or through comparative physiology), dividing these traits into three
adaptive categories: 1) locomotor energetics, 2) force resistance/​stabilization required of high
locomotor speeds, and 3) endurance (the ability to sustain locomotion over long distances).
In defining the endurance phenotype this way, we are careful not to conflate adaptations that
reduce energy costs of locomotion with traits that enhance endurance (see 39 for review and
discussion).To fully understand how natural selection led to human endurance performance, we
start by comparing human aerobic capabilities with those of our closest living relatives, and then
discuss the underlying traits that explain taxonomic differences.

Comparative approach: Physical activity in extant apes and humans

Comparative biology offers a valuable methodological approach to examine evolutionary
physiology in living taxa. By comparing performance or morphology in humans and our closest
living relatives, the great apes (Figure 11.1), we can better understand how, and potentially when,
major changes in evolutionary physiology occurred. We last shared a common ancestor with
the genus Pan (living chimpanzees and bonobos) sometime between 6 and 8 million years ago.
Any differences in endurance capabilities between humans and the Pan lineage should reflect
either selection for increased endurance along the human lineage, or selection for decreased
endurance along the chimpanzee and bonobo lineage. Thus, we can model changes in physical
activity patterns across human evolutionary history by examining morphology and behavior in
these taxa.
Field studies have tracked daily movement patterns in each of the three genera of great apes,
providing a window into activity patterns of our last common ancestor with chimpanzees (see
Figure 11.2). Studies of chimpanzees in the wild suggest they travel between 2 and 5 km per
day on the ground, complemented by about 100 m/​day of vertical climbing (1, 39, 41, 45).
Although fewer data exist, bonobos seem to travel less than chimpanzees, with daily travel
distances averaging 1.4 km/​day (3). For the other great apes, movement is significantly shorter,
with gorillas averaging 0.4–​2.6 km/​day (66), and orangutans averaging about 1 km/​day (11, 18,
56). Even less is known about the aerobic intensities of physical activity patterns in the great
apes. Data collected on walking velocities in chimpanzees suggest that most of their daily travel
occurs at low intensities (2). For example, average walking speeds for males (1.94 km/​hour) and

Orangutan

Gorilla

Chimpanzee

Human

0 5 10 15 20 25
Range of daily travel distances (km)

Figure 11.2 Range of daily walking distances for great apes. Human daily walking distances are longer
than any of the other great apes, suggesting the need for skeletal and physiological adaptations supporting
aerobic endurance. Estimates of walking distances for each species are taken from the literature. See text
for further details and references.

137
138

D.A. Raichlen, J.T. Webber, and H. Pontzer

females (2.20 km/​hour) correspond to low aerobic intensities in humans walking at comparable
speeds (<3 METs) (15). Interspecific differences in BMR and locomotor costs complicate this
comparison of aerobic intensities in humans and chimpanzees, however given slow speeds used
in the wild, it is unlikely chimpanzees engage in moderate intensity activity for any sustained
period of time.Thus, despite limited data, the current evidence suggests living great apes engage
in short daily travel at relatively slow speeds.
Comparing great ape activity to modern humans is complicated by changes in human
lifestyles in industrialized societies that greatly reduce physical activity. However, there are
extant human populations living lifestyles similar in some ways to those of our recent ancestors
(e.g., hunter-​gatherer populations), and we can use these groups as windows into physical
activity patterns in the past. While the great apes seem to live relatively sedentary lifestyles,
humans living in small-​scale societies are highly active for large portions of their day (13, 46).
For example, detailed recent studies of physical activity patterns in the Hadza, a hunting and
gathering population living in northern Tanzania, suggest that high activity levels are common
across the lifespan (30, 43, 44, 51). Hadza adults travel between 6 and 12 km/​day; however, there
is a high degree of variance, and adult males are known to travel over 20 km in a single day
during hunting forays (43). In other hunter-​gatherer populations, daily travel distances average
9.5–​14.1 km per day across a wide range of diverse habitats (i.e., rain forests and open savannahs)
(21, 31).
While intensities used during daily foraging vary considerably, we recently showed that
Hadza adults spend about 75 minutes per day in moderate to vigorous aerobic intensities (i.e.,
> 3 METs). Although similarly detailed data are not available for other hunting and gathering
groups, estimates of physical activity in these lifestyles are generally high, with foraging
populations regularly engaging in more MVPA each day than people in industrialized soci-
eties perform in 1 week (13, 51). However, despite long periods of time spent in MVPA, living
hunter-​gatherers rarely engage in highly vigorous activity (52). Intensities that are similar to
those experienced during long-​distance running (10–​14 METs) are performed for less than 10
minutes per day in Hadza hunter-​gatherers. During hunting in other groups (i.e.!Kung, Ache),
individuals walk at moderate to high speeds for long distances (~3 METs), punctuated only
by short sprints to capture prey (14, 20; see also Chapter 12), suggesting that across hunter-​
gatherers, moderate aerobic intensities dominate their physical activities.
Thus, compared to great apes, human hunter-​gatherers travel longer distances, and do so at
higher intensity levels, however most of this movement occurs in the moderate range of aerobic
intensities (3–​6 METs). If living hunter-​gatherers provide a window into activity levels in our
earlier ancestors, the endurance phenotype for humans should be associated with adaptations
that improve both endurance and sustained moderate intensity locomotion compared to
morphology seen in the great apes.

Morphological approach: Analyses of the human fossil record

The previous discussion of comparative activity patterns suggests a fundamental difference
between nonhuman ape-​like lifestyles, and the human hunting and gathering lifestyle. A key
element of reconstructing the evolution of human endurance activity is detailing how and
when lifestyles in human ancestors shifted from more ape-​ like to more human-​ like, and
specifically when hunting and gathering became the dominant human subsistence strategy.
Combining data from human fossils with paleoecological data, we can reconstruct how and
why hominin lifestyles underwent changes that altered physical activity patterns compared to
our last common ancestor with the chimpanzee lineage. For the purposes of reconstructing

138
139

Human endurance phenotype evolution

changes in activity levels and endurance capabilities, we suggest the hominin record can be
divided roughly into three key groups (see Figure 11.1): 1) the earliest hominins, 2) the genus
Australopithecus, and 3) the genus Homo.

Ecological pressures and changing lifestyles in human evolution

Reconstructions of locomotion and behavior in the earliest hominins generally suggest that,
although they walked bipedally, in many respects they more closely resemble nonhuman great
apes in activity patterns and behavior (37). Because of this similarity, we generally assume the
last common ancestor of humans and Pan (chimpanzees and bonobos) was more Pan-​like
(37). Further, as discussed above, all of the living apes share rather limited travel distances
and speeds, suggesting a low-​endurance phenotype was common among Miocene apes from
which the human lineage emerged. Thus, changes in endurance capabilities along the human
lineage are likely due to selection for increased time spent in moderate to vigorous aerobic
intensities in hominins, rather than a decrease in endurance capabilities in the chimpanzee/​
bonobo lineage.
There is little known about the endurance capabilities of the earliest hominins, Sahelanthropus
tchadensis, Orrorin tugenensis, and Ardipithecus. However, the retention of a grasping foot and
other ape-​like elements of the locomotor skeleton suggest the endurance capabilities of early
hominins were no better than those of living apes. Indeed, the adoption of a bipedal (two-​
legged) walking gait in early hominins may have diminished endurance capabilities, at least ini-
tially, by removing the forelimbs from their role in supporting body weight and thus requiring
more work from the hindlimb muscles (39).
The next major group of hominins, members of the genus Australopithecus (as well as the
closely related Paranthropus group), show changes in the locomotor skeleton that suggest
improved energy economy during walking (see ‘Skeletal adaptations to high-​speed, long-​
distance locomotion’). However, the Australopithecus postcranial skeleton contains a mix of traits
that resemble both modern humans and living apes, including skeletal features of the upper limb
and shoulder that suggest continued use of arboreal resources (12, 62, 65). Australopithecines
have craniodental adaptations that likely supported omnivorous diets comprised primarily of
fruits, and therefore a reliance on resources found mainly in forested habitats. During this time
period, the earth’s climate began a cooling trend that led to increasingly open, less heavily
forested habitats (52). For a taxon that was reliant on forest-​based resources for food, these
habitat changes likely required increased daily walking distances, providing a selection pressure
for increased energy economy. However, there is little evidence that Australopithecus species
shifted to a meat-​based diet, which would have necessitated a hunting and gathering lifestyle
associated with high aerobic endurance requirements.
Beginning around 2.0 million years ago, with the early evolution of the genus Homo, there
is a marked change in behavior that suggests hunting and gathering became an integral part of
the hominin lifestyle. At this time, members of the genus Homo increased their reliance on meat,
with evidence that there was some form of systematic hunting or power scavenging (stealing
carcasses from primary hunters), since fossil prey animals found at sites with stone tools suggest
our ancestors were gaining access to large sources of meat (9). As part of the development of
the hunting and gathering lifestyle, ancestral hominins became central place foragers, which
required a return to a home base following foraging bouts (30). Central place foraging allows
for food sharing in a lifestyle dominated by variance in procurement across individuals, however,
long-​distance foraging bouts necessitate higher speeds so that individuals can return to the cen-
tral place.Thus, we suggest the endurance phenotype is best understood as a suite of adaptations

139
140

D.A. Raichlen, J.T. Webber, and H. Pontzer

to movement at high speeds over long distances associated with this emergence of hunting and
gathering in the genus Homo.

Skeletal adaptations to high-​speed, long-​distance locomotion

Reconstructing lifestyle transitions, and how they impacted endurance capabilities during
human evolution, relies on analyses and interpretations of fossil skeletons. Although there
remains a lively debate over how best to reconstruct the evolutionary links between skeletal
morphology and locomotor performance, we believe there is an emerging consensus that the
ecological and lifestyle changes described above led to changes in the hominin musculoskeletal
system that supported a major transition in physical activity. In the following subsection, we
detail these morphological changes as they relate to three key aspects of the endurance pheno-
type: 1) energetics, 2) force resistance/​stabilization, and 3) endurance.

Skeletal adaptations for energy economy

Several skeletal features suggest changes from the earliest hominins to the origins of the genus
Homo are associated with selection to improve locomotor economy following the early evolu-
tion of bipedalism. The earliest hominins, including Sahelanthropus tchadensis, Orrorin tugenensis,
and Ardipithecus, likely walked with energy economy more similar to those of living great
apes, and substantially higher energy costs than living humans (39, 57). These reconstructions
are based mainly on the functional morphology of the pelvis and hindlimb which suggests
short limbs combined with more flexed limb postures (19), leading to large volumes of muscle
activated to support the body during each step.
However, beginning with the evolution of Australopithecus, we see changes to the lower
limb that indicate an improvement in locomotor economy. For example, the lower limbs of
australopiths are significantly longer than those of other great apes and earlier hominins (39).
Long limbs reduce energy costs by smoothing the path of the center of mass during walking,
minimizing the amount of muscle volume needed per step, and reducing the number of steps
taken per unit distance (38–​40, 42). In addition, australopiths had pelvic anatomy consistent with
full extension of the hip during locomotion, which would have greatly reduced energy costs
of walking compared with the crouched postures of chimpanzees (10, 19, 57). This increased
hip and knee extension in australopiths is confirmed by recent analyses of the fossil footprints
from Laetoli, Tanzania showing that, by at least 3.66 million years ago, australopiths walked
with mechanics generally similar to modern humans (49, 50). Given the inverse correlation
between limb length and energy costs of locomotion at both walking and running speeds, and
the reduced costs associated with an extended hip posture, selection for improved locomotor
economy was likely acting by 3–​4 million years ago. Thus, improved energy economy during
walking in australopiths may have laid the foundation for long-​distance foraging bouts in early
hunting and gathering members of the genus Homo.

Skeletal adaptations for high forces

While improved energy economy was beneficial at both low and high speeds of movement,
higher walking and running speeds bring other challenges. Specifically, higher speed movement
increases forces acting on the musculoskeletal system (8, 25, 28) and we should expect to see
adaptations to force resistance in fossil taxa moving at high speeds. In addition, high forces, espe-
cially at initial ground contact, generate large pitching moments at both the waist and neck, and

140
141

Human endurance phenotype evolution

if these are not resisted in some way, high-​speed motion would lead to potentially dangerous
instability.
Beginning with Homo erectus, hominin postcranial evolution is dominated by adaptations
for force resistance and stabilization mechanisms. Early members of the genus Homo have
increased lower limb joint surface areas, which are linked to resisting higher loading envir-
onments (16, 53). In comparisons with earlier hominins such as Australopithecus afarensis and
A. africanus, after taking into account differences in body size, H. erectus has increased femoral
head and tibial plateau breadth, and decreased humeral relative to femoral diaphyseal bone
strength compared with earlier hominins (16, 52). These changes in joint articular surface size
in the lower limb, combined with a shift in patterns of diaphyseal strength in the upper and
lower limb bones, suggests that H. erectus was likely managing higher lower limb loads during
locomotion.
Higher loads due to increased ground reaction forces often lead to pitching moments in both
the trunk and head in bipeds, which would require associated adaptations to maintain stability
(27). H. erectus and modern humans show changes in two areas associated with maintaining
stability during bipedal locomotion. First, pitching moments at the trunk are counteracted
by activation of the gluteus maximus muscle (27, 61). Modern humans have an enlarged glu-
teus maximus with an increased attachment area on the superior portion of the iliac blade
compared with other great apes (27). Evidence from the fossil record suggests this change in
muscular morphology likely occurred with the evolution of H. erectus, based on comparisons of
attachment site morphology on the pelvis (27).
In addition to managing trunk pitch, head pitch in response to increased ground reac-
tion forces may cause disruptions to the visual system that can be dangerous when moving at
high speeds on a complex landscape (58). To manage high-​pitch velocities, the vestibulo-​ocular
system must respond to head movements and adjust eye position to allow the maintenance
of vision (59, 60). Across mammals, increasing the radius of the semicircular canals is often
considered an adaptation to high head velocities, minimizing the potential for saturation of the
vestibulo-​ocular system (59, 60). Spoor et al. (59, 60) have shown that H. erectus and modern
humans have increased diameters of the posterior and anterior semicircular canals compared
with australopiths and other great apes.
Combined, a change in lower limb joint surface areas, a shift towards stronger lower limb
compared with upper limb long bones, increased gluteus maximus size, and enlarged semi-
circular canals suggest that H. erectus was adapted to a locomotor environment that included
increased loads potentially over long periods of time. These increased loads are likely linked to
increased walking and running speeds consistent with requirements of a hunting and gathering
lifestyle. Thus, by the evolution of the genus Homo, we see skeletal adaptations that are linked to
an increased reliance on higher aerobic intensity foraging behaviors.

Physiological adaptations for endurance

Evidence presented above suggests the human locomotor phenotype includes adaptations for
higher loads during locomotion, which likely come with increased velocity, and reduced loco-
motor economy, and could be advantageous for moving over long distances.While these skeletal
adaptations reflect the potential for a behavioral transition towards increased endurance, the
main elements that make up the endurance phenotype lie in soft tissue that does not preserve
in the fossil record (39). Using comparative biology, we can assess differences in endurance traits
that do not fossilize; however, a detailed exploration of the evolution of the endurance pheno-
type requires analysis of how the underlying genetics of these traits shifted over time.

141
142

D.A. Raichlen, J.T. Webber, and H. Pontzer

A key element of endurance is muscle fiber type distribution that impacts both mitochon-
drial density and fatigue resistance (5, 33, 34). Recent work shows that human lower limb
muscles are biased towards myosin heavy chain (MHC) type I fibers, whereas chimpanzees
and other primates have more mixed lower limb muscle fiber distributions. MHC I fibers have
a higher density of mitochondria, relying less on glycolysis, and therefore muscle glycogen,
allowing for an increase in the number of contractions possible prior to onset of fatigue (55).
In addition, humans have greater hindlimb muscle mass compared to chimpanzees (36, 39, 64),
which would increase muscle endurance given the greater density of mitochondria predicted
for muscles with a larger percentage of type I fibers.
Fiber-​type distributions are at least partially a product of underlying genetics. For example,
the alpha-​actinin-​3 (ACTN3) gene is linked with fiber-​type distribution, and individuals
possessing the 577X allele have a significantly higher density of type I fibers compared with
individuals possessing the 577R allele (67). In fact, elite endurance athletes include a higher
proportion of individuals homozygous for the 577X allele compared with nonathletes and
power-​oriented athletes (68). MacArthur et al. (29) show that this endurance-​linked mutation
likely arose between 40,000 and 60,000 years ago, during the evolution of modern H. sapiens,
suggesting increased selection pressures for improved endurance in populations of our own
species. Examinations of the underlying genetics behind these muscle variables may provide
further evidence of adaptive changes earlier in human evolution.
In addition to muscular traits that improve human endurance, humans also possess
thermoregulatory adaptations that some suggest evolved to facilitate endurance activities
in environments that would generate high heat stress (e.g., African savanna (23), but see
(4) for an alternative view). For example, while all Old World monkeys and apes sweat,
humans have a much higher percentage of eccrine subcutaneous sweat glands, and have
much higher sweat rates than other primates, including apes (7, 23). Combined with a loss
of dense fur, these adaptations would allow for enhanced thermoregulatory capabilities in
hot environments during endurance activities. A recent study demonstrated that the density
of eccrine sweat glands is under partial genetic control, through the engrailed-​1 (En1) gene,
which may provide a novel basis for exploring the evolution of human thermoregulatory
capabilities (17).
While analyses of ancestral anatomy detail a shift in lifestyle that likely required enhanced
endurance, soft tissue evidence of changing endurance capabilities does not fossilize. It is likely
that changes in mitochondrial density, muscle fiber-​type distributions, and thermoregulatory
abilities shifted with the evolution of the genus Homo. Future work in comparative genetics and
genomics can help evolutionary biologists clarify the timing of these adaptive changes in endur-
ance, and will help fill in gaps in the story of the evolution of the human endurance phenotype.

An evolutionary framework for the endurance phenotype

Our review of evidence for changes in locomotion and endurance during human evolution
suggests that a major transition occurred beginning around 2 million years ago. At this time,
with the earth’s climate changing, and associated increases in grasslands and decreases in forest
cover in east and south Africa, ancestral humans adopted a hunting and gathering lifestyle. From
comparative studies of living humans and great apes, we showed that this lifestyle change would
have involved an increase in travel distance and locomotor intensity (predominantly moderate
intensity physical activity). The hominin fossil record provides evidence for selection acting on
the locomotor skeleton to support this type of behavioral change. While selection for improved

142
143

Human endurance phenotype evolution

locomotor economy seems to have occurred prior to the adoption of hunting and gathering
(i.e., in Australopithecus), the evolution of the genus Homo is linked with adaptations to resist
high locomotor forces and maintain stability at higher locomotor speeds. Because adaptations
for endurance do not fossilize, it is difficult to know with certainty when traits associated with
aerobic capacity evolved; however, a plausible hypothesis is that an increase in the density of
slow-​twitch muscle fibers in the lower limb, and an associated increase in mitochondrial density,
was selected during the evolution of the genus Homo.

Evolutionary exercise physiology and human health

Based on this proposed evolutionary shift towards increased endurance activities, we can explore
the importance of an evolutionary framework for understanding the genetic links between
exercise (or lack thereof) and disease. Fundamental to this discussion is the understanding that
selection acted on human physiology to enhance reproductive success, rather than to improve
health (e.g., 24). Thus, any health-​enhancing effects of exercise that are associated with our
evolutionary history are linked to the reproductive fitness benefits of engaging in aerobic
activity.
Recently, Lieberman (24) and Raichlen and Alexander (46) detailed models for evolu-
tionary exercise physiology that suggest physiological systems evolved to respond plastically
to exercise stress. When faced with increased demands, systems from our musculoskeletal
system to our cardiovascular system and even our brains, respond by increasing capacity to
adapt to new stress. When exercise demands are removed, an energy-​saving mechanism is
initiated that leads to system atrophy, reducing the energy costs of tissue maintenance. Given
our reliance on physical activity and endurance behaviors to gain access to food, our physio-
logical systems do not function well with capacity reductions, often leading to negative health
outcomes.
Well-​known examples of these exercise-​induced capacity adaptations include increased per-
ipheral capillary density, increased heart mass, and increased cardiac stroke volume in individuals
engaged in aerobic exercise (46). Bone mineral density is increased through exercise, especially
at young ages, which can lead to a higher overall peak bone mass and reduced risk of developing
osteoporosis later in life (22). In addition, reductions in activity are associated with loss of muscle
mass and strength (sarcopenia) (22).
Finally, endurance exercise has beneficial effects on the brain, with reduced age-​related
atrophy and increased functional connectivity at young ages (46–​48). In fact, it is possible that
evolutionary changes in endurance behaviors may have altered brain aging early on during
the adoption of hunting and gathering (47). At this time, nearly 2 million years ago, human
ancestors were all homozygous for apolipoprotein E (APOE) ε4, an allele that carries increased
risk of cardiovascular disease and Alzheimer’s disease. However, individuals who exercise experi-
ence large risk reductions for each of these diseases. Raichlen and Alexander (47) argue that
the adoption of increased endurance-​related activities may have allowed for an increase in the
human lifespan in the context of genetic constraints such as APOE-​ε4.
This evolutionary approach to exercise physiology helps us understand how and why our
physiological systems respond to exercise. As a foundation for future studies, understanding the
role of intensity and exercise duration on capacity changes in physiology may help us better
determine optimal exercise patterns for maintaining the health of physiological systems. In
addition, knowledge of ancestral genotypes, and their interactions with physical activity, may
provide new avenues for examining precision exercise studies from an evolutionary perspective.

143
144

D.A. Raichlen, J.T. Webber, and H. Pontzer

Conclusions: What can genetics bring to the evolutionary

story and vice versa?
Here, we have argued that human evolution was driven, in many ways, by a shift from a
relatively sedentary lifestyle to an increase in endurance activities likely tied to a chan-
ging climate and the associated adoption of a hunting and gathering lifestyle. Adaptations
that appear during the evolution of H. erectus, beginning about 2 million years ago, suggest
major changes in the endurance capabilities of these human ancestors that persist in modern
H. sapiens. Thus, we owe our modern endurance athletic capabilities to key transitions
during human evolution.
While the human fossil record provides clues to the evolution of the human endurance
phenotype, a better understanding of the genetic and epigenetic underpinnings of the endur-
ance phenotype will not only help improve health and athletic performance, but will enhance
our understanding of human evolution. Thus, we suggest that analyses of human performance
genetics can both benefit from an evolutionary approach and can improve our understanding
of human evolutionary history. For example, testing the hypothesis that increased mitochon-
drial density evolved with the adoption of hunting and gathering in the genus Homo could
benefit from a directed effort to examine the history of genes associated with muscle fiber
type and mitochondrial density. A better understanding of the evolution of genes contrib-
uting to the endurance phenotype, and especially the timing and geographical variation of
their evolutionary history, will help us more clearly identify the selection pressures that led to
human endurance capabilities. By combining data from the fossil record with a systems genetic
approach to human exercise performance, we can, perhaps, use an evolutionary framework to
improve sports performance and health.

References
1. Amsler SJ. Energetic costs of territorial boundary patrols by wild chimpanzees. Am J Primatol
72: 93–​103, 2010.
2. Bates LA and Byrne RW. Sex differences in the movement patterns of free-​ranging chimpanzees
(Pan troglodytes schweinfurthii): foraging and border checking. Behav Ecol Sociobiol 64: 247–​255, 2009.
3. Beaune D, Bretagnolle F, Bollache L, Hohmann G, Surbeck M, Bourson C, and Fruth B.The
Bonobo–​Dialium positive interactions: seed dispersal mutualism. Am J Primatol 75: 394–​403, 2013.
4. Best A and Kamilar JM. The evolution of eccrine sweat glands in human and nonhuman primates.
J Hum Evol 117: 33–​43, 2018.
5. Bozek K, Wei Y, Yan Z, Liu X, Xiong J, Sugimoto M, Tomita M, Pääbo S, Pieszek R,
and Sherwood CC. Exceptional evolutionary divergence of human muscle and brain metabolomes
parallels human cognitive and physical uniqueness. PLoS Biol 12: e1001871, 2014.
6. Bramble DM and Lieberman DE. Endurance running and the evolution of Homo. Nature
432: 345–​352, 2004.
7. Carrier DR. The energetic paradox of human running and hominid evolution. Curr Anthropol
25: 483–​495, 1984.
8. Daoud AI, Geissler GJ, Wang F, Saretsky J, Daoud YA, and Lieberman DE. Foot strike and
injury rates in endurance runners: A retrospective study. Med Sci Sports Exerc 44: 1325–​1334, 2012.
9. Domínguez-​Rodrigo M. Hunting and scavenging by early humans: The state of the debate. J World
Prehist 16: 1–​54, 2002.
10. Foster AD, Raichlen DA, and Pontzer H. Muscle force production during bent-​knee, bent-​hip
walking in humans. J Hum Evol 65: 294–​302, 2013.
11. Galdikas BM. Orangutan diet, range, and activity at Tanjung Puting, Central Borneo. Int J Primatol
9: 1–​35, 1988.
12. Green DJ and Alemseged Z. Australopithecus afarensis scapular ontogeny, function, and the role of
climbing in human evolution. Science 338: 514–​517, 2012.

144
145

Human endurance phenotype evolution

13. Gurven M, Jaeggi AV, Kaplan H, and Cummings D. Physical activity and modernization among
Bolivian Amerindians. PLoS One 8: e55679, 2013.
14. Hill K and Hurtado AM. Hunter-​gatherers of the New World. Am Sci 77: 436–​443, 1989.
15. Jette M, Sidney K, and Blümchen G. Metabolic equivalents (METS) in exercise testing, exercise
prescription, and evaluation of functional capacity. Clin Cardiol 13: 555–​565, 1990.
16. Jungers WL. Lucy’s limbs: skeletal allometry and locomotion in Australopithecus afarensis. Nature
297: 676–​678, 1982.
17. Kamberov YG, Karlsson EK, Kamberova GL, Lieberman DE, Sabeti PC, Morgan BA, and
Tabin CJ. A genetic basis of variation in eccrine sweat gland and hair follicle density. Proc Natl Acad
Sci U S A 112: 9932–​9937, 2015.
18. Knott C, Beaudrot L, Snaith T, White S, Tschauner H, and Planansky G. Female-​female com-
petition in Bornean orangutans. Int J Primatol 29: 975–​997, 2008.
19. Kozma EE, Webb NM, Harcourt-​Smith WEH, Raichlen DA, D’Aout K, Brown MH,
Firestone E, Ross SR, Aerts P, and Pontzer H. Hip extensor mechanics and the evolution of
walking and climbing capabilities in humans, apes, and fossil hominins. Proc Natl Acad Sci U S A
115: 4134–​4139, 2018.
20. Lee RB. Kung Bushmen subsistence: An input-​ output analysis. In: Contributions to
Anthropology: Ecological Essays, edited by Damas D. Ottowa: National Museums of Canada, 1969,
p. 73–​94.
21. Leonard WR and Robertson ML. Comparative primate energetics and hominid evolution. Am J
Phys Anthropol 102: 265–​281, 1997.
22. Lieberman D. The Story of the Human body: Evolution, Health, and Disease. New York: Vintage, 2014.
23. Lieberman DE. Human locomotion and heat loss: An evolutionary perspective. Compr Physiol
5: 99–​117, 2015.
24. Lieberman DE. Is exercise really medicine? An evolutionary perspective. Curr Sports Med Rep
14: 313–​319, 2015.
25. Lieberman DE.What we can learn about running from barefoot running: An evolutionary medical
perspective. Exerc Sport Sci Rev 40: 63–​72, 2012.
26. Lieberman DE and Bramble DM.The evolution of marathon running. Sports Med 37: 288–​290, 2007.
27. Lieberman DE, Raichlen DA, Pontzer H, Bramble DM, and Cutright-​Smith E. The human
gluteus maximus and its role in running. J Exp Biol 209: 2143–​2155, 2006.
28. Lieberman DE, Venkadesan M, Werbel WA, Daoud AI, D’Andrea S, Davis IS, Mang’Eni
RO, and Pitsiladis Y. Foot strike patterns and collision forces in habitually barefoot versus shod
runners. Nature 463: 531, 2010.
29. MacArthur DG, Seto JT, Raftery JM, Quinlan KG, Huttley GA, Hook JW, Lemckert FA,
Kee AJ, Edwards MR, and Berman Y. Loss of ACTN3 gene function alters mouse muscle metab-
olism and shows evidence of positive selection in humans. Nat Genet 39: 1261, 2007.
30. Marlowe F. The Hadza: Hunter-​ Gatherers of Tanzania. Berkeley, CA: University of California
Press, 2010.
31. Marlowe FW. Hunter-​gatherers and human evolution. Evol Anthropol 14: 54–​67, 2005.
32. Marlowe FW and Berbesque JC. Tubers as fallback foods and their impact on Hadza hunter-​
gatherers. Am J Phys Anthropol 140: 751–​758, 2009.
33. Myatt JP, Schilling N, and Thorpe SK. Distribution patterns of fibre types in the triceps surae
muscle group of chimpanzees and orangutans. J Anat 218: 402–​412, 2011.
34. O’Neill MC, Umberger BR, Holowka NB, Larson SG, and Reiser PJ. Chimpanzee super
strength and human skeletal muscle evolution. Proc Natl Acad Sci U S A 114: 7343–​7734, 2017.
35. Office of Disease Prevention and Health Promotion. Appendix 1. Translating scientific evi-
dence about total amount and intensity of physical activity into guidelines. 2018. www.health.gov/​
paguidelines/​guidelines/​appendix1.aspx.
36. Payne RC, Crompton RH, Isler K, Savage R, Vereecke EE, Günther MM, Thorpe S, and
D’Août K. Morphological analysis of the hindlimb in apes and humans. I. Muscle architecture. J Anat
208: 709–​724, 2006.
37. Pilbeam DT and Lieberman DE. Reconstructing the last common ancestor of chimpanzees and
humans. In: Chimpanzees and Human Evolution, edited by Muller MN, Wrangham RW, and Pilbeam
DT. Cambridge, MA: The Belknap Press of Harvard University Press, 2017, p. 22–​141.
38. Pontzer H. Ecological energetics in early Homo. Curr Anthropol 53: S346–​S358, 2012.
39. Pontzer H. Economy and endurance in human evolution. Curr Biol 27: R613–​R621, 2017.

145
146

D.A. Raichlen, J.T. Webber, and H. Pontzer

40. Pontzer H. Effective limb length and the scaling of locomotor cost in terrestrial animals. J Exp Biol
210: 1752–​1761, 2007.
41. Pontzer H. Locomotor ecology and evolution in chimpanzees and humans. In: Chimpanzees and
Human Evolution, edited by Muller MN, Wrangham RW, and Pilbeam DT. Cambridge, MA: The
Belknap Press of Harvard University Press, 2017, p. 259–​285.
42. Pontzer H. Predicting the energy cost of terrestrial locomotion: A test of the LiMb model in
humans and quadrupeds. J Exp Biol 210: 484–​494, 2007.
43. Pontzer H, Raichlen DA, Wood BM, Emery Thompson M, Racette SB, Mabulla AZ, and
Marlowe FW. Energy expenditure and activity among Hadza hunter-​gatherers. Am J Hum Biol
27: 628–​637, 2015.
44. Pontzer H, Raichlen DA, Wood BM, Mabulla AZ, Racette SB, and Marlowe FW. Hunter-​
gatherer energetics and human obesity. PLoS One 7: e40503, 2012.
45. Pontzer H and Wrangham RW. Climbing and the daily energy cost of locomotion in wild
chimpanzees: implications for hominoid locomotor evolution. J Hum Evol 46: 315–​333, 2004.
46. Raichlen DA and Alexander GE. Adaptive capacity: An evolutionary neuroscience model linking
exercise, cognition, and brain health. Trends Neurosci 40: 408–​421, 2017.
47. Raichlen DA and Alexander GE. Exercise, APOE genotype, and the evolution of the human life-
span. Trends Neurosci 37: 247–​255, 2014.
48. Raichlen DA, Bharadwaj PK, Fitzhugh MC, Haws KA, Torre G-​A, Trouard TP, and
Alexander GE. Differences in resting state functional connectivity between young adult endurance
athletes and healthy controls. Front Hum Neurosci 10: 610, 2016.
49. Raichlen DA and Gordon AD. Interpretation of footprints from Site S confirms human-​like
bipedal biomechanics in Laetoli hominins. J Hum Evol 107: 134–​138, 2017.
50. Raichlen DA, Gordon AD, Harcourt-​Smith WE, Foster AD, and Haas Jr WR. Laetoli footprints
preserve earliest direct evidence of human-​like bipedal biomechanics. PLoS One 5: e9769, 2010.
51. Raichlen DA, Pontzer H, Harris JA, Mabulla AZ, Marlowe FW, Josh Snodgrass J, Eick G,
Colette Berbesque J, Sancilio A, and Wood BM. Physical activity patterns and biomarkers of car-
diovascular disease risk in hunter-​gatherers. Am J Hum Biol 29: e22919, 2017.
52. Rowan J and Reed KE. The paleoclimatic record and Plio-​ Pleistocene paleoenvironments.
In: Handbook of Paleoanthropology, edited by Henke W and Tattersall I. New York: Springer, 2015,
p. 465–​491.
53. Ruff CB, Burgess ML, Ketcham RA, and Kappelman J. Limb bone structural proportions and
locomotor behavior in AL 288–​1 (“Lucy”). PLoS One 11: e0166095, 2016.
54. Scheadler CM and Devor ST. VO2max measured with a self-​selected work rate protocol on an
automated treadmill. Med Sci Sports Exerc 47: 2158–​2165, 2015.
55. Schiaffino S and Reggiani C. Fiber types in mammalian skeletal muscles. Physiol Rev
91: 1447–​1531, 2011.
56. Singleton I, Knott CD, Morrogh-​Bernard HC, Wich SA, and van Schaik CP. Ranging behav-
iour of orangutan females and social organization. In: Orangutans: Geographic Variation in Behavioral
Ecology and Conservation, edited by Wich SA, Utami Atmoko SS, Setia TM, and van Schaik CP.
Oxford: Oxford University Press, 2009, p. 205–​213.
57. Sockol MD, Raichlen DA, and Pontzer H. Chimpanzee locomotor energetics and the origin of
human bipedalism. Proc Natl Acad Sci U S A 104: 12265–​12269, 2007.
58. Spoor F, Garland T, Krovitz G, Ryan TM, Silcox MT, and Walker A. The primate semicircular
canal system and locomotion. Proc Natl Acad Sci U S A 104: 10808–​10812, 2007.
59. Spoor F, Hublin J-​J, Braun M, and Zonneveld F.The bony labyrinth of Neanderthals. J Hum Evol
44: 141–​165, 2003.
60. Spoor F, Wood B, and Zonneveld F. Implications of early hominid labyrinthine morphology for
evolution of human bipedal locomotion. Nature 369: 645, 1994.
61. Stern JT. Anatomical and functional specializations of the human gluteus maximus. Am J Phys
Anthropol 36: 315–​339, 1972.
62. Stern JT. Climbing to the top: A personal memoir of Australopithecus afarensis. Evol Anthropol
9: 113–​133, 2000.
63. Takeshima N, Kobayashi F, Watanabe T, Tanaka K, Tomita M, and Pollock ML.
Cardiorespiratory responses to cycling exercise in trained and untrained healthy elderly: With special
reference to the lactate threshold. Appl Human Sci 15: 267–​273, 2001.

146
147

Human endurance phenotype evolution

64. Thorpe SK, Crompton RH, Guenther MM, Ker RF, and McNeill Alexander R. Dimensions
and moment arms of the hind-​and forelimb muscles of common chimpanzees (Pan troglodytes). Am
J Phys Anthropol 110: 179–​199, 1999.
65. Ward CV. Interpreting the posture and locomotion of Australopithecus afarensis: Where do we
stand? Am J Phys Anthropol 119: 185–​215, 2002.
66. Yamagiwa J and Basabose AK. Socioecological flexibility of gorillas and chimpanzees. In: Primates
and Cetaceans, edited by Yamagiwa J and Karczmarski L. Tokyo: Springer, 2014, p. 43–​74.
67. Yang N, MacArthur DG, Gulbin JP, Hahn AG, Beggs AH, Easteal S, and North K. ACTN3
genotype is associated with human elite athletic performance. Am J Hum Genet 73: 627–​631, 2003.

147
148

12
ENDURANCE
PHENOTYPE PRIMER
John C. Quindry and Michael D. Roberts

This chapter serves as a primer, summarizing individual physiological traits that contribute
to overall endurance, endurance-​related traits, and performance during endurance events. The
chapter also summarizes basic molecular pathways that modify traits associated with endurance
phenotypes and informs subsequent chapters on the genetic influences of these traits.

Introduction and evolutionary considerations of the endurance phenotype

Healthy humans have a remarkable capacity to engage in endurance activities. Genetic
understanding of endurance capacity describes an emergent group of well-​conserved traits
which span an organizational hierarchy from subcellular to organismal.This primer is founded on
classic understanding of human anthropometry, systems physiology, and tissue-​level adaptations
that inform endurance exercise capacity and adaptation. Descriptions of anthologized informa-
tion are provided in brief, while more recently discovered roles of cell signaling mechanisms
that ultimately underpin endurance capacity and adaptive responses will be covered in more
detail. The reader is directed to Chapter 11 for additional details of endurance and evolution.
For the sake of simplicity, human-​, animal-​, and cell culture-​based investigations are woven into
a common narrative which serves as a necessary background for subsequent understanding of
the genetic foundations of endurance.
Within a given species, endurance exercise speed is size limited, and generally inverse to total
body mass in terms of relative oxygen consumption (40, 42). However, when the importance
of endurance supersedes speed, larger species and individuals are more often benefitted by: 1)
increased metabolic flexibility favoring lipid utilization, and 2) fluid retention strategies that
preserve cardiovascular dynamics, in addition to heat tolerance (42). Proportional expressions of
these, and other, physiologic traits are often predictive of specialized endurance success across
athletic competitions (e.g., 5 km, 40 km, 160 km runs) (78).
Acute physiologic adjustments to prolonged physical activity, and endurance adaptations, are
largely independent of endurance performance capabilities (71). That is, prerequisite physio-
logic adjustments include acute alterations in cardiovascular work and blood flow redistribution,
matched to accelerated ventilatory function and parsimonious access to macronutrients (9). It is
important to remember, from an evolutionary perspective, that cellular “wisdom” self-​regulates

148
149

Endurance phenotype primer

in the name of homeostatic preservation (14). Thus, endurance exercise adaptations are the
cumulative effect of acute cellular stresses expressed across tissues and organ systems. Based
on this rationale, the following sections briefly overview the most well-​defined physiologic
responses to acute and chronic endurance exercise. Subsequent sections detail cell signaling
events responsible for endurance adaptations that underpin many of the physiologic outcomes
described forthright.

Acute and chronic physiologic responses to endurance exercise

Metabolic flexibility
The endurance phenotype is characterized by metabolic flexibility. This point is ultimately
underscored by natural selection of “thrifty” genes directed to promote energy storage and spare
carbohydrate use (14). Given the importance of duration to endurance-​type exercise, metabolic
flexibility is perhaps best characterized by the ability to the support ATP regeneration via mul-
tiple macronutrient substrates in order to maintain steady-​state workload demands.The topic of
metabolic flexibility is immense and beyond the scope of this chapter, but the reader is directed
to authoritative work on the topic (29). Suffice it to state that harnessing stored calories during
exercise involves time-​dependent macronutrient partitioning in a carbohydrate-​sparing fashion
(9). The topic is further nuanced in that the endurance phenotype responds to dietary manipu-
lation and can favor a number of macronutrient-​loading strategies including fat, carbohydrate,
and ketones (9, 80).

Cardiorespiratory responses
Endurance exercise evokes prominent alterations to the integrated cardiorespiratory response.
Elevated cardiac output is directly proportional to increased exercise intensity and reflects
elevations in both heart rate and stroke volume (SV), with the latter exhibiting a plateau at
intensities above moderate workloads (64). Of note, endurance-​trained athletes possess sig-
nificant increases in cardiac output for a given work intensity, primarily due to elevated SV
(58). More recent investigations, aided by advanced in vivo imaging, indicate that SV may not
plateau in elite distance runners as they exceed moderate intensity exercise (88).The respiratory
system also supports endurance activities through ventilation–​perfusion matching via increased
tidal volume and respiratory rate. Most of the increased tidal volume comes from decreases in
inspiratory reserve, a feature that is potentiated by muscular adaptations for inspiration with
endurance training (37, 62).

Blood flow redistribution

During endurance exercise, blood flow is redistributed (vasodilation and constriction) to match
tissue-​level changes in bioenergetics demand, while simultaneously mitigating heat accumu-
lation. Acute changes in blood flow ultimately influence oxygen consumption via the Fick
principle (VO2 = cardiac output − arteriovenous difference (a–vO2)). Accordingly, cardio-
respiratory responses require acute changes in blood flow and are accompanied by an increased
a–​vO2 across a capillary bed (41). For a more in-​depth discussion of this topic, the reader
is directed to an authoritative review (42), but it is notable that blood flow adaptations to
training are difficult to enumerate independent of the other physiologic systems (40, 78). While
integrated blood flow responses to thermoregulation are included shortly, it is important to

149
150

J.C. Quindry and M.D. Roberts

recognize that tissue-​level adaptations for improved blood flow responses to endurance training
are among the most well-​described of the mechanisms detailed below.

Neuroendocrine influence on metabolic flexibility

Metabolic flexibility and carbohydrate-​ sparing during endurance exercise is facilitated, in
part, by pairs of fast-​acting hormones. Insulin and glucagon are central to fast-​acting energy
regulation, with the latter being responsible for carbohydrate and free fatty acid liberation
during endurance exercise (16). Acute endurance exercise promotes energy availability through
time-​dependent glucagon release (77); a response that is blunted in trained individuals (19).
Epinephrine (E) and norepinephrine (NE) further mediate the fast-​acting hormonal influ-
ence on metabolic control, stimulating free fatty acid release and utilization simultaneous to
carbohydrate-​ sparing effects. These are highly important time-​ dependent effects as E/​ NE
levels rise gradually during endurance exercise. Not unlike glucagon, acute exercise-​induced
release of E/​NE to a given moderate intensity workload is blunted in trained individuals (90).
Collective attenuation of hormone release during exercise reveals an interesting feature about
the endurance phenotype. Specifically, the immediate physiologic stress of long-​duration mus-
cular activity is better tolerated in exercise-​trained individuals, a fact that speaks to innumerable
tissue-​level adaptations to be detailed shortly.

Thermoregulatory and blood volume perseveration

The aforementioned blood volume redistribution during endurance exercise is often complicated
by the physiologic need to thermoregulate, and subsequently, counteract the resultant hypovol-
emia (58). Recent investigation of acute ultra-​endurance physiology indicates SV will drift (first
decreasing, then increasing) to moderate ongoing metabolic demands simultaneous to hypovol-
emia and thermoregulatory efforts (50). Long-​term physiologic demands are also bolstered in
the endurance phenotype through increased red blood cell content and increased hematocrit
and red cell volume (55). Moreover, elevated exercise core temperatures potentiate the afore-
mentioned hormonal responses to better control fuel utilization during endurance exercise
(65).These adaptations are fundamental to improved thermoregulatory tolerance in endurance-​
trained individuals as compared to untrained individuals (58).

Section summary
From a comprehensive perspective, it is not surprising that the various aspects of organizational
physiology which support endurance exercise are universally exhibited across the talent spec-
trum from mall walkers to Olympic marathon champions. Moreover, all physiologic aspects
discussed herein are also subject to optimization via endurance training; a fact that is also
independent of aerobic talent. Foundational understanding of these tissue-​and cellular-​level
adaptations are described below.

Skeletal muscle and the myocardium: Responsiveness to endurance training

Endurance training promotes physiological adaptations in a variety of tissues. While describing
an exhaustive list of tissue-​level adaptations exceeds the scope of this chapter, selected thematic
skeletal muscle and myocardial adaptations are described below. Following months of endur-
ance training, well-​documented skeletal muscle adaptations include: 1) increased mitochondrial

150
151

Endurance phenotype primer

expression of the tricarboxylic acid (TCA) cycle and fatty acid oxidation enzymes; 2) increased
electron transport chain components (34); 3) expansion of mitochondrial volume via mito-
chondrial biogenesis (36); 4) majority transition of type IIX to IIA fibers, and limited transition
of type II to type I fibers (~5–​10%, quantified with histological myosin ATPase staining) (84);
and 5) an approximately 20% increase in skeletal muscle capillary content via angiogenesis (44).
Ethical constraints dictate that parallel understanding of myocardial adaptations to endur-
ance training are largely confined to data from animal studies. Notwithstanding, myocardial
adaptations generally mimic skeletal muscle responses, including increased fatty acid oxida-
tion enzymes (74), increased mitochondrial volume (79), and increased capillarization (12).
Furthermore, landmark human-​based studies report that months of endurance training increases
maximal SV (21) as well as cardiac output during exercise by approximately 10% (22).

Whole-​body adaptations resulting from skeletal muscle

and myocardial adaptations
The novel introduction of endurance training, adhering to several months of American College
of Sports Medicine prescriptive criteria, produces an approximately 15–​30% increase in max-
imal oxygen consumption (VO2max) (8). Subsequent scientific revelation introduced the idea
that high-​intensity interval training, involving interspersed intervals of supra-​VO2max running
or cycling with inactive-​or low-​intensity recovery, elicits similar improvements in aerobic cap-
acity. Notably, however, VO2max adaptations to endurance training are heterogenic across the
population. For instance, the HERITAGE Family Study revealed that 60 cycle training sessions
over 20 weeks produced individual VO2max increases ranging between no gain and a 100%
increase (8). Application of heritability calculations to these data suggest approximately 50% of
the VO2max adaptations to endurance training are genetically predisposed, and this topic is the
focus of a subsequent chapter (Chapter 14) on heritability and endurance capacity in humans
(7). Genetic contributors aside, the majority of VO2max adaptations to endurance training are
cardiovascular in nature. Physiologically, this response is marked by increases in maximal stroke
volume, as a primary influence on cardiac output, in addition to increased a–​vO2 difference
(e.g., increases in muscle fiber capillary number and mitochondrial volume) (52).
Beyond increasing VO2max, endurance training elevates lactate threshold (43), the work rate
beyond which incremental increases in exercise intensity elicit a nonlinear increase in blood lac-
tate concentration.Where untrained individuals experience lactate thresholds between 40% and
70% VO2max, endurance-​trained subjects exhibit a lactate threshold at or above 80% VO2max
(43). While the contributors to skeletal muscle fatigue involve accumulation of numerous
metabolites and ions (e.g., inorganic phosphate, lactate, and potassium and hydrogen ions) as
well as muscle glycogen depletion (28), lactate threshold is an important physiological param-
eter to consider, given that it is often more predicative of endurance performance than VO2max
(3). Numerous factors influence lactate threshold, including muscle fiber recruitment patterns,
skeletal muscle oxidative capacity, the rate of O2 delivery to muscle, glycolytic activation, and the
ability to clear lactate (28). However, the aforementioned doubling of skeletal muscle oxidative
capacity through increased mitochondrial function and volume, as well as capillary number, are
the primary determinants of increased lactate threshold values with endurance training (43).
Endurance training may also increase exercise economy, or the workload performed at a
given VO2. While improvements in mechanical efficiency are well documented, the adaptive
propensity is highly variable between individuals. In this regard, elite runners express a 30–​40%
difference in running economy (40). For example, athletes with a lower exercise economy ran
approximately 16.5 km/​hour at a VO2 of 55 ml O2/​kg/​min, whereas athletes with a higher

151
152

J.C. Quindry and M.D. Roberts

exercise economy ran over 19.0 km/​hour at this same VO2. Genetic factors, muscle fiber type,
fiber recruitment patterns during running, and external factors (e.g., wind resistance) likely
contribute to heterogeneity in running economy. However, adaptation is also trainable, in that
years of directed training produce 8–​14% increases in the running economy of elite endurance
athletes (40). While the nuanced mechanisms remain emergent, experts posit that a fast-​to-​slow
conversion of myosin isoforms over long-​term training periods may underpin adaptations in
economy (40).

Molecular pathways that facilitate endurance training adaptations

Cellular-​level adaptations to endurance training are widespread across skeletal muscle,
myocardium, and numerous other tissues. Upon the introduction of sustained exercise, intracel-
lular changes are acute and marked by a concert of biochemical alterations in metabolite levels,
protein signaling, and gene expression patterns. Over a series of training bouts, these molecular
events accumulate with tissue-​level adaptations to produce the aforementioned increases in
VO2max, lactate threshold, and exercise economy. Collectively, these responses contribute to
the genetics of cardiorespiratory fitness (detailed in Chapter 15), and are largely driven by mito-
chondrial adaptations (detailed in Chapter 16).
Given that exercise-​responsive molecules orchestrate a plethora of signaling events and
lead to extraordinary exercise adaptations, there is widespread interest in elucidating the
interrelationships among these molecules. In this regard, recent advances in powerful ‘omics’-​
based methods (e.g., metabolomics, phosphoproteomics, RNA sequencing) have been used
to characterize metabolite, kinase, transcription factor, and RNA transcript profile changes
following acute and chronic endurance training. For instance, RNA sequencing indicates that
more than 3400 skeletal muscle mRNA transcripts are altered by 12 weeks of endurance training
(47). While many altered transcripts associate with oxidative ATP production, 34 transcripts are
previously unidentified targets within exercised muscle. Moreover, mass spectrometry-​based
phosphoproteomics demonstrate that approximately 15 minutes of high-​intensity cycling exer-
cise alters the phosphorylation status of more than 560 skeletal muscle proteins (32). Canonical
protein signaling networks activated by endurance exercise include adenosine monophosphate-​
activated kinase (AMPK) and CaMK (as described below) in addition to a novel complement
of muscle proteins subject to exercise-​induced phosphorylation. Recent animal studies con-
firm that hundreds of myocardial proteins are acutely phosphorylated in response to tread-
mill running (31). Excitingly, the majority of these phospho-​sites have not previously been
associated with cardiac adaptive response to exercise.
Collectively, omics-​based studies conceptualize two critical points: 1) thousands of molecules
(e.g., kinases, transcription factors, mRNA transcripts, and epigenetic factors including
miRNAs) coordinate to facilitate endurance exercise adaptations; and 2) the functional signifi-
cance of molecular profiles in endurance exercised skeletal muscle, myocardia, and other tissues
remain unknown. Indeed, understanding molecular pathways activated in exercised tissues is just
emerging. A Common Fund Program from the National Institutes of Health (NIH), named the
Molecular Transducers of Physical Activity Consortium (MoTrPAC), was a 2015 NIH initiative
that identified a consortium to extensively investigate phosphoproteomic and transcriptomic
alterations in endurance exercised skeletal muscle and other tissues. Ongoing efforts will con-
tinue to validate previously identified exercise-​responsive molecules and identify new targets to
better understand the molecular landscape that underpins endurance training adaptations. The
following sections discuss the most well-​characterized molecules and signaling pathways that
promote endurance training adaptations at the cellular level to date. As noted above, one must

152
153

Endurance phenotype primer

appreciate the wealth of currently unidentified molecules and pathways that mediate endurance
exercise adaptations, as will soon be revealed.

PGC-​1α and VEGF

Peroxisome proliferator-​activated receptor gamma coactivator 1 alpha (PGC-​1α) and vascular
endothelial growth factor (VEGF) are perhaps the two most well-​studied molecular transcripts
(and corresponding proteins) upregulated in response to endurance training in skeletal muscle
and other tissues. Modern understanding dictates that PGC-​1α is a “master regulator” of mito-
chondrial biogenesis (39). Specifically, the PGC-​1α protein acts as a transcriptional coactivator
and associates with nuclear transcription factors to up-​or downregulate mRNA transcription
rates of other target genes. PGC-​1α, for example, regulates transcriptional activity of the peroxi-
some proliferator-​activated receptor gamma (PPARγ) nuclear receptor gene. PGC-​1α-​induced
increases in PPARγ DNA binding facilitates an upregulation of electron transport chain genes
(66). PGC-​1α also binds to and activates the nuclear respiratory factor-​1 (NRF1) transcription
factor in muscle cells, thereby upregulating genes that facilitate mitochondrial DNA replication
(86). These findings are bolstered by numerous reports suggesting endurance training promotes
mitochondrial biogenesis through the induction of PGC-​1α protein expression following one
or multiple bouts of training in animals (5) and humans alike (63). The effects of one exer-
cise bout on PGC-​1α mRNA and protein levels in human skeletal muscle are illustrated in
Figure 12.1.
Compelling evidence indicates that increased PGC-​1α mRNA is a foundational response to
acute exercise, inasmuch as potent trigger for downstream adaptive responses. However, the role
of PGC-​1α is not an exclusive mediator for endurance exercise adaptations. As a case in point,
global Pgc-​1α-​knockout mice express 17–​44% lower mitochondrial content in skeletal and
cardiac muscle as compared to wild-​type mice. Despite this fact, skeletal muscle mitochondrial
biogenesis still occurred in the knockout mice following 5 weeks of voluntary wheel running
compared to untrained knockout mice (2) (Figure 12.2). Other observations indicate endur-
ance exercise rapidly increases the mRNA expression of genes involved with skeletal muscle

10 1.6
Fold-change (VL muscle)

8 1.4

6 1.2

4 1.1

2 1.0

Pre IP 2h 24 h Pre IP 2h 24 h

PGC-1α mRNA PGC-1α protein

Figure 12.1 Figure demonstrating that PGC-​1α mRNA increases in human vastus lateralis (VL) muscle
up to 2 hours following exhaustive cycling at 65% VO2max.
Source: Figure adapted from Mathai et al. (49).

153
154

J.C. Quindry and M.D. Roberts

1.5
Mitochondrial yield
(mg/g tissue)

1.0

0.5

PGC-1α WT PGC-1α
KO mice mice KO mice

EX trained

Figure 12.2 Recent data demonstrate in global Pgc-​1α knockout (KO) mice that hindlimb muscle
mitochondrial content is lower than wild-​type controls. The solid and dashed lines represent the mean
and variance measures of un-exercised wild-type (WT) mice. EX=exercise.

Nuclear translocation and integration of PGC-1α

cle us into numerous transcription initiation complexes
myonu
cell
muscle

↑ transcription rates of genes regulating:

Exercise-induced signals leading to - mitochondrial biogenesis
the activation of the PGC-1α promoter - fatty acid metabolism
(e.g., p38 MAPK, CaMK, AMPK) - angiogenesis

Translation of PGC-1α mRNA into protein

Figure 12.3 Endurance exercise rapidly activates multiple pathways (e.g., p38 MAPK pathway, CaMK
pathway, and AMPK pathway) that subsequently activate the PGC-​1α promoter and increases mRNA
levels, and subsequent translation rates.

mitochondrial biogenesis via Nrf1/​2 transcription factor binding (e.g., cytochrome c and citrate
synthase) prior to PGC-​1α protein levels increasing after exercise (85). This important finding
indicates exercise-​induced mitochondrial biogenesis involves redundant stimuli that facilitate
exercise-​induced increases in mitochondrial content.
Beyond involvement with mitochondrial biogenesis, PGC-​1α upregulates genes involved
with skeletal muscle fatty acid metabolism (59), cardiac muscle mitochondrial function (4), and
hepatic gluconeogenesis (53). Therefore, PGC-​1α is a critical regulator of mitochondrial physi-
ology and metabolic homeostasis in numerous tissues. Figure 12.3 summarizes exercise-​induced
PGC-​1α mRNA and protein level upregulation of numerous metabolic processes.
Like PGC-​1α, VEGF mRNA and protein expression responses to endurance exercise have
been well-​studied in various tissues.VEGF is a growth factor expressed in numerous cell types
and secreted into the extracellular space. Once secreted, VEGF binds to cognate receptors
(VEGFR1–​3) on vascular endothelial cells initiating angiogenesis, or the formation of new capil-
laries (24). The VEGF family contains five proteins including VEGF-​A, placenta growth factor,
VEGF-​B, VEGF-​C, and VEGF-​D (72). VEGF-​A has been the most extensively studied form,
and notably, numerous spliced VEGF-​A mRNA variants (e.g., VEGF165, VEGF189, VEGF206)

154
155

Endurance phenotype primer

are translated into further differentiated VEGF-​A isoforms.These VEGF isoforms possess altered
receptor affinity binding (13), although the endurance exercise and training effects on mRNA
isoform expression and spliced variants in different tissue types are not completely described.
VEGF mRNA and protein levels are elevated in rodent (6) and human skeletal muscle
(67), as well as rodent myocardium (35) following acute and chronic endurance training. These
cross-​species data indicate VEGF is transcriptionally upregulated, rapidly translated, and secreted
in order to initiate and maintain angiogenesis in multiple tissues exposed to acute and chronic
exercise training scenarios. Advanced microscopy techniques have revealed, more mechanis-
tically, that human skeletal myofibers rapidly produce and secrete VEGF protein containing
vesicles into the extracellular space following extended duration cycling (33), and supports
VEGF as fundamental to the endurance phenotype.
While a variety of factors activate the VEGF promoter, hypoxia is the most potent stimulator
of VEGF mRNA expression. Classic investigations first revealed that lower muscle PO2 levels
during exercise rapidly activate the hypoxia-​inducible factor 1α (HIF-​1α) transcription factor
which, in turn, binds to and activates the VEGF gene promoter (81). Alternately, findings from
a study using the Pgc-​1α-​knockout mouse line, indicate that skeletal muscle AMPK activation
by either prolonged endurance exercise, or pharmacological activation via 5-​aminoimidazole-​
4-​carboxamide-​1-​beta-​D-​r ibofuranoside (AICAR), increases VEGF protein expression only in
wild-​type control mice (46). Moreover, in Pgc-​1α knockout mice, basal VEGF protein levels
are expressed at 60–​80% of control mouse levels and the resulting capillary number is 20%
lower than controls. Thus, skeletal muscle VEGF mRNA expression is also modulated through
the AMPK-​mediated activation of PGC-​1α in a HIF-​1α-​independent manner. More import-
antly, these studies reinforce the notion that: 1) multiple molecules promote endurance exer-
cise adaptations, and 2) PGC-​1α and VEGF are two critical signaling molecules that facilitate
physiological adaptations to endurance exercise with some degree of interdependence.

AMPK signaling
The 5ʹ AMP-​activated protein kinase (AMPK) is a hetero-​trimeric enzyme with a catalytic α
subunit and regulatory β and γ subunits. Notably, the catalytic subunit catalyzes the phosphoryl-
ation of cytoplasmic and nuclear proteins as part of signal transduction and transcription factor
activation. During exercise, AMPK is central to cellular energy sensing and homeostatic control
in response to: 1) skeletal ATP turnover and increased AMP concentrations which bind to and
activate AMPK (68), and 2) declining cellular glycogen levels which interact with AMPK β-​
subunit activity and reciprocal effects on subsequent glycogen level alterations (73).
Incumbent upon highly conserved metabolic regulators, AMPK is redundantly controlled
independently of cellular bioenergetic status. For instance, stress-​responsive molecules such as
LKB1 and CaMKKβ phosphorylate and increase AMPK activity through interactions with the
catalytic α-​subunit AMPK at the Thr172 residue (68). Resultant AMPK activation increases
cellular glucose uptake, facilitates fatty acid oxidation (54), promotes a decrease in anabolic
processes (e.g., protein and glycogen synthesis) (68), increases mitochondrial biogenesis (89),
and promotes fiber-​type shift from type II glycolytic to oxidative fibers (69). Figure 12.4
summarizes the upstream activators and downstream physiological effects of AMPK activation.
Cross-​species investigations indicate endurance exercise acutely increases phosphorylated
AMPK levels and activity in skeletal muscle (15). Moreover, acute endurance exercise increases
markers of AMPK activity in myocardium (17), aortic vessels (10), and pancreatic β-​cells (11),
reinforcing the concept that AMPK activation is evolutionarily conserved in metabolically
active tissues in cross-​species fashion. Emphasizing centrality to bioenergetics control, 4 weeks

155
 newgenrtpdf
156
Low AMPK activity Increased AMPK activity (during exercise)

upstream kinases
(e.g., LKB1, CaMKKβ)

catalytic subunit with

kinase domain P
Thr172
↑ ATP turnover
α γ
P
α γ β AM

β
↑ glycogenolysis
regulatory
subunits
glycogen

↑ phosphorylation of transcription
(cytoplasmic localization) factors, kinases, and other proteins
↑ cellular glucose uptake
↑ fatty acid oxidation
↓ anabolic processes
↑ activation of molecules involved with mitochondrial biogenesis
↑ activation of molecules that promote fast-to-slow fiber type shift

Figure 12.4 Left: the unstimulated AMPK enzyme; right: increased AMPK activity (during exercise).
157

Endurance phenotype primer

of AICAR administration in rodents induced a skeletal muscle gene expression profile that
roughly paralleled endurance training. As such, treadmill running endurance was enhanced by
44% in treated animals (56). However, some skepticism remains as to the essentiality of AMPK
activation facilitating endurance training adaptations in muscle. In this regard, basal mitochon-
drial markers were 20% lower in the gastrocnemius of AMPKα2-​knockout mice versus wild-​
type controls, although wheel-​running adaptations in knockout mice remained unmitigated
in terms of oxidative capacity (38). Additionally, it has been reported mitochondrial biogen-
esis occurs in AMPKα2-​knockout mice that undertake voluntary wheel running (69). These
findings led both research groups to conclude that AMPK signaling may be most important
for basal metabolic homeostasis, albeit redundant pathways or downstream substrates facilitate
adaptations to more extreme bioenergetics challenges such as endurance training. In spite of
seemingly disparate findings, the overwhelming body of evidence indicates that AMPK pathway
activation is critically involved in promoting the oxidative phenotype in metabolically active
tissues of exercised animals.

CaMKII–​p38/​MAPK
The calcium/​calmodulin-​dependent protein kinase II (CaMKII) enzyme facilitates endur-
ance training adaptations in skeletal muscle. While multiple classes of CaMK enzymes exist,
including CaMKI, CaMKII, and CaMKIV (75), only CaMKII isoforms are enriched in
skeletal muscle (70). Skeletal muscle CaMKII activity is increased by rising cytosolic cal-
cium concentrations during acute endurance exercise. This response results in increased cal-
cium/​calmodulin (Ca2+/​CaM) complexes and occurs in a species-​independent fashion (70).
In turn, Ca2+/​CaM binds to and activates CaMKII leading to autophosphorylation at the
Thr287 residue as well as increased activity. Similar to AMPK, CaMKII phosphorylates cyto-
plasmic and nuclear proteins to promote signal transduction and transcription factor activa-
tion. For instance, CaMKII phosphorylates numerous proteins involved with transcriptional
machinery formation in skeletal muscle (e.g., HDAC4, myogenin, SRF) and upregulates genes
associated with an endurance phenotype (70). As a proof of concept, pharmacologic inhib-
ition of CaMKII activity in vitro prevents calcium-​induced increases in PGC-​1α and mtTFA
protein expression as well as mitochondrial biogenesis in L6 myotubes (61). A follow-​up study
confirmed these findings in isolated rat epitrochlearis muscle, further demonstrating that
CaMKII phosphorylates p38 mitogen-​activated protein kinase (MAPK) in order to facilitate
PGC-​1α protein expression and markers of mitochondrial biogenesis, and additional evidence
exists suggesting CaMKII acutely phosphorylates p38/​MAPK in skeletal muscle to promote
an endurance phenotype (26).
CaMKII activation is also critical for increasing GLUT4 gene expression through acetyl-
ating histones that flank the GLUT4 promoter (60). This outcome is accomplished through
CaMKII phosphorylating and inactivating the HDAC4 histone deacteylase enzyme. Calcium/​
calmodulin-​sensitive CaMK kinase (CaMKK) is upstream of several CaMK enzymes, although
it does not phosphorylate CaMKII (82). However, CaMKK phosphorylates AMPK at Thr172
and increases AMPK activity in order to increase GLUT4 and CD36 sarcolemmal localization
as well as glucose and fatty acid uptake (1). Hence, increases in intracellular calcium during
endurance exercise activate parallel signaling pathways (i.e., CaMKII and AMPK) which pro-
mote skeletal muscle mitochondrial biogenesis and nutrient uptake. Figure 12.5 summarizes
the upstream activators and downstream physiological effects of CaMKII activation in skeletal
muscle.

157
158

J.C. Quindry and M.D. Roberts

cle us
myonu
cell HDAC4
muscle P P
P
P

CaMKII
P (active) P
p38 ↑ transcription rates of genes regulating:
CaMKII - mitochondrial biogenesis (PGC1-α, mTFAM)
(inactive) - glucose uptake (GLUT-4)
Ca++
CaM sarcoplasmic reticulum

Figure 12.5 Calcium (Ca2+) flux increases in the cytosol during endurance exercise.

SIRT1/​3
Sirtuin (SIRT) proteins encompass seven (SITR1–​7) deacetylase enzymes, although endurance
exercise activates SIRT1 and SIRT3 through increased nuclear and mitochondrial nicotinamide
adenine dinucleotide (NAD+) concentrations (83). SIRT1 regulates mitochondrial biogenesis
through deacetylating and increasing the integration of PGC-​1α into nuclear transcriptional
machinery (25). In muscle cells, SIRT1 overexpression increases PGC1α transcriptional activity
and mitochondrial gene expression, promoting fatty acid oxidation (25). However, other studies
report no reduction in mitochondrial function or number when SIRT1 protein or deacetylase
activity is knocked out in myotubes or murine skeletal muscle (83). Conflicting data make it
difficult to conclude whether endurance exercise-​induced increases in nuclear NAD+ and sub-
sequent SIRT1-​mediated PGC-​1α deacetylation are required for mitochondrial adaptations.
The SIRT3 deacetylase enzyme is localized to mitochondria and enhances mitochondrial
function like SIRT1. SIRT3 deacetylates and activates TCA cycle and β-​oxidation enzymes
(83), which during endurance exercise logically elevates NAD+ concentrations, increasing fuel
substrate oxidation rates. However, SIRT3 activation may attenuate mitochondrial biogenesis
given that overexpression of SIRT3 in C2C12 myotubes depresses mitochondrial protein syn-
thesis through the deacetylation of mitochondrial ribosomal protein L1 (87). Hence, during
energy-​deprived states (e.g., exercise or fasting), SIRT3 seemingly operates to upregulate the
rate of mitochondrial fuel oxidation while shunting consumptive pathways such as mitochon-
drial biogenesis.

NRF2 signaling
Upregulation of antioxidant defense mechanisms is another noteworthy adaptation of endur-
ance training in skeletal muscle and myocardial tissue. Nuclear respiratory factor 2 (NRF2) is
the master regulator of antioxidant defense mechanisms of this response (57). In confirmation,
endurance training increases NRF2 signaling across multiple tissues including skeletal muscle,
myocardium, kidney, brain, liver, testes, and prostate (18). Multiple studies report acute tread-
mill bouts increase skeletal muscle and myocardium NRF2 protein levels as well as the expres-
sion of NRF2-​dependent genes (e.g., CAT, SOD2, HO-​1, NQO1, and glutathione synthesis
genes) (18).
NRF2 is a member of the leucine zipper family of transcription factors. In basal, unstressed
states, NRF2 localizes to the cytosol via association with KEAP1. Elevations in reactive
oxygen species (ROS) during endurance exercise modify cysteine residues on KEAP1. This
event promotes the dissociation of KEAP1 from NRF2, and allows NRF2 to translocate into
the nucleus. Upon nuclear entry, NRF2 binds promoter regions of the genome containing

158
159

Endurance phenotype primer

antioxidant response elements altering expression of aforementioned endogenous antioxidant

defenses and antiapoptotic regulators (e.g., Bcl-​2) (76). Collectively, NRF2 activation through
acute and chronic endurance training bolsters antioxidant defenses in numerous tissues and
promotes mitochondrial homeostatic maintenance. Additionally, NRF2 nuclear localization
and DNA binding post exercise contributes to mitochondrial biogenesis (85), further under-
scoring a critical role in endurance training adaptations.

eNOS and nitric oxide signaling

Nitric oxide (NO) is also fundamental to endurance training adaptations in skeletal muscle,
myocardium, and vascular tissues. NO is produced endogenously via nitric oxide synthase
(NOS) enzymes via L-​arginine substrate utilization. The primary function of NO is to facili-
tate vasodilation through vascular smooth muscle relaxation (27). NOS isozymes are expressed
in tissue-​independent fashion (e.g., nNOS in nerve and skeletal muscle, iNOS in immune
cells, and eNOS in endothelial nitric oxide) (51). Chronic endurance training increases skeletal
muscle and myocardium NO levels (23). Phosphorylation of eNOS is observed post exercise,
indicative of heightened activity in cardiac and skeletal muscle capillary beds (30).
Beyond vasodilation, NO also affects molecular signaling cascades in skeletal muscle and
other tissues. For instance, eNOS knockout mice present impairments in basal mitochon-
drial function and a dysregulation in CamKII signaling (45). Moreover, NO influences VEGF
mRNA and protein regulation in several exercised tissues (20). As yet another redundant factor
in promoting mitochondrial biogenesis, NO directly activates skeletal muscle AMPK signaling,
increasing PGC-​1α mRNA and protein expression (48). Hence, elevation of NO during and
following endurance exercise initiates signaling cascades which collectively increase mitochon-
drial number and function as well as angiogenesis.

Conclusion
One should appreciate the complex, tissue-​wide, molecular signaling phenomena that occur
in response to acute and chronic endurance training. Thematic of these outcomes, exercise
adaptations operate through parallel or redundant signaling mechanisms including PGC-​1α,
VEGF, and AMPK. Foundational understanding of these and other molecules will give rise to
new horizons in related mechanisms, yet to be elucidated through future omics-​based endeavors
and related loss of function in whole animal and cell culture models. Indeed, the interface
between genetic responses to exercise will likely shed new light on classic understanding of
endurance phenotypes and the underpinning physiologic responses necessary for long-​duration
exercise adaptations. Many of the most pressing scientific questions are already posed within
this text.

References
1. Abbott MJ, Edelman AM, and Turcotte LP. CaMKK is an upstream signal of AMP-​activated pro-
tein kinase in regulation of substrate metabolism in contracting skeletal muscle. Am J Physiol Regul
Integr Comp Physiol 297: R1724–​1732, 2009.
2. Adhihetty PJ, Uguccioni G, Leick L, Hidalgo J, Pilegaard H, and Hood DA. The role of PGC-​
1alpha on mitochondrial function and apoptotic susceptibility in muscle. Am J Physiol Cell Physiol
297: C217–​225, 2009.
3. Allen WK, Seals DR, Hurley BF, Ehsani AA, and Hagberg JM. Lactate threshold and distance-​
running performance in young and older endurance athletes. J Appl Physiol 58: 1281–​1284, 1985.

159
160

J.C. Quindry and M.D. Roberts

4. Arany Z, He H, Lin J, Hoyer K, Handschin C, Toka O, Ahmad F, Matsui T, Chin S, Wu

PH, Rybkin, II, Shelton JM, Manieri M, Cinti S, Schoen FJ, Bassel-​Duby R, Rosenzweig A,
Ingwall JS, and Spiegelman BM.Transcriptional coactivator PGC-​1 alpha controls the energy state
and contractile function of cardiac muscle. Cell Metab 1: 259–​271, 2005.
5. Baar K, Wende AR, Jones TE, Marison M, Nolte LA, Chen M, Kelly DP, and Holloszy JO.
Adaptations of skeletal muscle to exercise: rapid increase in the transcriptional coactivator PGC-​1.
FASEB J 16: 1879–​1886, 2002.
6. Birot OJ, Koulmann N, Peinnequin A, and Bigard XA. Exercise-​induced expression of vas-
cular endothelial growth factor mRNA in rat skeletal muscle is dependent on fibre type. J Physiol
552: 213–​221, 2003.
7. Bouchard C, Daw EW, Rice T, Perusse L, Gagnon J, Province MA, Leon AS, Rao DC,
Skinner JS, and Wilmore JH. Familial resemblance for VO2max in the sedentary state: the
HERITAGE family study. Med Sci Sports Exerc 30: 252–​258, 1998.
8. Bouchard C, and Rankinen T. Individual differences in response to regular physical activity. Med
Sci Sports Exerc 33: S446–​451, 2001.
9. Burke LM. Re-​examining high-​fat diets for sports performance: Did we call the “nail in the coffin”
too soon? Sports Med 45 Suppl 1: S33–​49, 2015.
10. Cacicedo JM, Gauthier MS, Lebrasseur NK, Jasuja R, Ruderman NB, and Ido Y.Acute exercise
activates AMPK and eNOS in the mouse aorta. Am J Physiol Heart Circ Physiol 301: H1255–​1265, 2011.
11. Calegari VC, Zoppi CC, Rezende LF, Silveira LR, Carneiro EM, and Boschero AC. Endurance
training activates AMP-​activated protein kinase, increases expression of uncoupling protein 2 and
reduces insulin secretion from rat pancreatic islets. J Endocrinol 208: 257–​264, 2011.
12. Carlsson S, Ljungqvist A, Tornling G, and Unge G. The myocardial capillary vasculature in
repeated physical exercise. An experimental investigation in the rat. Acta Pathol Microbiol Scand A
86: 117–​119, 1978.
13. Cebe Suarez S, Pieren M, Cariolato L, Arn S, Hoffmann U, Bogucki A, Manlius C, Wood
J, and Ballmer-​Hofer K. A VEGF-​A splice variant defective for heparan sulfate and neuropilin-​1
binding shows attenuated signaling through VEGFR-​2. Cell Mol Life Sci 63: 2067–​2077, 2006.
14. Chakravarthy MV, and Booth FW. Eating, exercise, and “thrifty” genotypes: Connecting the dots
toward an evolutionary understanding of modern chronic diseases. J Appl Physiol 96: 3–​10, 2004.
15. Coffey VG, Zhong Z, Shield A, Canny BJ, Chibalin AV, Zierath JR, and Hawley JA. Early
signaling responses to divergent exercise stimuli in skeletal muscle from well-​trained humans. FASEB
J 20: 190–​192, 2006.
16. Coker RH, and Kjaer M. Glucoregulation during exercise: The role of the neuroendocrine system.
Sports Med 35: 575–​583, 2005.
17. Coven DL, Hu X, Cong L, Bergeron R, Shulman GI, Hardie DG, and Young LH. Physiological
role of AMP-​activated protein kinase in the heart: Graded activation during exercise. Am J Physiol
285: E629–​636, 2003.
18. Done AJ, and Traustadottir T. Nrf2 mediates redox adaptations to exercise. Redox Biol
10: 191–​199, 2016.
19. Drouin R, Lavoie C, Bourque J, Ducros F, Poisson D, and Chiasson JL. Increased hepatic glu-
cose production response to glucagon in trained subjects. Am J Physiol 274: E23–​28, 1998.
20. Dulak J, Jozkowicz A, Dembinska-​Kiec A, Guevara I, Zdzienicka A, Zmudzinska-​
Grochot D, Florek I, Wojtowicz A, Szuba A, and Cooke JP. Nitric oxide induces the synthesis
of vascular endothelial growth factor by rat vascular smooth muscle cells. Arterioscler Thromb Vasc Biol
20: 659–​666, 2000.
21. Ehsani AA, Hagberg JM, and Hickson RC. Rapid changes in left ventricular dimensions and mass
in response to physical conditioning and deconditioning. Am J Cardiol 42: 52–​56, 1978.
22. Ekblom B, Astrand PO, Saltin B, Stenberg J, and Wallstrom B. Effect of training on circulatory
response to exercise. J Appl Physiol 24: 518–​528, 1968.
23. Gavin TP, Spector DA, Wagner H, Breen EC, and Wagner PD. Nitric oxide synthase inhib-
ition attenuates the skeletal muscle VEGF mRNA response to exercise. J Appl Physiol (1985)
88: 1192–​1198, 2000.
24. Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A, Abramsson A, Jeltsch M,
Mitchell C, Alitalo K, Shima D, and Betsholtz C. VEGF guides angiogenic sprouting utilizing
endothelial tip cell filopodia. J Cell Biol 161: 1163–​1177, 2003.

160
161

Endurance phenotype primer

25. Gerhart-​Hines Z, Rodgers JT, Bare O, Lerin C, Kim SH, Mostoslavsky R, Alt FW, Wu Z,
and Puigserver P. Metabolic control of muscle mitochondrial function and fatty acid oxidation
through SIRT1/​PGC-​1alpha. EMBO J 26: 1913–​1923, 2007.
26. Gibala MJ, McGee SL, Garnham AP, Howlett KF, Snow RJ, and Hargreaves M. Brief intense
interval exercise activates AMPK and p38 MAPK signaling and increases the expression of PGC-​
1alpha in human skeletal muscle. J Appl Physiol (1985) 106: 929–​934, 2009.
27. Gielen S, Sandri M, Erbs S, and Adams V. Exercise-​induced modulation of endothelial nitric
oxide production. Curr Pharm Biotechnol 12: 1375–​1384, 2011.
28. Gladden LB. Lactate metabolism: a new paradigm for the third millennium. J Physiol 558: 5–​30, 2004.
29. Goodpaster BH, and Sparks LM. Metabolic flexibility in health and disease. Cell Metab
25: 1027–​1036, 2017.
30. Grijalva J, Hicks S, Zhao X, Medikayala S, Kaminski PM, Wolin MS, and Edwards JG.
Exercise training enhanced myocardial endothelial nitric oxide synthase (eNOS) function in diabetic
Goto-​Kakizaki (GK) rats. Cardiovasc Diabetol 7: 34, 2008.
31. Guo H, Isserlin R, Emili A, and Burniston JG. Exercise-​responsive phosphoproteins in the heart.
J Mol Cell Cardiol 111: 61–​68, 2017.
32. Hoffman NJ, Parker BL, Chaudhuri R, Fisher-​Wellman KH, Kleinert M, Humphrey SJ,
Yang P, Holliday M, Trefely S, Fazakerley DJ, Stockli J, Burchfield JG, Jensen TE, Jothi
R, Kiens B, Wojtaszewski JF, Richter EA, and James DE. Global phosphoproteomic analysis
of human skeletal muscle reveals a network of exercise-​regulated kinases and AMPK substrates. Cell
Metab 22: 922–​935, 2015.
33. Hoier B, Prats C, Qvortrup K, Pilegaard H, Bangsbo J, and Hellsten Y. Subcellular local-
ization and mechanism of secretion of vascular endothelial growth factor in human skeletal muscle.
FASEB J 27: 3496–​3504, 2013.
34. Holloszy JO. Adaptation of skeletal muscle to endurance exercise. Med Sci Sports 7: 155–​164, 1975.
35. Iemitsu M, Maeda S, Jesmin S, Otsuki T, and Miyauchi T. Exercise training improves aging-​
induced downregulation of VEGF angiogenic signaling cascade in hearts. Am J Physiol Heart Circ
Physiol 291: H1290–​1298, 2006.
36. Irrcher I, Adhihetty PJ, Joseph AM, Ljubicic V, and Hood DA. Regulation of mitochondrial
biogenesis in muscle by endurance exercise. Sports Med 33: 783–​793, 2003.
37. Johnson BD, Aaron EA, Babcock MA, and Dempsey JA. Respiratory muscle fatigue during
exercise: Implications for performance. Med Sci Sports Exerc 28: 1129–​1137, 1996.
38. Jorgensen SB, Treebak JT, Viollet B, Schjerling P, Vaulont S, Wojtaszewski JF, and Richter
EA. Role of AMPKalpha2 in basal, training-​, and AICAR-​induced GLUT4, hexokinase II, and mito-
chondrial protein expression in mouse muscle. Am J Physiol 292: E331–​339, 2007.
39. Jornayvaz FR, and Shulman GI. Regulation of mitochondrial biogenesis. Essays Biochem
47: 69–​84, 2010.
40. Joyner MJ. Modeling: Optimal marathon performance on the basis of physiological factors. J Appl
Physiol (1985) 70: 683–​687, 1991.
41. Joyner MJ. Preclinical and clinical evaluation of autonomic function in humans. J Physiol
594: 4009–​4013, 2016.
42. Joyner MJ, and Casey DP. Regulation of increased blood flow (hyperemia) to muscles during exer-
cise: A hierarchy of competing physiological needs. Physiol Rev 95: 549–​601, 2015.
43. Joyner MJ, and Coyle EF. Endurance exercise performance: The physiology of champions. J Physiol
586: 35–​44, 2008.
44. Klausen K, Andersen LB, and Pelle I. Adaptive changes in work capacity, skeletal muscle
capillarization and enzyme levels during training and detraining. Acta Physiol Scand 113: 9–​16, 1981.
45. Lee-​Young RS, Ayala JE, Hunley CF, James FD, Bracy DP, Kang L, and Wasserman DH.
Endothelial nitric oxide synthase is central to skeletal muscle metabolic regulation and enzymatic
signaling during exercise in vivo. Am J Physiol Regul Integr Comp Physiol 298: R1399–​R1408, 2010.
46. Leick L, Hellsten Y, Fentz J, Lyngby SS, Wojtaszewski JF, Hidalgo J, and Pilegaard H.
PGC-​1alpha mediates exercise-​induced skeletal muscle VEGF expression in mice. Am J Physiol
297: E92–​103, 2009.
47. Lindholm ME, Giacomello S, Werne Solnestam B, Fischer H, Huss M, Kjellqvist S, and
Sundberg CJ. The impact of endurance training on human skeletal muscle memory, global isoform
expression and novel transcripts. PLoS Genet 12: e1006294, 2016.

161
162

J.C. Quindry and M.D. Roberts

48. Lira VA, Brown DL, Lira AK, Kavazis AN, Soltow QA, Zeanah EH, and Criswell DS. Nitric
oxide and AMPK cooperatively regulate PGC-​1 in skeletal muscle cells. J Physiol 588: 3551–​3566, 2010.
49. Mathai AS, Bonen A, Benton CR, Robinson DL, and Graham TE. Rapid exercise-​induced
changes in PGC-​ 1alpha mRNA and protein in human skeletal muscle. J Appl Physiol (1985)
105: 1098–​1105, 2008.
50. Mattsson CM, Stahlberg M, Larsen FJ, Braunschweig F, and Ekblom B. Late cardiovascular
drift observable during ultraendurance exercise. Med Sci Sports Exerc 43: 1162–​1168, 2011.
51. McAllister RM, Newcomer SC, and Laughlin MH. Vascular nitric oxide: Effects of exercise
training in animals. Appl Physiol Nutr Metab 33: 173–​178, 2008.
52. McGavock JM, Warburton DE, Taylor D, Welsh RC, Quinney HA, and Haykowsky MJ.
The effects of prolonged strenuous exercise on left ventricular function: A brief review. Heart Lung
31: 279–​292, 2002.
53. Meirhaeghe A, Crowley V, Lenaghan C, Lelliott C, Green K, Stewart A, Hart K, Schinner
S, Sethi JK, Yeo G, Brand MD, Cortright RN, O’Rahilly S, Montague C, and Vidal-​Puig
AJ. Characterization of the human, mouse and rat PGC1 beta (peroxisome-​proliferator-​activated
receptor-​gamma co-​activator 1 beta) gene in vitro and in vivo. Biochem J 373: 155–​165, 2003.
54. Merrill GF, Kurth EJ, Hardie DG, and Winder WW. AICA riboside increases AMP-​activated pro-
tein kinase, fatty acid oxidation, and glucose uptake in rat muscle. Am J Physiol 273: E1107–​1112, 1997.
55. Montero D and Lundby C. Red cell volume response to exercise training: Association with aging.
Scand J Med Sci Sports 27: 674–​683, 2017.
56. Narkar VA, Downes M, Yu RT, Embler E, Wang YX, Banayo E, Mihaylova MM, Nelson
MC, Zou Y, Juguilon H, Kang H, Shaw RJ, and Evans RM. AMPK and PPARdelta agonists are
exercise mimetics. Cell 134: 405–​415, 2008.
57. Nguyen T, Nioi P, and Pickett CB. The Nrf2-​antioxidant response element signaling pathway and
its activation by oxidative stress. J Biol Chem 284: 13291–​13295, 2009.
58. Nielsen B, Hales JR, Strange S, Christensen NJ, Warberg J, and Saltin B. Human circulatory
and thermoregulatory adaptations with heat acclimation and exercise in a hot, dry environment. J
Physiol 460: 467–​485, 1993.
59. Nikolic N, Rhedin M, Rustan AC, Storlien L, Thoresen GH, and Stromstedt M. Overexpression
of PGC-​1alpha increases fatty acid oxidative capacity of human skeletal muscle cells. Biochem Res Int
2012: 714074, 2012.
60. Ojuka EO, Goyaram V, and Smith JA. The role of CaMKII in regulating GLUT4 expression in
skeletal muscle. Am J Physiol 303: E322–​331, 2012.
61. Ojuka EO, Jones TE, Han DH, Chen M, and Holloszy JO. Raising Ca2+ in L6 myotubes
mimics effects of exercise on mitochondrial biogenesis in muscle. FASEB J 17: 675–​681, 2003.
62. Olafsson S and Hyatt RE.Ventilatory mechanics and expiratory flow limitation during exercise in
normal subjects. J Clin Invest 48: 564–​573, 1969.
63. Pilegaard H, Saltin B, and Neufer PD. Exercise induces transient transcriptional activation of the
PGC-​1alpha gene in human skeletal muscle. J Physiol 546: 851–​858, 2003.
64. Poliner LR, Dehmer GJ, Lewis SE, Parkey RW, Blomqvist CG, and Willerson JT. Left ven-
tricular performance in normal subjects: A comparison of the responses to exercise in the upright and
supine positions. Circulation 62: 528–​534, 1980.
65. Powers SK, Howley ET, and Cox R. A differential catecholamine response during prolonged exer-
cise and passive heating. Med Sci Sports Exerc 14: 435–​439, 1982.
66. Puigserver P, Wu Z, Park CW, Graves R, Wright M, and Spiegelman BM. A cold-​inducible
coactivator of nuclear receptors linked to adaptive thermogenesis. Cell 92: 829–​839, 1998.
67. Richardson RS, Wagner H, Mudaliar SR, Henry R, Noyszewski EA, and Wagner PD. Human
VEGF gene expression in skeletal muscle: effect of acute normoxic and hypoxic exercise. Am J Physiol
277: H2247–​H2252, 1999.
68. Richter EA, and Ruderman NB. AMPK and the biochemistry of exercise: implications for human
health and disease. Biochem J 418: 261–​275, 2009.
69. Rockl KS, Hirshman MF, Brandauer J, Fujii N, Witters LA, and Goodyear LJ. Skeletal muscle
adaptation to exercise training: AMP-​activated protein kinase mediates muscle fiber type shift.
Diabetes 56: 2062–​2069, 2007.
70. Rose AJ, Frosig C, Kiens B, Wojtaszewski JF, and Richter EA. Effect of endurance exercise
training on Ca2+ calmodulin-​dependent protein kinase II expression and signalling in skeletal muscle
of humans. J Physiol 583: 785–​795, 2007.

162
163

Endurance phenotype primer

71. San-​Millan I, and Brooks GA. Assessment of metabolic flexibility by means of measuring blood
lactate, fat, and carbohydrate oxidation responses to exercise in professional endurance athletes and
less-​fit individuals. Sports Med 48: 467–​479, 2018.
72. Shibuya M. vascular endothelial growth factor (VEGF) and its receptor (VEGFR) signaling in angio-
genesis: A crucial target for anti-​and pro-​angiogenic therapies. Genes Cancer 2: 1097–​1105, 2011.
73. Steinberg GR, Watt MJ, McGee SL, Chan S, Hargreaves M, Febbraio MA, Stapleton D, and
Kemp BE. Reduced glycogen availability is associated with increased AMPKalpha2 activity, nuclear
AMPKalpha2 protein abundance, and GLUT4 mRNA expression in contracting human skeletal
muscle. Appl Physiol Nutr Metab 31: 302–​312, 2006.
74. Stuewe SR, Gwirtz PA, Agarwal N, and Mallet RT. Exercise training enhances glycolytic and
oxidative enzymes in canine ventricular myocardium. J Mol Cell Cardiol 32: 903–​913, 2000.
75. Takemoto-​Kimura S, Suzuki K, Horigane SI, Kamijo S, Inoue M, Sakamoto M, Fujii
H, and Bito H. Calmodulin kinases: Essential regulators in health and disease. J Neurochem
141: 808–​818, 2017.
76. Tebay LE, Robertson H, Durant ST, Vitale SR, Penning TM, Dinkova-​Kostova AT, and
Hayes JD. Mechanisms of activation of the transcription factor Nrf2 by redox stressors, nutrient cues,
and energy status and the pathways through which it attenuates degenerative disease. Free Radic Biol
Med 88: 108–​146, 2015.
77. Trefts E, Williams AS, and Wasserman DH. Exercise and the regulation of hepatic metabolism.
Prog Mol Biol Transl Sci 135: 203–​225, 2015.
78. van der Zwaard S, van der Laarse WJ, Weide G, Bloemers FW, Hofmijster MJ, Levels K,
Noordhof DA, de Koning JJ, de Ruiter CJ, and Jaspers RT. Critical determinants of combined
sprint and endurance performance: an integrative analysis from muscle fiber to the human body.
FASEB J 32: 2110–​2123, 2018.
79. Vettor R, Valerio A, Ragni M, Trevellin E, Granzotto M, Olivieri M, Tedesco L, Ruocco
C, Fossati A, Fabris R, Serra R, Carruba MO, and Nisoli E. Exercise training boosts eNOS-​
dependent mitochondrial biogenesis in mouse heart: role in adaptation of glucose metabolism. Am J
Physiol 306: E519–​E528, 2014.
80. Volek JS, Freidenreich DJ, Saenz C, Kunces LJ, Creighton BC, Bartley JM, Davitt PM,
Munoz CX, Anderson JM, Maresh CM, Lee EC, Schuenke MD, Aerni G, Kraemer WJ,
and Phinney SD. Metabolic characteristics of keto-​adapted ultra-​endurance runners. Metabolism
65: 100–​110, 2016.
81. Wagner PD. The critical role of VEGF in skeletal muscle angiogenesis and blood flow. Biochem Soc
Trans 39: 1556–​1559, 2011.
82. Wayman GA, Tokumitsu H, Davare MA, and Soderling TR. Analysis of CaM-​kinase signaling
in cells. Cell Calcium 50: 1–​8, 2011.
83. White AT and Schenk S. NAD(+)/​NADH and skeletal muscle mitochondrial adaptations to exer-
cise. Am J Physiol 303: E308–​E321, 2012.
84. Wilson JM, Loenneke JP, Jo E, Wilson GJ, Zourdos MC, and Kim JS. The effects of endurance,
strength, and power training on muscle fiber type shifting. J Strength Cond Res 26: 1724–​1729, 2012.
85. Wright DC, Han DH, Garcia-​Roves PM, Geiger PC, Jones TE, and Holloszy JO. Exercise-​
induced mitochondrial biogenesis begins before the increase in muscle PGC-​1α expression. J Biol
Chem 282: 194–​199, 2007.
86. Wu Z, Puigserver P, Andersson U, Zhang C, Adelmant G, Mootha V, Troy A, Cinti S,
Lowell B, Scarpulla RC, and Spiegelman BM. Mechanisms controlling mitochondrial biogenesis
and respiration through the thermogenic coactivator PGC-​1. Cell 98: 115–​124, 1999.
87. Yang Y, Cimen H, Han MJ, Shi T, Deng JH, Koc H, Palacios OM, Montier L, Bai Y, Tong
Q, and Koc EC. NAD+-​dependent deacetylase SIRT3 regulates mitochondrial protein synthesis by
deacetylation of the ribosomal protein MRPL10. J Biol Chem 285: 7417–​7429, 2010.
88. Zhou B, Conlee RK, Jensen R, Fellingham GW, George JD, and Fisher AG. Stroke volume
does not plateau during graded exercise in elite male distance runners. Med Sci Sports Exerc
33: 1849–​1854, 2001.
89. Zong H, Ren JM, Young LH, Pypaert M, Mu J, Birnbaum MJ, and Shulman GI. AMP kinase
is required for mitochondrial biogenesis in skeletal muscle in response to chronic energy deprivation.
Proc Natl Acad Sci U S A 99: 15983–​15987, 2002.
90. Zouhal H, Jacob C, Delamarche P, and Gratas-​Delamarche A. Catecholamines and the effects
of exercise, training and gender. Sports Med 38: 401–​423, 2008.

163
164

13
HERITABILITY OF
ENDURANCE TRAITS FROM
ANIMAL RESEARCH MODELS
Joshua J. Avila, Sean M. Courtney, and Michael P. Massett

Introduction
A variety of studies in humans indicate that genetics contribute significantly to individual
variation in both baseline or intrinsic exercise capacity and the response to training, with her-
itability estimates of approximately 50% for each of these phenotypes (detailed in Chapter 14).
Despite evidence of a genetic component influencing these phenotypes, the need for large
sample sizes and variation in training paradigms have significantly limited replication of per-
tinent results in humans (22). Alternatively, three strategies have been employed to investigate
the genetic basis of complex traits in rodent models: selective breeding, screening of multiple
inbred strains for exercise phenotypes followed by quantitative trait loci (QTL) analyses, and
candidate gene studies. Transgenic or knockout mice generated to alter expression level of
genes known to be relevant to acute or chronic exercise typically exhibit marked changes
in exercise performance or responses to training (11, 18, 20, 24, 45, 46). However, for the
candidate gene approach to be fruitful, the gene must already be known to be involved in a
pathway related to exercise or training responses (3). In contrast, other approaches can be used
to identify the genetic basis for a complex trait with no a priori bias toward a particular gene,
protein, tissue, or organ system. One such approach is selective breeding for endurance exer-
cise capacity or the change in exercise capacity in response to training. Repeated selection
on a specific exercise phenotype should enrich for alleles associated with the selected trait,
resulting in divergent lines with markedly different phenotypes (10). The phenotypic vari-
ance attributed to additive genetic variance, or narrow-​sense heritability, can be estimated
in this selected population by regression, using phenotype values from parents and their
offspring. Conversely, screening multiple inbred rodent strains for a specific phenotypic
trait takes advantage of the natural genetic and phenotypic variation among strains. One
benefit of using inbred rodents for analysis of complex traits is that strains inbred for at
least 20 generations are essentially homozygous at all loci (55). Thus, under standardized
environmental conditions, phenotypic differences observed among inbred strains are due
predominantly to genetic variation. The proportion of phenotypic variation attributed
to total genetic variation (i.e., broad-​sense heritability) can then be estimated from these
inbred strain populations. Two strains with differing phenotypes can be crossed and the
offspring used for traditional linkage (i.e., QTL) analysis. Although several hundred

164
165

Heritability of animal endurance traits

second-​generation offspring are typically used for linkage analysis, the power to detect
QTL is dependent on multiple factors, including the magnitude of the phenotypic diffe-
rence between the parental inbred strains and the QTL effect size (21). Alternatively, if a
large number of inbred strains have been phenotyped, a genome-​wide association study
(GWAS) can be performed (6, 30, 59). For GWAS in mice, 30 or more inbred strains are
generally sufficient to detect QTL of moderate effect, but the resolution and power are
improved by utilizing a larger number of inbred strains (30, 59). Both linkage analysis and
GWAS have been used to identify genomic regions and/​or putative candidate genes for
endurance exercise capacity or responses to training.
There are three commonly used paradigms for exercise training in rodents –​swimming,
voluntary wheel running, and “forced” wheel or treadmill running –​and each has been used to
study the genetic basis of exercise capacity and responses to training.Voluntary wheel running
and treadmill running are known to induce adaptations in mice associated with endurance
exercise training (2, 12, 15, 28, 43, 56, 60). However, there are inherent differences between the
two paradigms (34) such that there is little correlation between treadmill running performance
and voluntary wheel-​r unning performance among mouse strains (36, 38, 39), suggesting the
genetic factors influencing these two phenotypes are different. The genetic and environmental
control of physical activity in rodents are discussed further in Chapters 5, 7, 8, 9, and 10 of
this book.
This chapter will focus on treadmill running and swimming as a means to assess endurance
exercise capacity and as an exercise paradigm to elicit responses to endurance training. The
majority of these studies have utilized a graded exercise test to assess endurance exercise cap-
acity reported as maximal time, distance, speed, or work. Changes in these variables, i.e., the
difference between post-​training exercise capacity and pretraining exercise capacity, represent
the response to training. Human exercise studies suggest the genes influencing variation in
pretraining or intrinsic endurance exercise capacity are different than those that determine the
responses to training (7, 8, 54) (see Chapter 14). There is support for this from rodent-​based
studies as well (4, 33, 43, 44). Therefore, these phenotypes will be discussed individually in the
subsequent sections.

Intrinsic endurance exercise capacity

Selected strains/​artificial selection
Starting with a founder population of heterogeneous N:NIH rats, Koch and Britton
developed a two-​way artificial selection model of endurance running capacity based on max-
imal treadmill running performance (31). After 11 generations of selection, high-​capacity
runner (HCR) and low-​capacity runner (LCR) selected lines differed by approximately 660
m (~4.5-​fold) in treadmill running endurance (63). By generation 28, the difference increased
to about 840 m (~8.3 fold) (52). In HCR, improved oxygen utilization was attributed to
increased capillary density, higher oxygen extraction, and higher activity of oxidative meta-
bolic enzymes (25, 26). Proteins involved in mitochondrial synthesis and function are also
elevated in muscle from HCR rats (63). Thus, phenotypes associated with endurance exer-
cise training cosegregated with the selected phenotype. Preliminary phenotypic and genetic
analyses were conducted on F1 and F2 populations developed from parental LCR and HCR
lines (52). The exercise phenotype in F1 and F2 generations was intermediate to the parental
lines. Narrow-​sense heritability calculated for maximal running distance based on parental
through F2 generation offspring was 60%. That is somewhat higher than heritability reported

165
166

J.J. Avila et al.

for running distance within the HCR and LCR lines (45%). Thus, the HCR/​LCR lines are a
viable model for QTL mapping and identification of genetic factors contributing to variation
in endurance exercise capacity.

Inbred strain comparisons

Although not specifically selected for exercise capacity, strain differences in endurance exercise
capacity measured by treadmill running and swimming have been reported for inbred mice and
rats (5, 29, 36, 39, 58). In a survey of 11 inbred rat strains, the difference in endurance exercise
performance measured by treadmill running is approximately 2.5-​fold between the highest
(DA/​OlaHsd (DA)) and lowest (Copenhagen (COP)) performing strains (5). In mice, maximum
treadmill running speed is approximately double in the highest (FVB/​NJ (FVB)) versus lowest
(C57BL/​6J (B6)) performing strain (36, 58). Treadmill running distance during a maximal
exercise test varied by 3.6-​fold between ten inbred mouse strains (highest: BALB/​cJ, lowest:
A/​J) (39). Based on differences in treadmill running performance among inbred strains of mice,
broad-​sense heritability estimates for endurance exercise capacity range from 31% to 73% (36,
39), whereas in rats heritability estimates range from 39% to 50% (5, 32). The large differences
between inbred strains and high heritability provide evidence that endurance exercise capacity
is significantly influenced by genetic background.

Genetic mapping
Although transgenic/​ knockout mouse approaches have identified some genes that influ-
ence exercise capacity, the data supporting putative candidate genes based on genome-​wide
approaches is lacking. Using traditional genome-​wide linkage analysis, several QTL for endur-
ance exercise capacity have been identified in rats (61) and mice (40, 42, 44). In the first study
to identify QTL for aerobic running capacity in rats,Ways et al. utilized an F2 population based
on low-​performing COP and high-​performing DA strains for linkage analysis (61). Significant
and suggestive QTL for aerobic running capacity were identified on rat chromosomes 16
(LOD 4.0) and 3 (LOD 2.2), respectively. To confirm these QTL, two congenic strains were
developed by introgressing either a region of chromosome 3 or chromosome 16 from DA rats
into the genetic background of COP rats (62). Exercise capacity was similar between chromo-
some 3 congenic and COP rats. Chromosome 16 congenic rats had significantly greater aer-
obic running capacity compared with COP rats, supporting the presence of a QTL for aerobic
running capacity on this chromosome. The congenic region on chromosome 16 spans nearly
the entire chromosome and requires additional fine-​mapping to identify candidate genes for
aerobic running capacity.
Lightfoot et al. (39) screened several strains of mice for endurance exercise perform-
ance using a graded treadmill test and identified DBA/​2J and BALB/​cJ as low-​and high-​
performing strains, respectively. In a subsequent study, significant and suggestive QTL for
maximal endurance exercise performance were identified on chromosomes X (LOD 2.26)
and 8 (LOD 1.19), respectively, in F2 mice derived from a cross between these strains (40).The
confidence interval for the QTL on mouse chromosome 8 overlaps with the two running
capacity QTL on rat chromosome 16 (the QTL regions on rat chromosome 16 are syn-
tenic to regions of mouse chromosome 8) (61). This concordance between rats and mice
supports the premise that a portion of the genetic regulation of endurance exercise capacity
is conserved among species.

166
167

Heritability of animal endurance traits

Similarly, Massett and Berk performed a strain screen using three inbred mouse strains and
three F1 hybrid strains (43). Significant differences in run time and work were identified among
strains. The highest (FVB) and lowest (B6) performing strains were used to derive an inter-
cross F2 population (44). Significant QTL for pretraining endurance exercise capacity were
identified on chromosomes 14 (LOD 3.72) and 19 (LOD 3.63) in a genome-​wide linkage
scan (Table 13.1). Suggestive QTL (LOD ≤3.4) were identified on chromosomes 1, 2, 3, and
8. Similarly, a reciprocal intercross-​breeding scheme was used to generate F2 mice from inbred
129S1/​SvImJ (high-​performing) and NZW/​LacJ (low-​performing) strains (42). A significant
QTL was identified on chromosome 5 (LOD 4.26) in female mice only (Table 13.1). Suggestive
QTL (LOD ≤3.0) were identified on chromosomes 6, 9, and 12. Subsequently, data from (FVB
× B6)F2 and (NZW × 129S1)F2 populations were combined to increase the resolution of shared
QTL and identify new QTL not identified in individual crosses. In the combined cross-​analysis,
a significant novel QTL for pretraining exercise time was identified on chromosome 12 (LOD
3.6) (Table 13.1). Results from this combined cross-​analysis demonstrate that this approach can
be useful in identifying novel QTL and reducing the confidence interval of previously iden-
tified QTL. The success of combining two mapping populations lends support for the use of
larger and more diverse mapping populations to identify the genetic basis for exercise capacity.
To verify the presence of a QTL for endurance exercise capacity on chromosome 14,
Courtney and Massett utilized a chromosome substitution strain (CSS) derived from A/​J (A) and
B6 inbred strains, denoted as B6.A14 (14). CSS mice are made by substituting a single chromo-
some from a donor inbred strain on the genetic background of a host inbred strain (recipient).
Phenotypic differences between the recipient or background strain mice and CSS mice support
the presence of a QTL on the substituted chromosome for the phenotype being measured.
B6.A14 mice carry chromosome 14 from the low-​performing A donor inbred strain on the
genetic background of a host B6 inbred strain. Exercise capacity in B6.A14 mice is significantly
lower than in B6 mice, suggesting the presence of a QTL on the substituted chromosome for
exercise capacity. Subsequent linkage analysis in a (B6.A14 × B6)F2 population identified a
significant QTL for exercise time (LOD 2.28) and a suggestive QTL for work (LOD 2.19) in
male mice. In contrast, no QTL was identified for exercise time, but a suggestive QTL for work
(LOD 1.8) was identified in female (B6.A14 × B6)F2 mice, suggesting that the genetic archi-
tecture underlying exercise capacity is different in males and females of these strains. Using a
similar strategy, Kvedaras et al. measured exercise capacity in inbred B6 mice and mice from a
CSS carrying A/​J chromosome 10 on a B6 background (B6.A10) (35). Running distance was
significantly lower in B6.A10 mice compared to inbred B6 indicating the presence of at least
one QTL on chromosome 10. No sex differences were reported for this CSS. Collectively, these
results support the presence of QTL for endurance exercise capacity on chromosomes 10 and
14 in A and B6 strains.
As an alternative to traditional linkage analyses, association mapping can be performed on
phenotype data from a large number of inbred strains. To this end, endurance exercise capacity
was measured in 34 strains of inbred mice (13). Exercise time in the highest performing strain
(C58/​J) was 2.7 times that of the lowest performing strain (A/​J). These same strains showed a
16.5-​fold difference in work performed. Subsequent genome-​wide association mapping iden-
tified significant associations (P ≤1 × 10-​6) for exercise time on chromosomes 1, 2, 7, 11, and
13. The average confidence interval for the significant QTL was approximately 450 kb, which
is markedly smaller than QTL identified by linkage analysis. The interval on chromosome 2
contains one gene, Nfatc2. This gene is associated with pathological cardiac hypertrophy (9),
but has yet to be confirmed as a candidate gene for endurance exercise capacity. The genetic

167
168

Table 13.1 Significant quantitative trait loci (QTL) for intrinsic endurance exercise capacity

Study Population Phenotype Chromosome Peak, Mb CI, Mb Genes of

interest

Lightfoot et al., (C × 2J)F2 Time, min 8 74.8 –​

2007
X 127.9 Hsh2d
Massett et al., (FVB × Work, kg:​m 14 7.5 0–​73 Myh6, Myh7,
2009 B6)F2 Myoz1, Sgcg
19 41.1 14–​60 Ankrd2, Btaf1,
Fxn, Plce1
Massett et al., (NZW × Time, min 5 71.0 52–​129 Atp2a2, Myl2,
2015 129S1)F2 Nos1, Sgcb
Combined Time, min 12 105.6 89–​110 Bdkrb2, Tdp1
cross
14 50.8 20–​72 Myh6, Myh7,
Myoz1, Sgcg
Courtney and Inbred Time, min 1 180.7 178–​181 Kif26b
Massett, strains
2012
2 168.4 168.2–​168. 5 Nfactc2
7 16.97 16.97–​16.99 Sae1
11 21.6 21–​23 Mdh1
11 24.5 24.5–​25.6 –​
11 70.3 70.2–​70.8 Chrne, Eno3
13 58.95 Ntrk2
Ways et al., (COP × Distance, m 8* 43.1 33–​61 Sgcz, Slc25a4
2002 DA)F2
8* 18.0 14.5–​25 Angpt2
2* 25.2 25–​49 Crat, Mymk
Ways et al., Congenic Distance, m 8* 9–​75 Irs2, Slc27a1
2007
14* 25–​42 Slc18a3
Courtney and (B6.A14 × Time, min 14 104.9 99–​114 Ednrb
Massett, B6)F2
2014
Kvedaras B6.A10 Distance, m 10 –​
et al., 2017 CSS

Peak, position of QTL peak in megabases (Mb); CI, 95% confidence interval in Mb; (C × 2J)F2,
cross between BALB/​cJ and DBA/​2J strains; (FVB × B6)F2, cross between FVB/​NJ and C57BL/​6J
strains; (NZW × 129S1)F2, cross between NZW/​LacJ and 129S1/​SvImJ strains; Combined cross, data
combined from (FVB × B6)F2 and (NZW × 129S1)F2 populations; (COP × DA)F2, cross between
Copenhagen and DA rat strains; Congenic, COP.DA-​(D16Rat12-​D16Rat90)/​Mco rat congenic strain;
(B6.A14 × B6)F2, cross between C57BL/​6J-​Chr14A/​J/​NaJ and C57BL/​6J strains; B6.A10 CSS, C57BL/​
6J-​Chr10A/​J/​NaJ; *, Rat genome positions have been converted to approximate syntenic regions in mouse
genome. All QTL peak positions and intervals were mapped to build 37.2 of the National Center for
Biotechnology Information mouse genome. Genes of Interest, genes within region associated with
exercise performance, or cardiac, and/​or skeletal muscle structure/​function; Hsh2d, hematopoietic SH2
domain containing; Myh6, myosin, heavy polypeptide 6, cardiac muscle, alpha; Myh7, myosin, heavy
polypeptide 7, cardiac muscle, beta; Myoz1, myozenin 1; Sgcg, sarcoglycan, gamma (dystrophin-​associated
glycoprotein); Ankrd2, ankyrin repeat domain 2 (stretch responsive muscle); Btaf1, B-​TFIID TATA-​box
169

Heritability of animal endurance traits

binding protein associated factor 1; Fxn, frataxin; Plce1, phospholipase C, epsilon 1; Atp2a2, ATPase, Ca2+
transporting, cardiac muscle, slow twitch 2; Myl2, myosin, light polypeptide 2, regulatory, cardiac, slow;
Nos1, nitric oxide synthase 1, neuronal; Sgcb, sarcoglycan, beta (dystrophin-​associated glycoprotein);
Bdkrb2, bradykinin receptor, beta 2; Tdp1, tyrosyl-​DNA phosphodiesterase 1; Kif26b, kinesin family
member 26B; Nfactc2, nuclear factor of activated T cells, cytoplasmic, calcineurin dependent 2; Sae1,
SUMO1 activating enzyme subunit 1; Mdh1, malate dehydrogenase 1, NAD (soluble); Chrne, cholinergic
receptor, nicotinic, epsilon polypeptide; Eno3, enolase 3, beta muscle; Ntrk2, neurotrophic tyrosine kinase,
receptor, type; Sgcz, sarcoglycan zeta; Slc25a4, solute carrier family 25 (mitochondrial carrier, adenine
nucleotide translocator), member 4; Angpt2, angiopoietin 2; Crat, carnitine acetyltransferase; Mymk,
myomaker, myoblast fusion factor; Irs2, insulin receptor substrate 2; Slc27a1, solute carrier family 27 (fatty
acid transporter), member 1; Slc18a3, solute carrier family 18 (vesicular monoamine), member 3; Ednrb,
endothelin receptor type B.

marker D20S857 was linked to the response to exercise training (change in VO2max) in African
Americans in the HERITAGE Family Study (54) and maps to a region on human chromosome
20.This region is syntenic with the putative QTL on mouse chromosome 2. Hence, association
mapping using a large panel of inbred strains provides a higher-​resolution approach to gene
discovery for endurance exercise capacity and related traits.

Responses to exercise training

Selected strains/​artificial selection
Similar to their approach for endurance exercise capacity, Koch et al. created a rat model of
low and high responders to aerobic exercise training to investigate the genetic basis for the
response to exercise training (33). They utilized heterogeneous N/​NIH rats in a two-​way
artificial selection based on changes in maximal running distance to develop high-​response
trainer (HRT) and low-​response trainer (LRT) lines. Rats trained 3 days per week for 8
weeks. The workload progressively increased each session, exposing all rats to the same
absolute training stimulus. After 15 generations of selection, HRT improved their exer-
cise capacity by about 223 m, whereas LRT exhibited a reduction in distance of about 65
m. Estimated narrow-​sense heritability for the response to training at generation 15 was
relatively low (10%) compared with pretraining (34%) and post-​training (43%) exercise cap-
acity. Additionally, the response to training was not dependent upon endurance exercise cap-
acity as both groups had similar pretraining exercise capacity. This was also reflected in the
low genetic correlation between endurance exercise capacity and change in distance (0.14),
suggesting these phenotypes are determined by different genetic factors. Physiologically,
LRT animals have a dysregulated metabolic profile characterized by insulin resistance,
increased adipose mass, and decreased angiogenesis following exercise when compared to
HRT (37). Thus, physiological adaptations to training appear to cosegregate with the change
in distance in HRT and LRT lines. At this time, no genetic analysis utilizing the HRT and
LRT lines has been published.

Inbred strain comparisons

Several studies have compared training responses across multiple inbred strains of mice or
rats. All have consistently reported high-​responding strains having training responses at least
three to four times higher than the lowest responding strains (4, 33, 43). In ten inbred strains
of rats, training responses differed by 3.9-​fold following 8 weeks of treadmill training at the

169
170

J.J. Avila et al.

same absolute workload. Strain means ranged from −80 m (PVG) to +239 m (LEW), while
individual responses ranged from −438 m to +754 m (33). Massett and Berk reported that
responses to 4 weeks of exercise training at the same relative workload (~65% max) varied sig-
nificantly across three inbred (BALB/​cJ, C57BL/​6J, FVB/​NJ) and three F1 hybrid (CB6F1/​
J, B6FF1, FB6F1) mouse strains (43). Among the inbred strains, the response to treadmill
training was greatest in FVB/​NJ and smallest in BALB/​cJ mice. The C57BL/​6J strain had
a low response to training that was similar to BALB/​cJ and significantly less than FVB/​NJ.
Responses in the hybrid strains varied relative to each other and relative to their respective
parental strains. Mean responses in hybrids were greater than both parental strains (CB6F1/​
J), intermediate to their parents (FB6F1), or similar to the lower performing parental strain
(B6FF1), implying that responses to training are complex, significantly influenced by genetic
background, and potentially influenced by maternal inheritance (43). Using a swim training
model, adaptations to 5 weeks of training were compared in six inbred strains of mice (29).The
change in exercise capacity was not reported; however, there was an approximately 6.5-​fold
difference in exercise capacity after training between the highest (632 minutes) and lowest (97
minutes) performing strains. Similar to Massett and Berk, the BALB/​cByJ as well as A/​J were
identified as low-​responding strains. The C57BL/​6J and DBA/​2J were the highest responding
strains in that study.
In a large strain screen, Avila et al. characterized the response to exercise training in 24 gen-
etically diverse strains of inbred mice following 4 weeks of treadmill training (4). As expected,
endurance exercise capacity was significantly different between strains when expressed as time
(range: 21–​42 minutes) and work performed (range: 0.42–​3.89 kg:​m). In response to training,
changes in exercise capacity were significantly different between strains, ranging from −2.2 to
+8.7 minutes. The change in exercise capacity expressed as work performed also varied from
a decrease of 0.24 kg:​m to an increase of 2.30 kg:​m. Collectively, the surveys of inbred rat and
mouse strains indicate that responses to training vary markedly among strains and are signifi-
cantly influenced by genetic background.
Despite reporting similar ranges of responses among strains, the estimated genetic contri-
bution to this variation is not consistent. Estimated heritability for responses to training in
inbred rats was 13% (33).This value is similar to that observed in selected lines developed from
heterogeneous N/​NIH rats. In contrast, genetic background accounted for 37–​58% of the
variance in responses to training in inbred mice (4, 43). In (FVB × B6)F2 mice estimated her-
itability for the change in work was 53% (44). The estimated contribution by genetic factors
to training responses in mice is comparable to that reported for change in VO2max following
exercise training in humans (7, 23, 50). The differences between heritability estimates for
responses to training in rats and mice can be partly attributed to differences in environ-
mental variance. The training paradigms differed significantly between rat and mouse models
(8 weeks vs. 4 weeks, absolute vs. relative intensity). Therefore, it is not unexpected that
the contributions of environmental and genetic factors to phenotypic variance might differ.
Furthermore, heritability estimates are population specific. Regardless, studies of training
responses from rats and mice indicate that this response is a heritable trait with a significant
genetic component.

Quantitative trait locus mapping

While the majority of linkage and association studies have focused on endurance exercise cap-
acity (Table 13.1), a few studies have been conducted to identify the genetic basis for the response

170
171

Heritability of animal endurance traits

to training. An intercross breeding scheme between low-​responding B6 and high-​responding

FVB strains was used to generate (FVB × B6)F2 offspring for linkage analysis (44). After 4 weeks
of treadmill training, changes in exercise capacity ranged from −1.67 to +4.55 kg:m. Suggestive
QTL for the change in work were identified on chromosomes 11 (LOD 2.30) and 14 (LOD
2.25) (Table 13.2). In a subsequent study, a reciprocal intercross breeding strategy was used to
generate an F2 population from low-​responding NZW and high-​responding 129S1 strains (42).
Training responses in (NZW × 129S1)F2 mice were highly variable with a mean increase in
exercise capacity of 1.5 ± 3.7 minutes (range: −10.1 to 12.4 minutes). Cross direction (female
NZW × male 129S1 or female 129S1 × male NZW) did not significantly affect the response
to training. Suggestive QTL for response to training were identified on chromosomes 1 (LOD
2.47) and 6 (LOD 2.73) (Table 13.2) with the QTL on chromosome 1 identified in female
mice only.
Both linkage studies for exercise training responses incorporated relatively small populations
of animals (<300 each). Therefore, a combined cross-​analysis was performed utilizing both F2
populations (42). Genome-​wide linkage analysis yielded a significant QTL on chromosome 6
(LOD 3.7) for change in exercise time (Table 13.2). Several suggestive QTL (LOD <3.0) also
were identified using this approach (Table 13.2). Although the confidence interval for the QTL
on chromosome 6 was relatively large (71 Mb), these findings support the use of combining
data from multiple crosses to identify QTL for responses to training not observed in smaller
mapping populations.

Mouse–​rat–​human comparative genomics

For most of the animal studies discussed above, the QTL intervals for endurance exercise cap-
acity and responses to training are relatively large. These large intervals contain hundreds of
genes, many of which are not obviously associated with exercise capacity or training responses.
One approach to prioritizing regions or genes for future investigation is to identify regions of
chromosomal synteny across species for endurance exercise and training response QTL (17). For
example, QTL for endurance exercise capacity and responses to training on mouse chromo-
some 14 are concordant with regions of the human genome identified in the HERITAGE
Family Study linked to exercise capacity in the sedentary state and training-​induced changes
in maximal oxygen consumption (ΔVO2max) (8, 42, 44, 54). Single nucleotide polymorphisms
(SNPs) associated with training responses in humans also map to regions sytenic with QTL on
chromosome 14 (53, 57). One of these SNPs, in the gene mitochondrial intermediate peptidase
(MIPEP), was used to predict training responses in humans based on baseline gene expression
in skeletal muscle (57). This collection of evidence suggests that mouse chromosome 14 should
be considered for more detailed analyses of the genetic basis for endurance exercise capacity
and responses to training.
Similarly, QTL intervals on chromosomes 6, 8, and 19 align with concordant QTL from
other species. QTL for endurance exercise capacity in rats and mice overlap on chromosome 8
(40, 61). Several linkage markers for VO2max in the sedentary state in humans (8, 54) also fall
within these QTL intervals suggesting potential concordance among mouse, rat, and human
QTL. On chromosome 19, QTL for endurance exercise capacity and responses to training
identified in mice contain genes associated with training responses (BTAF1, PIP5K1B) (16,
57), and fitness (ANKRD22) (51) in humans. The significant QTL for responses to training in
mice on chromosome 6 contains Cpvl, a carboxypeptidase gene. This transcript was included in
a group of genomic predictors for responses to exercise training in humans (57). Collectively,

171
172

J.J. Avila et al.

Table 13.2 Significant and suggestive quantitative trait loci (QTL) for responses to endurance exercise
training

Study Population Phenotype Chromosome Peak, CI, Mb Genes of interest

Mb

Massett et al., (FVB × B6)F2 Change in work, 11 72.6 47–​109 Ace, Myh1, Myocd,
2009 kg:​m Myh4, Slc2a4
14 69.8 0–​122 Myh6, Myh7,
Myoz1, Sgcg
Massett et al., (NZW × Change in time, 1 42.7 21–​166 Mstn, Myog Tnnt2
2015 129S1)F2 min
6 86.8 33–​127 Chrm2, Kcna5,
Mtpn, Pparg
Combined Change in time, 6 58.2* 41–​112 Cav3, Dysf
cross min
8 95.3 82–​117 Mmp2, Rln3
11 42 33–​113 Ace, Myh1, Myocd,
Myh4, Slc2a4
19 27.5 9–​37 Ankrd1, Anxa1,
Btaf1, Fxn

Peak, position of QTL peak in megabases (Mb); CI, 95% confidence interval in Mb; (FVB × B6)
F2, cross between FVB/​NJ and C57BL/​6J strains; (NZW × 129S1)F2, cross between NZW/​LacJ and
129S1/​SvImJ strains; Combined cross, data combined from (FVB × B6)F2 and (NZW × 129S1)F2
populations; *, indicates significant QTL. All QTL peak positions and intervals were mapped to build
37.2 of the National Center for Biotechnology Information mouse genome. Genes of Interest, genes
within region associated with exercise performance, or cardiac, and/​or skeletal muscle structure/​function;
Ace, angiotensin I converting enzyme (peptidyl-​dipeptidase A) 1, Myh1, myosin, heavy polypeptide 1,
skeletal muscle, adult; Myh4, myosin, heavy polypeptide 4, skeletal muscle; Myocd, myocardin; Slc2a4,
solute carrier family 2 (facilitated glucose transporter), member 4; Myh6, myosin, heavy polypeptide
6, cardiac muscle, alpha; Myh7, myosin, heavy polypeptide 7, cardiac muscle, beta; Myoz1, myozenin 1;
Sgcg, sarcoglycan, gamma (dystrophin-​associated glycoprotein); Mstn, myostatin; Myog, myogenin; Tnnt2,
troponin T2, cardiac; Chrm2, cholinergic receptor, muscarinic 2, cardiac; Kcna5, potassium voltage-​gated
channel, shaker-​related subfamily, member 5; Mtpn, myotrophin; Pparg, peroxisome proliferator activated
receptor gamma; Cav3, caveolin 3; Dysf, dysferlin; Mmp2, matrix metallopeptidase 2; Rln3, relaxin 3;
Ankrd1, ankyrin repeat domain 1 (cardiac muscle); Anxa1, annexin A1; Btaf1, B-​TFIID TATA-​box
binding protein associated factor 1.

these examples suggest that there is some concordance among mouse, rat, and human QTL for
endurance exercise capacity and responses to training. However, the large confidence intervals
from linkage analyses in rodents require some fine mapping to effectively utilize comparative
mapping as an approach for candidate gene discovery.

Summary and future directions

Data from animal models confirm that endurance exercise capacity and responses to training
are heritable traits and these traits are determined by multiple genetic factors. The high herit-
ability estimates for endurance exercise capacity in inbred mice suggest that a high percentage
of the variation in exercise capacity is determined by genetic factors. The heritability estimates
for the response to training are lower, likely reflecting the greater influence of environmental

172
173

Heritability of animal endurance traits

factors and gene by environment interactions on the response to training.This complex nature
of the responses to exercise training has made identifying genetic factors challenging. The
two inbred line crosses yielded moderately significant QTL for responses to training. The
combined cross-​analyses strengthened the evidence for a QTL on chromosome 6, but the
region is quite large.The wide variation in training responses in the inbred mouse strains tested
thus far suggests that additional loci can be identified using the GWAS approach. Any overlap
between the relatively small regions identified using this approach and previously identified
QTL would strengthen the evidence that these regions contain putative candidate genes influ-
encing responses to training.
Some of the difficulty in identifying genetic factors is related to the complex nature of the
phenotype. To maximize the ability to detect genetic variants and identify candidate genes,
several factors should be considered going forward. A well-​defined phenotype is paramount to
successful gene identification. The published exercise testing and training paradigms are quite
variable (4, 19, 28, 33, 49).Therefore, some consideration should be given to better defining the
measurement of exercise capacity in mice and the outcome variable (i.e., time, work, oxygen
consumption) used to define it. Similarly, an appropriate training regimen eliciting maximal
changes in exercise capacity, but capable of being utilized for large numbers of animals, needs
to be identified. Additionally, assessment of the adaptations to training in mice should be
expanded to better identify the appropriate target organ for gene and protein mapping studies.
For example, functional changes in cardiac performance with endurance training have yet to be
assessed in multiple strains of mice. Finally, inbred strain surveys have typically utilized animals
of one sex (4, 13, 29, 36, 39, 43, 58). However, a few sex-​specific QTL were identified in F2
populations (14, 42), suggesting that the genetic architecture underlying these traits differs.
Therefore, expanding our knowledge of the contribution of sex to these exercise phenotypes
could facilitate discovery of the genetic networks underlying endurance exercise capacity and
the responses to training.
To date, individual line crosses and GWAS approaches have been used identify genomic
regions of interest. Although the line crosses have yielded some significant loci, the identified
regions contain hundreds of genes. The GWAS approach applied to endurance exercise cap-
acity identified several small regions containing only a few genes, but with 34 inbred strains
was still somewhat underpowered (6, 30). The current data for training responses in inbred
strains will likely yield similar results but are underpowered for identifying genetic factors
influencing training responses. Future studies should focus on the GWAS approach, but also
utilize the recently developed mouse and rat models designed to maximize genetic heterogen-
eity. Approximately 500 or less Diversity Outbred and/​or Collaborative Cross mice have been
used to identify QTL for physiological and behavioral traits (1, 41). A similar number might
be sufficient for genetic analyses of endurance exercise capacity and/​or training responses.
Heterogeneous stock rats also could be used for higher resolution mapping; however, the
number of animals required (>1000) could be prohibitive for exercise training studies (64).
Conversely, examining smaller populations as a whole with meta-​analytic approaches could cir-
cumvent the need for conducting such large training studies (27). Finally, combining genome-​
wide gene expression with genetic linkage or association mapping has successfully identified
candidate genes for metabolic and disease-​related traits (47, 48). Utilizing such a systems biology
approach could expedite identification of candidate genes or proteins underlying the variation
in endurance exercise capacity and the responses to training, and better define the genetic archi-
tecture underlying these complex phenotypes.

173
174

J.J. Avila et al.

References
1. Abu-​Toamih Atamni HJ, Ziner Y, Mott R, Wolf L, Iraqi FA. Glucose tolerance female-​specific
QTL mapped in collaborative cross mice. Mamm Genome 28: 20–​30, 2017.
2. Allen DL, Harrison BC, Maass A, Bell ML, Byrnes WC, Leinwand LA. Cardiac and skeletal
muscle adaptations to voluntary wheel running in the mouse. J Appl Physiol 90: 1900–​1908, 2001.
3. Andreux PA, Williams EG, Koutnikova H, Houtkooper RH, Champy M-​F, Henry H,
Schoonjans K, Williams RW, Auwerx J. Systems genetics of metabolism: The use of the BXD
murine reference panel for multiscalar integration of traits. Cell 150: 1287–​1299, 2012.
4. Avila JJ, Kim SK, Massett MP. Differences in exercise capacity and responses to training in 24
inbred mouse strains. Front Physiol 8: 974, 2017.
5. Barbato JC, Koch LG, Darvish A, Cicila GT, Metting PJ, Britton SL. Spectrum of aerobic
endurance running performance in eleven inbred strains of rats. J Appl Physiol 85: 530–​536, 1998.
6. Bennett BJ, Farber CR, Orozco L, Kang HM, Ghazalpour A, Siemers N, Neubauer M,
Neuhaus I, Yordanova R, Guan B, Truong A, Yang W, He A, Kayne P, Gargalovic P,
Kirchgessner T, Pan C, Castellani LW, Kostem E, Furlotte N, Drake TA, Eskin E, Lusis AJ.
A high-​resolution association mapping panel for the dissection of complex traits in mice. Genome Res
20: 281–​290, 2010.
7. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, Pérusse L, Leon AS, Rao DC.
Familial aggregation of VO(2max) response to exercise training: results from the HERITAGE Family
Study. J Appl Physiol 87: 1003–​1008, 1999.
8. Bouchard C, Rankinen T, Chagnon YC, Rice T, Pérusse L, Gagnon J, Borecki I, An P, Leon
AS, Skinner JS, Wilmore JH, Province M, Rao DC. Genomic scan for maximal oxygen uptake
and its response to training in the HERITAGE Family Study. J Appl Physiol 88: 551–​559, 2000.
9. Bourajjaj M, Armand A-​S, da Costa Martins PA, Weijts B, van der Nagel R, Heeneman
S, Wehrens XH, De Windt LJ. NFATc2 is a necessary mediator of calcineurin-​dependent cardiac
hypertrophy and heart failure. J Biol Chem 283: 22295–​22303, 2008.
10. Britton SL, Koch LG. Animal genetic models for complex traits of physical capacity. Exerc Sport Sci
Rev 29: 7–​14, 2001.
11. Calvo JA, Daniels TG, Wang X, Paul A, Lin J, Spiegelman BM, Stevenson SC, Rangwala SM.
Muscle-​specific expression of PPARgamma coactivator-​1alpha improves exercise performance and
increases peak oxygen uptake. J Appl Physiol 104: 1304–​1312, 2008.
12. Chow LS, Greenlund LJ, Asmann YW, Short KR, McCrady SK, Levine JA, Nair KS. Impact
of endurance training on murine spontaneous activity, muscle mitochondrial DNA abundance, gene
transcripts, and function. J Appl Physiol 102: 1078–​1089, 2007.
13. Courtney SM, Massett MP. Identification of exercise capacity QTL using association mapping in
inbred mice. Physiol Genomics 44: 948–​955, 2012.
14. Courtney SM, Massett MP. Effect of chromosome substitution on intrinsic exercise capacity in
mice. F1000Res 3: 9, 2014.
15. De Angelis K, Wichi RB, Jesus WRA, Moreira ED, Morris M, Krieger EM, Irigoyen
MC. Exercise training changes autonomic cardiovascular balance in mice. J Appl Physiol
96: 2174–​2178, 2004.
16. Dias RG, Silva MSM, Duarte NE, Bolani W, Alves CR, Junior JRL, da Silva JL, de Oliveira
PA, Alves GB, de Oliveira EM, Rocha CS, Marsiglia JDC, Negrao CE, Krieger EM, Krieger
JE, Pereira AC. PBMCs express a transcriptome signature predictor of oxygen uptake responsiveness
to endurance exercise training in men. Physiol Genomics 47: 13–​23, 2015.
17. DiPetrillo K, Wang X, Stylianou IM, Paigen B. Bioinformatics toolbox for narrowing rodent
quantitative trait loci. Trends Genet 21: 683–​692, 2005.
18. Fan W, Waizenegger W, Lin CS, Sorrentino V, He M-​X, Wall CE, Li H, Liddle C, Yu RT,
Atkins AR, Auwerx J, Downes M, Evans RM. Pparδ promotes running endurance by preserving
glucose. Cell Metab 25: 1186–​1193.e4, 2017.
19. Ferreira JCB, Rolim NPL, Bartholomeu JB, Gobatto CA, Kokubun E, Brum PC.
Maximal lactate steady state in running mice: effect of exercise training. Clin Exp Pharmacol Physiol
34: 760–​765, 2007.

174
175

Heritability of animal endurance traits

20. Fewell JG, Osinska H, Klevitsky R, Ng W, Sfyris G, Bahrehmand F, Robbins J. A tread-

mill exercise regimen for identifying cardiovascular phenotypes in transgenic mice. Am J Physiol
273: H1595–​605, 1997.
21. Flint J, Valdar W, Shifman S, Mott R. Strategies for mapping and cloning quantitative trait genes
in rodents. Nat Rev Genet 6: 271–​286, 2005.
22. Hagberg JM, Rankinen T, Loos RJF, Pérusse L, Roth SM, Wolfarth B, Bouchard C. Advances
in exercise, fitness, and performance genomics in 2010. Med Sci Sports Exerc 43: 743–​752, 2011.
23. Hamel P, Simoneau JA, Lortie G, Boulay MR, Bouchard C. Heredity and muscle adaptation to
endurance training. Med Sci Sports Exerc 18: 690–​696, 1986.
24. Haubold KW, Allen DL, Capetanaki Y, Leinwand LA. Loss of desmin leads to impaired volun-
tary wheel running and treadmill exercise performance. J Appl Physiol 95: 1617–​1622, 2003.
25. Henderson KK, Wagner H, Favret F, Britton SL, Koch LG, Wagner PD, Gonzalez NC.
Determinants of maximal O(2) uptake in rats selectively bred for endurance running capacity. J Appl
Physiol 93: 1265–​1274, 2002.
26. Howlett RA, Gonzalez NC, Wagner HE, Fu Z, Britton SL, Koch LG, Wagner PD. Selected
contribution: skeletal muscle capillarity and enzyme activity in rats selectively bred for running
endurance. J Appl Physiol 94: 1682–​1688, 2003.
27. Kang EY, Han B, Furlotte N, Joo JWJ, Shih D, Davis RC, Lusis AJ, Eskin E. Meta-​analysis
identifies gene-​by-​environment interactions as demonstrated in a study of 4,965 mice. PLoS Genet
10: e1004022, 2014.
28. Kemi OJ, Loennechen JP, Wisløff U, Ellingsen Ø. Intensity-​controlled treadmill running in
mice: cardiac and skeletal muscle hypertrophy. J Appl Physiol 93: 1301–​1309, 2002.
29. Kilikevicius A, Venckunas T, Zelniene R, Carroll AM, Lionikaite S, Ratkevicius A, Lionikas
A. Divergent physiological characteristics and responses to endurance training among inbred mouse
strains. Scand J Med Sci Sports 23: 657–​668, 2013.
30. Kirby A, Kang HM, Wade CM, Cotsapas C, Kostem E, Han B, Furlotte N, Kang EY,
Rivas M, Bogue MA, Frazer KA, Johnson FM, Beilharz EJ, Cox DR, Eskin E, Daly MJ.
Fine mapping in 94 inbred mouse strains using a high-​density haplotype resource. Genetics
185: 1081–​1095, 2010.
31. Koch LG, Britton SL. Artificial selection for intrinsic aerobic endurance running capacity in rats.
Physiol Genomics 5: 45–​52, 2001.
32. Koch LG, Meredith TA, Fraker TD, Metting PJ, Britton SL. Heritability of treadmill running
endurance in rats. Am J Physiol 275: R1455–​60, 1998.
33. Koch LG, Pollott GE, Britton SL. Selectively bred rat model system for low and high response to
exercise training. Physiol Genomics 45: 606–​614, 2013.
34. Kregel KC, Allen DL, Booth FW, Fleshner MR, Henriksen EJ, Musch TI, O’Leary DS, Parks
CM, Poole DC, Ra’anan AW, Sheriff DD. Resource Book for the Design of Animal Exercise Protocols.
Bethesda, MD: American Physiological Society, 2006.
35. Kvedaras M, Minderis P, Fokin A, Ratkevicius A, Venckunas T, Lionikas A. Forced running
endurance is influenced by gene(s) on mouse chromosome 10. Front Physiol 8: 9, 2017.
36. Lerman I, Harrison BC, Freeman K, Hewett TE, Allen DL, Robbins J, Leinwand LA. Genetic
variability in forced and voluntary endurance exercise performance in seven inbred mouse strains. J
Appl Physiol 92: 2245–​2255, 2002.
37. Lessard SJ, Rivas DA, Alves-​Wagner AB, Hirshman MF, Gallagher IJ, Constantin-​Teodosiu
D, Atkins R, Greenhaff PL, Qi NR, Gustafsson T, Fielding RA, Timmons JA, Britton SL,
Koch LG, Goodyear LJ. Resistance to aerobic exercise training causes metabolic dysfunction and
reveals novel exercise-​regulated signaling networks. Diabetes 62: 2717–​2727, 2013.
38. Lightfoot JT, Turner MJ, Daves M, Vordermark A, Kleeberger SR. Genetic influence on daily
wheel running activity level. Physiol Genomics 19: 270–​276, 2004.
39. Lightfoot JT, Turner MJ, Debate KA, Kleeberger SR. Interstrain variation in murine aerobic
capacity. Med Sci Sports Exerc 33: 2053–​2057, 2001.
40. Lightfoot JT, Turner MJ, Knab AK, Jedlicka AE, Oshimura T, Marzec J, Gladwell W, Leamy
LJ, Kleeberger SR. Quantitative trait loci associated with maximal exercise endurance in mice. J
Appl Physiol 103: 105–​110, 2007.

175
176

J.J. Avila et al.

41. Logan RW, Robledo RF, Recla JM, Philip VM, Bubier JA, Jay JJ, Harwood C, Wilcox T,
Gatti DM, Bult CJ, Churchill GA, Chesler EJ. High-​precision genetic mapping of behavioral
traits in the diversity outbred mouse population. Genes Brain Behav 12: 424–​437, 2013.
42. Massett MP, Avila JJ, Kim SK. Exercise capacity and response to training quantitative trait loci
in a NZW X 129S1 intercross and combined cross analysis of inbred mouse strains. PLoS One
10: e0145741, 2015.
43. Massett MP, Berk BC. Strain-​dependent differences in responses to exercise training in inbred and
hybrid mice. Am J Physiol Regul Integr Comp Physiol 288: R1006–​13, 2005.
44. Massett MP, Fan R, Berk BC. Quantitative trait loci for exercise training responses in FVB/​NJ and
C57BL/​6J mice. Physiol Genomics 40: 15–​22, 2009.
45. Momken I, Lechêne P, Ventura-​Clapier R, Veksler V. Voluntary physical activity alterations in
endothelial nitric oxide synthase knockout mice. Am J Physiol Heart Circ Physiol 287: H914–​H920, 2004.
46. Nie Y, Sato Y, Wang C, Yue F, Kuang S, Gavin TP. Impaired exercise tolerance, mitochondrial
biogenesis, and muscle fiber maintenance in miR-​133a-​deficient mice. FASEB J 30: 3745–​3758, 2016.
47. Parker CC, Gopalakrishnan S, Carbonetto P, Gonzales NM, Leung E, Park YJ, Aryee E,
Davis J, Blizard DA, Ackert-​Bicknell CL, Lionikas A, Pritchard JK, Palmer AA. Genome-​
wide association study of behavioral, physiological and gene expression traits in outbred CFW mice.
Nat Genet 48: 919–​926, 2016.
48. Parks BW, Sallam T, Mehrabian M, Psychogios N, Hui ST, Norheim F, Castellani LW,
Rau CD, Pan C, Phun J, Zhou Z, Yang W-​P, Neuhaus I, Gargalovic PS, Kirchgessner TG,
Graham M, Lee R, Tontonoz P, Gerszten RE, Hevener AL, Lusis AJ. Genetic architecture of
insulin resistance in the mouse. Cell Metab 21: 334–​346, 2015.
49. Petrosino JM, Heiss VJ, Maurya SK, Kalyanasundaram A, Periasamy M, LaFountain RA,
Wilson JM, Simonetti OP, Ziouzenkova O. Graded maximal exercise testing to assess mouse
cardio-​metabolic phenotypes. PLoS One 11: e0148010, 2016.
50. Prud’homme D, Bouchard C, Leblanc C, Landry F, Fontaine E. Sensitivity of maximal aerobic
power to training is genotype-​dependent. Med Sci Sports Exerc 16: 489–​493, 1984.
51. Rampersaud E, Nathanson L, Farmer J, Meshbane K, Belton RL, Dressen A, Cuccaro M,
Musto A, Daunert S, Deo S, Hudson N, Vance JM, Seo D, Mendez A, Dykxhoorn DM,
Pericak-​Vance MA, Goldschmidt-​Clermont PJ. Genomic signatures of a global fitness index in
a multi-​ethnic cohort of women. Ann Hum Genet 77: 147–​157, 2013.
52. Ren Y-​Y, Overmyer KA, Qi NR, Treutelaar MK, Heckenkamp L, Kalahar M, Koch LG,
Britton SL, Burant CF, Li JZ. Genetic analysis of a rat model of aerobic capacity and metabolic
fitness. PLoS One 8: e77588, 2013.
53. Rice TK, Sarzynski MA, Sung YJ, Argyropoulos G, Stütz AM, Teran-​Garcia M, Rao DC,
Bouchard C, Rankinen T. Fine mapping of a QTL on chromosome 13 for submaximal exercise
capacity training response: The HERITAGE Family Study. Eur J Appl Physiol 112: 2969–​2978, 2012.
54. Rico-​Sanz J, Rankinen T, Rice T, Leon AS, Skinner JS, Wilmore JH, Rao DC, Bouchard C.
Quantitative trait loci for maximal exercise capacity phenotypes and their responses to training in the
HERITAGE Family Study. Physiol Genomics 16: 256–​260, 2004.
55. Silver LM. Mouse Genetics: Concepts and Applications. New York: Oxford University Press, 1995.
56. Swallow JG, Garland T, Carter PA, Zhan WZ, Sieck GC. Effects of voluntary activity and gen-
etic selection on aerobic capacity in house mice (Mus domesticus). J Appl Physiol 84: 69–​76, 1998.
57. Timmons JA, Knudsen S, Rankinen T, Koch LG, Sarzynski M, Jensen T, Keller P, Scheele
C, Vollaard NBJ, Nielsen S, Akerström T, MacDougald OA, Jansson E, Greenhaff PL,
Tarnopolsky MA, van Loon LJC, Pedersen BK, Sundberg CJ, Wahlestedt C, Britton SL,
Bouchard C. Using molecular classification to predict gains in maximal aerobic capacity following
endurance exercise training in humans. J Appl Physiol 108: 1487–​1496, 2010.
58. Totsuka Y, Nagao Y, Horii T, Yonekawa H, Imai H, Hatta H, Izaike Y, Tokunaga T, Atomi
Y. Physical performance and soleus muscle fiber composition in wild-​derived and laboratory inbred
mouse strains. J Appl Physiol 95: 720–​727, 2003.
59. Wang J, Liao G, Usuka J, Peltz G. Computational genetics: from mouse to human? Trends Genet
21: 526–​532, 2005.
60. Waters RE, Rotevatn S, Li P, Annex BH, Yan Z. Voluntary running induces fiber type-​specific
angiogenesis in mouse skeletal muscle. Am J Physiol Cell Physiol 287: C1342–​C1348, 2004.
61. Ways JA, Cicila GT, Garrett MR, Koch LG. A genome scan for loci associated with aerobic
running capacity in rats. Genomics 80: 13–​20, 2002.

176
177

Heritability of animal endurance traits

62. Ways JA, Smith BM, Barbato JC, Ramdath RS, Pettee KM, DeRaedt SJ, Allison DC, Koch
LG, Lee SJ, Cicila GT. Congenic strains confirm aerobic running capacity quantitative trait loci on
rat chromosome 16 and identify possible intermediate phenotypes. Physiol Genomics 29: 91–​97, 2007.
63. Wisløff U, Najjar SM, Ellingsen O, Haram PM, Swoap S, Al-​Share Q, Fernström M, Rezaei
K, Lee SJ, Koch LG, Britton SL. Cardiovascular risk factors emerge after artificial selection for low
aerobic capacity. Science 307: 418–​420, 2005.
64. Woods LCS, Mott R. Heterogeneous stock populations for analysis of complex traits. Methods Mol
Biol 1488: 31–​44, 2017.

177
178

14
HERITABILITY OF
ENDURANCE TRAITS FROM
HUMAN RESEARCH MODELS
Jacob L. Barber and Mark A. Sarzynski

Chapter 13 summarized published work related to the heritability of endurance traits at baseline
and in response to exercise training in animal models. This chapter will summarize published
work that quantifies genetic contributions to endurance-​related traits both in the intrinsic state
(i.e., in the sedentary state at baseline) and following exercise training using traditional family
and twin study approaches in humans. Please refer to Chapter 12 for a primer on the endurance
phenotype.

Heritability of intrinsic levels of endurance traits

There are large interindividual differences in endurance-​related phenotypes, which may be
affected by age, sex, physical activity, and fitness levels (13). Additionally, numerous twin and
family studies have provided evidence that endurance-​related traits clearly aggregate in families
and have a significant genetic component in the intrinsic state.Twin and family studies represent
the ideal design to investigate the influences of genes and the environment on a trait given that
family members not only share similar genetic profiles but are also exposed to similar environ-
mental influences.The rationale is that the trait levels will be more similar for individuals having
the same genes (monozygotic (MZ) twins) or sharing about 50% of their genes (dizygotic (DZ)
twins, parent–​offspring, siblings) compared to unrelated individuals (between twin pairs or fam-
ilies). From comparisons of intrapair differences between MZ and DZ twins, it is possible to
separate the relative contribution of genotype and environment for any attribute by deriving
a coefficient of heritability. Heritability is defined as the proportion of observed phenotypic
variance in a population that is attributable to individual genetic differences. If MZ twins show
more similarity for a given trait than DZ twins, this provides evidence of a significant genetic
component of the trait, whereas if MZ and DZ twins exhibit equal similarity for a trait, it is
likely the environment influences the trait more than genetic factors. Similarly, in family studies
a pattern of significant correlations between parents and offspring and between siblings, but
not between spouses, suggests that the familial resemblance is primarily due to genetic factors.
For example, a study of 42 brothers and 66 DZ and 106 MZ twins of both sexes found that
the interclass correlations of absolute maximal oxygen uptake or VO2max (l/​min) were 0.57,
0.84, and 0.92 (all P <0.01), respectively (12). The HERITAGE Family Study examined the

178
179

Heritability of human endurance traits

5000 Maximal heritability: 51%

4500

4000
Baseline VO2max (ml/min)

3500

3000

2500

2000

1500

1000 Overall mean for all subjects:

Family mean
500 2466 ml O2
0
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96
Families ordered by mean VO2max value from lowest to highest

Figure 14.1 Mean baseline VO2max across white families from the HERITAGE Family Study, with
families ranked by family mean.

familial resemblance for VO2max in 426 sedentary adults from 86 nuclear families. The authors
found that intrinsic levels of VO2max aggregate in families, as there was about 2.7 times more
variance between different families than within families for adjusted VO2max (9) (Figure 14.1).
It can be seen in Figure 14.1 that some families in HERITAGE tended to have below-​average
VO2max values, while others had above-​average values.
Furthermore, the authors estimated that the genetic heritability of intrinsic VO2max was 51%
in HERITAGE (9).This heritability estimate is similar to the aforementioned study of 106 MZ
twins and 66 DZ twins that found heritability estimates ranging from 38% to 47% for VO2max
(12). However, to date the calculated heritability estimates of VO2max show wide variation
across studies, with twin studies tending to report higher estimates (Table 14.1) (18, 24, 26, 33).
A meta-​analysis of seven twin and sibling studies in children to young adults (N=1088) found
weighted heritability estimates of 59% and 72% for absolute (ml/​min) and relative (ml/​kg/​min)
VO2max (33). In 2017, a systematic review and meta-​analysis of twin and family studies was
performed on the heritability of intrinsic VO2max (24). Six twin studies and two family studies
were included in the meta-​analysis, which found that the weighted means of the heritability
of absolute VO2max and VO2max adjusted for body weight and adjusted for fat-​free mass were
0.68 (95% confidence interval (CI) 0.59–​0.77), 0.56 (95% CI 0.47–​0.65) (Figure 14.2), and
0.44 (95% CI 0.13–​0.75), respectively (24). The authors found that estimation method, study
type, and exercise mode did not significantly influence the heritability estimates of VO2max,
while the heritability estimates were significantly higher in studies comprised of only males
(0.69; 95% CI 0.55–​0.83) compared to studies including both males and females (0.49; 95%
CI 0.42–​0.56). Thus, sex differences appear to partially explain the heterogeneity in the herit-
ability estimates of absolute and relative VO2max. This could be the result of sex differences in
body composition, as heritability estimates of absolute VO2max tend to be higher compared to
those adjusting for body weight or composition. Importantly, no study of only females has been
performed to better address and compare sex differences in the heritability of VO2max. Results

179
180

J.L. Barber and M.A. Sarzynski

Table 14.1 Heritability estimates for endurance-​related traits in the intrinsic/​untrained state

Trait Study design Number of Heritability References

studies

VO2max Nuclear families 6 0.20–​0.52 18, 24, 26, 33

(adjusted for body weight) Twins 8 0.38–​0.93
Submaximal aerobic Nuclear families 4 0.00–​0.74 5, 18, 24, 26, 33
performance Twins 3 0.00–​0.90
(e.g.,VO2 at 50% or 80%
of maximum)

Study Heritability (95% CI) %Weight

Klissouras et al. 1971 0.93 (0.59, 1.28) 5.04
Bouchard et al. 1986 0.39 (0.21, 0.57) 11.53
Fagard et al. 1991 0.78 (0.53, 1.03) 8.14
Sundet et al. 1994 0.62 (0.57, 0.67) 22.37
Rodas et al. 1998 0.52 (0.09, 0.95) 3.49
Bouchard et al. 1998 0.52 (0.40, 0.64) 16.85
Foraita et al. 2015 0.42 (0.28, 0.56) 14.66
Schutte et al. 2016 0.55 (0.45, 0.65) 17.91
Overall (I2 = 63.6%, P = 0.007) 0.56 (0.47, 0.65) 100.00
0 0.5 1

Figure 14.2 Meta-​analysis of heritability estimates of intrinsic VO2max adjusted by body weight.

from HERITAGE found the strongest familial correlations of intrinsic VO2max between
mothers and daughters and that the best heritability models incorporated maternal inheritance
with the father’s contribution being environmental (9). The authors estimated the maternal
heritability of intrinsic VO2max ranged from 29% to 36% and most likely represents the con-
tribution of mitochondrial DNA, which is transmitted by the mother to the zygote with no
contribution from the father.The strong maternal inheritance could also be mediated by in utero
maternal epigenetic factors.
In addition to VO2max, several other maximal and submaximal endurance-​related traits have
been shown to have significant genetic heritability levels as estimated from twin and family
studies (Table 14.1) (5, 18, 24, 26, 33). In the study of 66 DZ twins and 106 MZ twins of both
sexes, the range of heritability estimates for relative total work output during 90 minutes of
maximal exercise, maximal oxygen pulse, and maximal ventilation were 0.66–​0.70, 0.49–​0.55,
and 0.51–​0.95, respectively (12). A study of 22 pairs of DZ and 31 pairs of MZ twins found
that heritability estimates of submaximal endurance traits adjusted for body weight were high
for low-​power outputs and became nonsignificant at higher power outputs (15). The estimated
heritability of VO2 (ml/​kg/​min) at submaximal absolute intensities was 0.90 at 50 W, 0.78 at
75 W, 0.46 at 100 W, and 0.0 at 125 and 150 W, respectively (15). The estimated heritability of
VO2 at submaximal relative intensities was nonsignificant for VO2 at a heart rate of 150 and 0.63
for VO2 at a respiratory exchange ratio of 0.95 (20). Similarly, in HERITAGE the heritability

180
181

Heritability of human endurance traits

Study Heritability (95% CI) % Weight

Perusse et al. 1987 (PO at HR 150) 0.00 (–0.06, 0.06) 6.37
Bouchard et al. 1989 (VO2 at 50 W) 0.90 (0.67, 1.13) 5.63
Bouchard et al. 1989 (VO2 at 75 W) 0.78 (0.55, 1.01) 5.63
Bouchard et al. 1989 (VO2 at 100 W) 0.46 (0.23, 0.69) 5.63
Bouchard et al. 1989 (VO2 at 125 W) 0.00 (–0.23, 0.23) 5.63
Bouchard et al. 1989 (VO2 at 150 W) 0.00 (–0.23, 0.23) 5.63
Fagard et al. 1991 (VO2 at HR 150) 0.16 (–0.09, 0.41) 5.56
Fagard et al. 1991 (VO2 at RER 0.95) 0.63 (0.38, 0.88) 5.56
Fagard et al. 1991 (PO at HR 150) 0.54 (0.29, 0.79) 5.56
Fagard et al. 1991 (PO at RER 0.95) 0.57 (0.32, 0.82) 5.56
Gaskill et al. 2001 (VO2 at VT in Caucasian) 0.58 (0.45, 0.71) 6.16
0.54 (0.36, 0.72) 5.91
Gaskill et al. 2001 (VO2 at VT in African)
Perusse et al. 2001 (VO2 at 50 W) 0.70 (0.59, 0.81) 6.23
Perusse et al. 2001 (VO2 at 60% VO2max) 0.52 (0.41, 0.63) 6.23
Perusse et al. 2001 (VO2 at 80% VO2max) 0.48 (0.37, 0.59) 6.22
Perusse et al. 2001 (PO at 60% VO2max) 0.68 (0.57, 0.79) 6.23
Perusse et al. 2001 (PO at 80% VO2max) 0.74 (0.63, 0.85) 6.23
Overall (I2=95.0%, P<0.001) 0.49 (0.33, 0.65) 100.00
–0.5 0 0.5 1

Figure 14.3 Meta-​analysis of heritability estimates of submaximal endurance phenotypes in the

untrained state.

estimates were 0.52 and 0.48 for VO2 at 60% and 80% of VO2max, respectively (27), while
heritability estimates for VO2 at ventilatory threshold were 0.58 for white and 0.54 for black
individuals (21). The heritability estimates for submaximal endurance traits are summarized
in the systematic review and meta-​analysis by Miyamoto-​Mikami et al. (24), which found a
weighted mean heritability estimate of 0.49 (95% CI 0.33–​0.65) for submaximal endurance
traits (Figure 14.3).
Lastly, one twin study examined longitudinal changes in genetic influences of intrinsic
endurance-​related traits. Specifically, the heritability of maximal walking speed over 10 m and a
6-​min walking endurance test were estimated in 63 MZ and 67 DZ pairs of older (63–​75 years)
twin sisters at baseline and 3 years later (25). The authors found that the heritability estimate of
walking speed was similar at baseline (56%) and follow-​up (60%), while heritability of walking
endurance increased from baseline (40%) to follow-​up (60%). Additional analyses showed the
genetic influences at baseline and follow-​up were highly correlated for both traits (rg=0.71–​
0.72), suggesting expression of the same genetic factors at both time points. However, the
results also suggest a small proportion of newly expressed genetic factors at follow-​up for both
traits (25). A study of older twins followed over 6 years found an increase in genetic variance of
self-​reported physical functioning (17). The authors postulate that these results conform to the
genetic-​evolutionary theories of aging that suggest genetic variability in the underlying traits
for physical functioning tend to vary more at older ages due to changes in gene expression in
different body systems or the weakening influence of beneficial genes that were strongly active
in early life (17, 27).

181
182

J.L. Barber and M.A. Sarzynski

Human variation in endurance training responses

Both animal and human studies provide strong evidence that there is a significant genetic
component to endurance-​related traits. Importantly, the genetic component is thought to
affect both the intrinsic level of endurance in the untrained state and responsiveness to endur-
ance training. In general, the heritability estimates tend to be lower for endurance response
phenotypes as compared to intrinsic phenotypes.This section will briefly summarize important
studies highlighting variation in endurance trait response to exercise training and the role of
genetic factors in these responses.
The beneficial effects of endurance exercise training on several health outcomes and phys-
ical performance have been well documented (28). However, numerous studies have also shown
large heterogeneity in the adaptation to exercise training (13). In the early 1980s, Dr. Claude
Bouchard led a series of carefully controlled and standardized exercise training studies that
showed the individual differences in training-​induced changes in several endurance-​related
phenotypes were large, with the range between low and high responders reaching several-​
fold (6, 7, 10, 23, 34). The most extensive data on individual differences in trainability come
from the HERITAGE Family Study, in which healthy but sedentary subjects followed a highly
standardized, fully monitored and enforced, laboratory-​ based endurance-​training program
consisting of 60 exercise sessions over 20 weeks (204 families, 742 subjects) (11). In the entire
HERITAGE cohort, the average increase in VO2max was 384 ml/​min with a standard deviation
of 202 ml/​min, while individual training responses varied from no change to increases of more
than 1000 ml/​min (8, 13). This heterogeneity in the response of VO2max has been shown in
several endurance training studies (31), as shown in Figure 14.4.
Interestingly, the marked interindividual differences inVO2max responsiveness in HERITAGE
was not accounted for by baseline VO2max level, age, sex, or race. The HERITAGE Family
Study is uniquely qualified to examine the contributions of these factors on responsiveness
to standardized exercise training given its study design. Specifically, the HERITAGE cohort
consisted of black and white families with both parents and three or more adult offspring that
completed the same standardized and supervised endurance training program (11).Thus, analyses

2200

1800
Delta VO2max (ml/min)

1400

1000

600

200

-200

-600

-1000
1 101 201 301 401 501 601 701 801 901 1001 1101 1201 1301 1401 1501 1601 1701
Individuals

Figure 14.4 Heterogeneity of VO2max response to endurance exercise training across 1724 participants
from 8 studies and 14 different interventions.

182
183

Heritability of human endurance traits

in HERITAGE could examine the independent or combined effects of age (parents vs. offspring
or continuous variable), sex, and race (black vs. white) on exercise responsiveness. For example,
the same heterogeneity in response levels can be found in those who began the program with a
low VO2max and in those who were initially above the VO2max median (13). A similar pattern
of interindividual variation in training responses was observed for other endurance-​related
phenotypes in HERITAGE, including submaximal exercise stroke volume, cardiac output, heart
rate, and blood pressure (15, 35, 36). These data demonstrate that interindividual variability in
response to exercise training is common across diverse populations and exercise programs.

Heritability of the response of endurance traits to exercise training

Results from twin and family studies suggest that the marked interindividual differences in the
trainability of phenotypes governing endurance performance is highly familial and genetically
determined. Key twin and family studies examining these issues are summarized below.

Twin studies
A study of ten pairs of MZ twins subjected to 20 weeks of endurance training found a wide
range of individual training gains in absolute VO2max (0–​41%), but the distribution of responses
was not random among the twins. The intraclass correlation of VO2max response (l/​min) was
0.77, indicating that members of the same twin pair responded similarly to training, whereas
there was almost eight times (F ratio=7.8) more variance between pairs of twins than within
pairs for absolute VO2max response (29). These results were replicated in two other studies of
endurance training in MZ twins (16, 22), which showed intraclass correlations in response of
VO2max (l/​min) of 0.44 and 0.65. Pooled results of the trainability of VO2max in MZ twins
from the three above-​mentioned studies show there is about nine times more variance between
genotypes than within genotypes in the response of VO2max to standardized endurance training
protocols (32). Similarly, pooled analysis of four twin studies (N=41 pairs) found the within-​
MZ pair correlation for relative VO2max response to endurance training was 0.38 (95% CI
0.04–​0.64) (37). A study of nine pairs of MZ twins aged 11–​14 years that completed 6 months
of endurance training used the between-​pair variance to represent the heritability and found
the heritability of relative VO2max response to be 46% (19).
A significant intrapair resemblance of 0.60 for change in VO2max was also observed in a study
of seven male MZ twin pairs subjected to a negative energy balance protocol over a period of
3 months (14). The daily energy deficit was induced entirely by exercise training performed
twice daily on a cycle ergometer, 9 out of 10 days, over a period of 93 days while subjects were
kept on a constant daily energy and nutrient intake (14). In addition to VO2max, there were
significant within-​pair resemblances in training response for submaximal exercise measures of
VO2 (r=0.87 and 0.76 at 50 and 150 W, respectively), respiratory quotient (r=0.86 and 0.87 at
100 and 150 W, respectively), and heart rate (r=0.70 and 0.63 at 50 and 150 W, respectively)
(14). Lastly, two studies of MZ twins have also shown significant within-​pair resemblance in
training response for total power output during a 90-​min maximal exercise test, with the ratio
of between pairs to within pairs variances ranging from 5.5 to 11, and the intraclass coefficient
for twin resemblance in response ranging from 0.69 to 0.83 (16, 22).
In summary, these studies showed that twin pairs are significantly more alike than unrelated
individuals in their cardiorespiratory responses to standardized endurance training programs.
Thus, a substantial genetic component exists in exercise training-​induced changes in endurance-​
related phenotypes.

183
184

J.L. Barber and M.A. Sarzynski

Family studies
To date, the only family study to examine the heritability of endurance-​related phenotypes to
exercise training is the HERITAGE Family Study. In HERITAGE, the increase in absolute
VO2max in 481 individuals from 99 two-​generation families of Caucasian descent showed 2.5
times more variance between families than within families (8). Furthermore, under the most
parsimonious model, the maximal heritability of absolute VO2max response adjusted for age and
sex was 47%. The authors found that adjusting for baseline VO2max did not change the herit-
ability estimate, indicating that the familial and genetic factors underlying intrinsic VO2max and
VO2max response to training appear to be different. These findings are generally in agreement
with those from pharmacological interventions. Combined analyses of genome-​wide associ-
ation study results from three trials of statin efficacy (N=3928) showed there was no overlap
between the genetic variants associated with variation in the lipid levels of the cohort at base-
line compared with those associated with the change in lipid levels (4). Thus, available evidence
both from exercise and pharmacological interventions appear to show that the genetic factors
influencing response to intervention differ from those affecting the intrinsic state of complex
phenotypes.
In addition to VO2max, significant familial aggregation and the heritability of training-​
induced changes in several other endurance-​related phenotypes, including submaximal aerobic
performance (27), resting and submaximal stroke volume, cardiac output, and heart rate were
found in HERITAGE (1–​3, 30). The response to exercise training of all measured submaximal
endurance performance phenotypes showed significant familial resemblance in white families
in HERITAGE, with maximal heritability values of 57% for ΔVO2 at 50 W, 23% and 44% for
ΔVO2 at 60% and 80% of VO2max, and 33% and 45% for ΔPower output at 60% and 80%
of VO2max (27). Analyses in HERITAGE also found a moderate familial component in the
response of VO2 at the ventilatory threshold to endurance exercise training in white subjects
(heritability of 22%) and a larger component in black subjects (heritability of 51%) (21).
Submaximal exercise stroke volume and cardiac output were also characterized by a signifi-
cant familial aggregation in response to endurance training in HERITAGE, as the between-​
family variation in adjusted stroke volume and cardiac output training responses at 50 W were
1.5 to 2.2 times greater than the within-​family variation (2). In white families of HERITAGE,
maximal heritability estimates were 29% and 38% for stroke volume and cardiac output training
responses at 50 W, and 24% and 30% for the training-​induced changes in stroke volume and
cardiac output at 60% of VO2max (12). In the entire HERITAGE cohort, there was 1.8 times
more variance between families than within families for the response of submaximal heart rate
(13), with maximal heritability estimates reaching 34% and 29% for heart rate training responses
at 50 W and 60% of VO2max, respectively, in white families (1).

Conclusion
Heterogeneity in the levels of complex traits is a normal biological phenomenon, which is
not limited to the endurance-​related phenotypes discussed in this chapter. It is ubiquitous
and commonly observed in both the intrinsic (untrained) state and in response to training.
Furthermore, interindividual differences in endurance traits are not randomly distributed, as
they are characterized by significant familial aggregation. Significant heritability estimates have
been reported for both intrinsic levels and training-​induced changes in several endurance-​
related phenotypes, confirming that familial factors account for some of the interindividual
variation in these phenotypes. Both twin and family studies show that individuals with the

184
185

Heritability of human endurance traits

same genotype are more likely to have similar baseline levels of endurance traits, as well as
respond more similarly to training than those with different genotypes. However, the evidence
also suggests that there are separate genetic components affecting endurance-​related phenotypes
at baseline in the untrained state and in response to exercise training. Better understanding the
genetic factors underlying interindividual differences in responsiveness to acute and regular
exercise will allow for a better understanding of the biology of exercise response, as well as the
potential to utilize genetic information in personalized exercise medicine.

References
1. An P, Perusse L, Rankinen T, Borecki IB, Gagnon J, Leon AS, Skinner JS, Wilmore JH,
Bouchard C, and Rao DC. Familial aggregation of exercise heart rate and blood pressure in
response to 20 weeks of endurance training: The HERITAGE family study. Int J Sports Med
24: 57–​62, 2003.
2. An P, Rice T, Gagnon J, Leon AS, Skinner JS, Bouchard C, Rao DC, and Wilmore JH. Familial
aggregation of stroke volume and cardiac output during submaximal exercise: The HERITAGE
Family Study. Int J Sports Med 21: 566–​572, 2000.
3. An P, Rice T, Perusse L, Borecki IB, Gagnon J, Leon AS, Skinner JS, Wilmore JH, Bouchard
C, and Rao DC. Complex segregation analysis of blood pressure and heart rate measured before and
after a 20-​week endurance exercise training program: The HERITAGE Family Study. Am J Hypertens
13: 488–​497, 2000.
4. Barber MJ, Mangravite LM, Hyde CL, Chasman DI, Smith JD, McCarty CA, Li X, Wilke
RA, Rieder MJ, Williams PT, Ridker PM, Chatterjee A, Rotter JI, Nickerson DA, Stephens
M, and Krauss RM. Genome-​wide association of lipid-​lowering response to statins in combined
study populations. PLoS One 5: e9763, 2010.
5. Beunen GP, Peeters MW, and Malina RM. Twin studies in sports performance. In: Genetic and
Molecular Aspects of Sport Performance, edited by Bouchard C and Hoffman EP.West Sussex, UK: Wiley-​
Blackwell, 2011, p. 101–​109.
6. Bouchard C. Human adaptability may have a genetic basis. In: Risk Reduction and Health
Promotion: Proceedings of the 18th Annual Meeting of the Society of Prospective Medicine, edited by Landry
F. Ottawa: Canadian Public Health Association, 1983, p. 463–​476.
7. Bouchard C. Individual differences in the response to regular exercise. Int J Obes Relat Metab Disord
19 Suppl 4: S5–​8, 1995.
8. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, Perusse L, Leon AS, and Rao
DC. Familial aggregation of VO(2max) response to exercise training: Results from the HERITAGE
Family Study. J Appl Physiol 87: 1003–​1008, 1999.
9. Bouchard C, Daw EW, Rice T, Perusse L, Gagnon J, Province MA, Leon AS, Rao DC,
Skinner JS, and Wilmore JH. Familial resemblance for VO2max in the sedentary state: The
HERITAGE family study. Med Sci Sports Exerc 30: 252–​258, 1998.
10. Bouchard C, Dionne FT, Simoneau JA, and Boulay MR. Genetics of aerobic and anaerobic
performances. Exerc Sport Sci Rev 20: 27–​58, 1992.
11. Bouchard C, Leon AS, Rao DC, Skinner JS, Wilmore JH, and Gagnon J. The HERITAGE
family study. Aims, design, and measurement protocol. Med Sci Sports Exerc 27: 721–​729, 1995.
12. Bouchard C, Lesage R, Lortie G, Simoneau JA, Hamel P, Boulay MR, Perusse L, Theriault
G, and Leblanc C. Aerobic performance in brothers, dizygotic and monozygotic twins. Med Sci
Sports Exerc 18: 639–​646, 1986.
13. Bouchard C, and Rankinen T. Individual differences in response to regular physical activity. Med
Sci Sports Exerc 33: S446–​451, 2001.
14. Bouchard C, Tremblay A, Despres JP, Theriault G, Nadeau A, Lupien PJ, Moorjani S,
Prudhomme D, and Fournier G. The response to exercise with constant energy intake in identical
twins. Obes Res 2: 400–​410, 1994.
15. Bouchard C, Tremblay A, Nadeau A, Despres JP, Theriault G, Boulay MR, Lortie G,
Leblanc C, and Fournier G. Genetic effect in resting and exercise metabolic rates. Metabolism
38: 364–​370, 1989.

185
186

J.L. Barber and M.A. Sarzynski

16. Boulay MR LG, Simoneau JA, Bouchard C. Sensitivity of maximal aerobic power and capacity
to anaerobic training is partly genotype dependent. In: Sport and Human Genetics, edited by Malina R,
and Bouchard C. Champaign, IL: Human Kinetics, 1986, p. 173–​181.
17. Christensen K, Frederiksen H, Vaupel JW, and McGue M. Age trajectories of genetic variance
in physical functioning: A longitudinal study of Danish twins aged 70 years and older. Behav Genet
33: 125–​136, 2003.
18. Costa AM, Breitenfeld L, Silva AJ, Pereira A, Izquierdo M, and Marques MC. Genetic inher-
itance effects on endurance and muscle strength: An update. Sports Med 42: 449–​458, 2012.
19. Danis A, Kyriazis Y, and Klissouras V. The effect of training in male prepubertal and pubertal
monozygotic twins. Eur J Appl Physiol 89: 309–​318, 2003.
20. Fagard R, Bielen E, and Amery A. Heritability of aerobic power and anaerobic energy generation
during exercise. J Appl Physiol 70: 357–​362, 1991.
21. Gaskill SE, Rice T, Bouchard C, Gagnon J, Rao DC, Skinner JS, Wilmore JH, and Leon AS.
Familial resemblance in ventilatory threshold: The HERITAGE Family Study. Med Sci Sports Exerc
33: 1832–​1840, 2001.
22. Hamel P, Simoneau JA, Lortie G, Boulay MR, and Bouchard C. Heredity and muscle adapta-
tion to endurance training. Med Sci Sports Exerc 18: 690–​696, 1986.
23. Lortie G, Simoneau JA, Hamel P, Boulay MR, Landry F, and Bouchard C. Responses of max-
imal aerobic power and capacity to aerobic training. Int J Sports Med 5: 232–​236, 1984.
24. Miyamoto-​Mikami E, Zempo H, Fuku N, Kikuchi N, Miyachi M, and Murakami H.
Heritability estimates of endurance-​related phenotypes: A systematic review and meta-​analysis. Scand
J Med Sci Sports 28: 834–​845, 2018.
25. Ortega-​Alonso A, Sipila S, Kujala UM, Kaprio J, and Rantanen T. Longitudinal changes
in genetic and environmental influences on older women’s walking ability. Scand J Med Sci Sports
19: 669–​677, 2009.
26. Perusse L. Role of genetic factors in sport performance: Evidence from family studies. In: Genetic
and Molecular Aspects of Sport Performance, edited by Bouchard C and Hoffman EP. West Sussex,
UK: Wiley-​Blackwell, 2011, p. 90–​100.
27. Perusse L, Gagnon J, Province MA, Rao DC, Wilmore JH, Leon AS, Bouchard C, and
Skinner JS. Familial aggregation of submaximal aerobic performance in the HERITAGE Family
study. Med Sci Sports Exerc 33: 597–​604, 2001.
28. Physical Activity Guidelines Advisory Committee. Physical Activity Guidelines Advisory Committee
Report, 2008. Washington, DC: Department of Health and Human Services, 2008.
29. Prud’homme D, Bouchard C, Leblanc C, Landry F, and Fontaine E. Sensitivity of maximal
aerobic power to training is genotype-​dependent. Med Sci Sports Exerc 16: 489–​493, 1984.
30. Rice T, An P, Gagnon J, Leon AS, Skinner JS, Wilmore JH, Bouchard C, and Rao DC.
Heritability of HR and BP response to exercise training in the HERITAGE Family Study. Med Sci
Sports Exerc 34: 972–​979, 2002.
31. Ross LM, Church TS, Blair SN, Durstine JL, Hagberg JM, Martin CK, Rankinen T, Ross R,
Bouchard C, and Sarzynski MA. Prevalence of Vo2max low response across nine aerobic exercise
interventions. Med Sci Sports Exerc 49: 838, 2017.
32. Sarzynski MA, Rankinen T, and Bouchard C. Twin and family studies of training responses.
In: Genetic and Molecular Aspects of Sport Performance, edited by Bouchard C and Hoffman EP. West
Sussex, UK: Wiley-​Blackwell, 2011, p. 110–​120.
33. Schutte NM, Nederend I, Hudziak JJ, Bartels M, and de Geus EJ. Twin-​sibling study and meta-​
analysis on the heritability of maximal oxygen consumption. Physiol Genomics 48: 210–​219, 2016.
34. Simoneau JA, Lortie G, Boulay MR, Marcotte M, Thibault MC, and Bouchard C. Inheritance
of human skeletal muscle and anaerobic capacity adaptation to high-​intensity intermittent training.
Int J Sports Med 7: 167–​171, 1986.
35. Wilmore JH, Stanforth PR, Gagnon J, Rice T, Mandel S, Leon AS, Rao DC, Skinner JS, and
Bouchard C. Cardiac output and stroke volume changes with endurance training: The HERITAGE
Family Study. Med Sci Sports Exerc 33: 99–​106, 2001.
36. Wilmore JH, Stanforth PR, Gagnon J, Rice T, Mandel S, Leon AS, Rao DC, Skinner JS, and
Bouchard C. Heart rate and blood pressure changes with endurance training: The HERITAGE
Family Study. Med Sci Sports Exerc 33: 107–​116, 2001.
37. Zadro JR, Shirley D, Andrade TB, Scurrah KJ, Bauman A, and Ferreira PH. The beneficial
effects of physical activity: Is it down to your genes? A systematic review and meta-​analysis of twin
and family studies. Sports Med Open 3: 4, 2017.

186
187

15
GENETIC CONTRIBUTIONS TO
CARDIORESPIRATORY FITNESS
Reuben Howden, Benjamin D.H. Gordon, and Ebony C. Gaillard

Introduction
In the animal kingdom, humans are impressive runners.While human sprint speed does not com-
pare favorably with other animals (e.g., ~10 m/​s for ~15 secs vs. 15–​20 m/​s for several minutes
in horses), human capacity for endurance running is seldom rivaled (1). Cardiopulmonary
function is a key feature of endurance running capacity and therefore the cardiopulmonary
system was an important component in the genetic development (evolution) of Homo.
Cardiorespiratory fitness (CRF) is broadly defined as the capacity of circulatory and respira-
tory systems to deliver energy substrates and oxygen to muscles contributing to prolonged
exercise. The principal components of CRF have been described beginning with the work of
A.V. Hill and colleagues early in the 20th century (2). Low CRF is well established as an indi-
cator of all-​cause mortality risk (3, 4). Moreover, the health benefits of engaging in habitual
exercise training that likely leads to improved CRF are well described (e.g., 5–​8). Since CRF is
determined mainly by cardiovascular and pulmonary capacity during exercise, this chapter will
focus on these traits and their genetic underpinnings.
The standard measure for CRF is maximum capacity for oxygen consumption (VO2max),
which is a robust and reproducible measure with a reliability coefficient of at least 0.95 (9, 10).
However,VO2 max is defined by the Fick equation Q Å~A– VO2 difference, where Q is cardiac
output (l/min) and A–VO2 difference is the difference in oxygen content between arterial and
venous blood, an index for cellular oxygen uptake. The Fick equation does not consider pul-
monary function and is therefore a contradiction in terms as a measure of CRF (oxygen supply
vs. oxygen uptake, i.e., exercising muscle does not uptake all oxygen supplied). Predicting cardio-
pulmonary fitness via VO2max assessment has become an important determinant of not only the
degree of adaptation to exercise training, but also mortality risk (11).The latter raises the question
of whether a high CRF (via genetics or exercise training) is more or less important than simply
engaging in physical activity regardless of its influence on CRF: genetics versus environment.
It has been argued that reaching a high level of performance (e.g., CRF) is solely dependent
upon accumulating hours of so-​called deliberate practice and is not limited by individual geno-
type (12, 13). Ericsson’s model proposed that all individuals possess the necessary genetics to
become expert performers.This suggests that any individual with sufficient determination could
become competitive at the elite level, including sports performance, as long as they accumulate

187
188

R. Howden et al.

enough deliberate practice hours (the common figure suggested is 10,000 hours). However, it
has been reported that athletes in sports where CRF is important can achieve elite status with
considerably less “deliberate practice” than 10,000 hours (14, 15), suggesting a capacity for
accelerated progression or responsiveness to training; ergo, a genetic component. In any case,
the premise of Ericsson’s “deliberate practice” model has been criticized in several reports, not
least of all in a recent meta-​analysis on deliberate practice and human performance in a number
of categories, which found the deliberate practice only explained 18% of sports performance
variability (16). This does not discount the critical importance of deliberate practice (training)
to reach maximum possible individual performance. Indeed, failure of individuals to reach their
maximum potential at a given activity has long been associated with a lack of commitment to
practice or training (17). However, it does appear that a considerable proportion of human CRF
levels are not explained by deliberate practice. Contributing factors to this large proportion of
variability may include opportunity, environment, socioeconomic status, ethnic background,
and sex, among others.
Interestingly, there is considerable variation between sedentary individuals in CRF, suggesting
innate capacity (18). Evidence from twin human and inbred mouse (19) studies, as well as
interindividual differences and sexual dimorphism (20), suggests genetic contributions to CRF
and responses to exercise training (18, 21). These contributions are described in more detail in
Chapters 13 and 14 of this book. To understand more about specific genes that may contribute
to CRF, it is prudent to consider factors that determine or limit CRF and the genetics modi-
fying those factors.

Cardiac function
Cardiac function is a major component of CRF and therefore is a key feature of the Fick
equation describing VO2. The job of the heart is to generate blood pressure and flow, moving
deoxygenated blood through the pulmonary vasculature for gas exchange and then oxygenated,
nutrient-​rich blood through to the exchange vessels. Baseline cardiac function can provide
insight into individual fitness levels. For example, lower resting heart rate (HR) suggests a larger
stroke volume (SV) for a given cardiac output (Q). Variation in human HR carries approxi-
mately 50% heritability (22). However, in inbred mice, significant variation in resting HR, com-
prising a 296.4 bpm difference between 129/​Svlm and Castaneous strains has been reported
(23). Interestingly, there does not appear to be a relationship between untrained resting HR (23)
and voluntary wheel running in inbred mice (24), suggesting that a low sedentary resting HR,
usually associated with fitness, is not related to motivation for exercise. Strong evidence exists
for resting HR being a heritable trait, with estimates ranging from approximately 51% to 65%
(22, 25–​28).
A higher maximum Q, within certain limits, results in higher CRF as determined by VO2max
(Fick equation). In the HERITAGE Family Study, heritability of SV and Q was approximately
40%, suggesting a moderate genetic influence on baseline cardiac function. Moreover, assessed at
different submaximal exercise intensities (% VO2max) heritability estimates for SV and Q were
similar to baseline (41–​46%) (29).
Variation in HR responses to exercise, before and after training, also suggest a genetic com-
ponent to cardiac adaptation to exercise training. In HERITAGE, HRs at 50 W workload
(HR50) were assessed before and after a 20-​week endurance exercise training program (30).
Interestingly, the mean training-​induced reduction in HR50 was almost matched by the standard
deviation. This variation in training responsiveness produced a heritability estimate of 34%,
which may be important not only for exercise performance improvements, but also the degree

188
189

Cardiorespiratory fitness

to which mortality risk can be lowered. Previous reports have suggested that improvements in
CRF are important in lowering mortality risk (11). Therefore, presumably the slope of the cor-
relation between improved CRF and mortality risk may be influenced by genetics. However,
this remains a controversial idea particularly in twin or rodent studies when habitual physical
activity began in adulthood (31–​33).

Cardiac function, QTL, and candidate genes

Limited work has been done to identify chromosomal regions (quantitative trait loci (QTL))
that associate with cardiac function. Inbred (30 strains) and recombinant inbred (29 AXB/​BXA
strains) mouse strain distributions for cardiac function were reported as continuous in quiescent
mice (23). This suggests that cardiac phenotypes related to CRF such as resting HR are com-
plex, meaning many genes are likely involved, which is not surprising. Others reported similar
inbred strain distributions, although recorded resting HRs were generally higher, possibly due
to using older mice and electrocardiographs (ECGs) were recorded in mice that were not qui-
escent (34). The conditions under which ECGs for cardiac function calculations are recorded is
critical to consider since, for example, HR is highly sensitive to even minor stimuli (e.g., sound,
smell, and vibration) detected by the research participant or animal which influences results.
This is an important consideration, because although the genetic influences on inbred mouse
HR responses to exercise have not been reported, there is evidence that genetic factors influ-
encing cardiac function may differ when stress levels change (e.g., rest vs. exercise vs. training;
see below).
QTL identified as associated with resting HR include mouse chromosomes 6 (23), 2, and
15 (35), rat chromosomes 2 (36), 3 (37) and 8 (38), and human chromosome 4 (39). However,
these QTL are quite broad and therefore is it challenging to assess concordance between species.
In any case, as stated previously, subtle differences in the precise conditions under which HRs
were recorded and their influence on “resting” HR also make such a comparison difficult. Some
of the previously identified HR QTL overlap with regions also reported to be associated with
other cardiovascular phenotypes (e.g., blood pressure) (36), which should not be surprising, but
does increase confidence in the importance of these chromosomal regions in cardiovascular
regulation.
Proposed candidate genes within published QTL for resting HR include, but are not limited
to, corticotropin-​releasing factor receptor 2 and neuropeptide Y on mouse chromosome 6
(23), which have been associated with HR regulation elsewhere (40–​42). Cholinergic receptor,
nicotinic, polypeptide-​α (Chrna1) on mouse chromosome 2 arising from a comparison between
BALB/​J and CBA/​CaJ mice was identified as a candidate gene for resting HR (35). The influ-
ence of the parasympathetic nervous system on HR is well known and therefore, it is plausible
that Chrna1 could be important in HR regulation.
Studies reporting HR QTL are at least 10 years old and little has been done in terms
of successful identification of genes within these QTL that have a functional influence on
cardiovascular function. This is a problem well recognized especially for phenotypes that are
influenced by many genes (complex traits) (43), but this may improve as the research tools
needed to unravel the complexity of physiological phenomena continues to develop (44, 45).
Candidate genes associated with a complex and easily influenced phenotype such as HR are
likely numerous. Cardiac function is influenced by multiple factors, including but not limited
to peripheral neural inputs, central nervous system function, vascular function, and pulmonary
function.The complexity of interactions within the cardiorespiratory system (46) makes genetic
regulation very difficult to dissect. However, there has been progress in identifying candidates

189
190

R. Howden et al.

for which there is evidence of an association between cardiac function and exercise (e.g., 47–​
51), which could be extrapolated to CRF. It is, however, beyond the scope of this chapter to
discuss all reported genes associated with cardiac function or indeed CRF and therefore we
discuss prominent examples only.
Aging is associated with reductions in cardiac function, but exercise training to improve
CRF has been shown to attenuate this decline (52–​54). The genetic influences on cardiac
function adaptation to exercise training in the aged appear to be associated with an attenuation
of the known switch in myosin heavy chain isoform from α to β and therefore lower sarco-
plasmic reticulum Ca2+ ATPase activity (55, 56). This effect of exercise training was reported in
aged rats and was controlled via an attenuated age-​related decline in myocyte thyroid receptor
(TR) protein expression (47), which has been associated with rat cardiac hypertrophy (57).
A similar effect on TR protein expression was also reported after exercise training in post-​
myocardial infarction (58). Taken together, these studies suggest that exercise training induced
improvements in cardiac function and therefore CRF may in part operate through changes in
TR expression during aging or pathology.
An increase in left ventricular mass is a well-​established adaptation to exercise training
that improves CRF. Genes that have been reported as associated with left ventricular mass
include, but are not limited to, angiotensin-​converting enzyme (ACE; (59, 60, 61), peroxisome
proliferator-​activated receptor alpha (PPARα; 62), G-​protein β3 subunit (GNB3; 63) and aldos-
terone synthase (CYP11B2; 64).
A polymorphism in the ACE gene (deletion/​insertion) has received significant attention
in association with ventricular hypertrophy with (59, 60) and without (61) hypertension.
Moreover, associations between ACE polymorphisms and athletic ability have already been
reviewed (65). It is well known that ACE plays a key role in vascular function by producing the
vasoconstrictor angiotensin II and degrades vasodilator kinins (66). It is also well established that
increases in sympathetic nerve activity during exercise lead to increased production of angio-
tensin II, which acts as a nonadrenergic vasoconstrictor contributing to blood flow redistribu-
tion supporting active muscle function. This role of angiotensin II in blood flow redistribution
is a critical component to maximizing delivery of oxygen and nutrient-​rich blood to active
muscle during exercise and therefore CRF. Moreover, ACE has been associated with exercising
muscle function, especially in sprinters (67), but it is beyond the scope of this chapter to discuss
exercising muscle function in detail.
Calcineurin is a well-​ known component of calcium signaling, especially in myocar-
dial hypertrophy (68). Indeed, calcineurin deficiency in mice was associated with lower
cardiac mass (69). Moreover, calcineurin polymorphisms were reported to influence car-
diac adaptations to cycle-​ergometry training, assessed by postexercise HR responses. Other
polymorphisms in the same gene are associated with resting HR, which is of well-​established
clinical significance (70).
Interestingly, there is evidence that genes influencing cardiac function at rest in untrained
subjects are not necessarily important in cardiac adaptations induced by exercise training. For
example, heme oxygenase 1 (HMOX1), which is involved in heme catabolism under conditions
of oxidative stress and hypoxia among other stressors, has been shown to influence SV and
Q at rest and in response to exercise. However, HMOX1 was not associated with SV and Q
adaptations induced by exercise training (71).
Differential genetic regulators of the same phenotypes have been shown previously. A QTL
associated with baseline HR (chromosome 6) (23) was different from a QTL associated with
HR responses to oxidative stress (chromosomes 3, 5, and 9) (72). This serves to illustrate the
complexity of genetic influences on cardiac function, a major component of CRF.

190
191

Cardiorespiratory fitness

Respiratory function
The respiratory system functions to move oxygen (O2) from ambient air into blood and for
excess carbon dioxide (CO2) to be move from blood to ambient air (gas exchange), a critical
component of CRF. Dalton’s law of partial pressures describes the relative concentrations of
gases in a mixture (e.g., blood gases) where the total pressure of a gas mixture is the sum of the
partial pressures exerted by each gas within that mixture. For example, the partial pressure of
oxygen (PO2) in pulmonary capillary blood is lower than the PO2 in alveolar space and there-
fore oxygen will move (diffuse) into the blood and bind to hemoglobin. Conversely, pulmonary
capillary blood PCO2 is higher than alveolar PCO2, which drives movement of CO2 from
blood into alveolar space.
Factors effecting gas exchange between the alveoli and blood are described by Fick’s law
of diffusion where surface area for gas exchange, the distance gases must travel in order to
move from one space to another, and the partial pressure difference between the two spaces
principally effect diffusion in biological and nonbiological systems. This process is facilitated
by the anatomy and physiology of the respiratory system, for example, the close proximity of
pulmonary capillaries to alveoli reducing the distance gases must travel during diffusion or the
density of the erythrocytes in pulmonary capillaries for oxygen to bind to thus helping to main-
tain gas concentration gradients in the right direction.
In addition to gas exchange governed by diffusion, intercostal and diaphragmatic muscle
function are key components of pulmonary capacity. Moreover, oxygenation of blood is
facilitated by the pulmonary and cardiovascular systems being regulated in concert. For example,
pulmonary inspiration is accompanied by vagal nerve inhibition, increasing Q and pulmonary
blood flow when pulmonary oxygen levels are highest.
Pulmonary function tests are used to measure pulmonary capacity in health and disease, and
include measures of forced vital capacity (FVC), forced expiratory volume in 1 second (FEV1),
and forced expiratory flow at 25–​75% FVC (FEF25–​75). Attenuation of age-​related decline and
improvement in respiratory function as well as exercise performance are all associated with
habitual physical activity (73). Interestingly, the highest level of pulmonary function has been
reported in elite athletes and although increases in diffusion capacity are possible for habitual
exercisers, it appears unlikely that nonelite athletes will reach the pulmonary capacities of elite
athletes. This suggests a genetic influence on pulmonary capacity.

Heritability of and QTL for pulmonary function

Early estimates of pulmonary function heritability between parents and offspring in FEV1 and
FEF25–​75 were between 40% and 50% (74). In Swedish twins, heritability estimates for FVC and
FEV1 were reported as 48% and 67% respectively (75) (mean age ~65 years), which was similar
to an earlier estimate for FEV1 of 77% in American male twins between 42 and 56 years of age
(76). Interestingly, Hubert et al. (1982) also reported that the variation in FVC was explained by
pack-​years of smoking, as well as genetically driven factors related to body size, demonstrating
a gene–​environment interaction. Heritability estimates for pulmonary function are quite con-
sistent compared to cardiac phenotypes, perhaps because pulmonary function is less sensitive to
acute environmental influences.
There are a limited number of studies regarding QTL analyses associated with pulmonary
function. Part of the complication is the number of potential phenotypes of interest that are com-
monly used to assess different aspects of pulmonary function. Using minute ventilation, breathing
frequency, and tidal volume to represent resting pulmonary function in a recombinant inbred

191
192

R. Howden et al.

mouse strain set (29 strains), no significant phenotypic differences or QTL were found (23). In
human studies, a region on chromosome 6 was associated with FEV1 and the FEV1/​FVC ratio
(77). This suggests that if functional tests of pulmonary function are used to assess phenotypes,
it is more likely that heritability estimates and pulmonary function QTL are identified. There is
much work to be done regarding genetic influences of pulmonary function in relation to CRF.

Pulmonary function and genetics

One approach to increase oxygen-​carrying capacity and therefore CRF is living at high altitude
and training at low altitude. The premise of this strategy is to stimulate red blood cell produc-
tion at high altitude while maintaining high-​quality training at low altitude. Extensive study
of the efficacy of “live high, train low” has revealed responders or nonresponders with regard
to increased altitude-​induced erythropoietin (EPO) stimulation (78). For example, altitude
exposure-​induced increases in packed cell volume was matched by an increase in VO2max in
responders only. The differential responsiveness to altitude-​induced improvements in oxygen-​
carrying capacity between individuals suggests a genetic component.
Hypoxia-​inducible factor 1α (HIF-​1α) mediates EPO expression, in response to low ambient
PO2 at high altitude, resulting in an increase in red blood cell count, improved gas exchange,
increasing oxygen-​carrying capacity and CRF. As suggested earlier, results from reports on this
topic have been inconsistent with increases (79, 80), no change (81, 82), or no response in red
blood cell production despite increases in plasma EPO concentrations (83), although the length
of time spent at altitude may also be an important factor in the interindividual differences in
responses to hypoxia exposure (84).
Alveolar angiogenesis in response to exercise training improves pulmonary function and
potentially CRF. The genetic influences on alveolar angiogenesis are quite well characterized
and are mediated by VEGF and FGF (85). Increased vascularization of alveoli induced by exer-
cise training could potentially improve pulmonary gas exchange and therefore CRF but it has
not been demonstrated in humans yet. Conversely, some studies have had difficulty in dem-
onstrating that alveolar capillaries specifically are capable of an angiogenic response during
induced pulmonary hypertension, at least in rats (86, 87).

Hypoxia, pulmonary function, and genetics

The reduction in ambient PO2 with height above sea level and therefore VO2max means that
altitude presents increasing limitations on pulmonary function (88) and CRF.At sea level, arterial
PO2 (PaO2) remains close to resting levels even as exercise intensity increases because ambient
PO2 (101.3 kPa) is high enough for effective gas exchange (89, 90). However, as ambient PO2
declines with increasing altitude, resting PaO2 levels become progressively more difficult to
achieve. Low PaO2 at high altitude has been reported previously (91–​94) making pulmonary
gas exchange increasingly more difficult, particularly during exercise (e.g., mountaineering). In
an attempt to maintain PaO2, pulmonary function responses to reduced ambient PO2 include
increases in pulmonary ventilation (Ve). Pugh et al. (1964) demonstrated that a 2 l/​min VO2
at sea level resulted in a 40 l/​min Ve. However, at 5800 m altitude, a Ve of about 180 l/​min
was required to achieve the same VO2. At 7400 m altitude, the same VO2 was not possible (95).
Exceptionally low VO2max levels are present at Mt. Everest’s summit, estimated at less than 20
ml/​min/​kg compared to a typical sea-​level VO2max of 50–​70 ml/kg/min (91, 95). Previous
to the first “oxygenless” ascent to the summit of Mt. Everest by Messner and Habeler in 1978,

192
193

Cardiorespiratory fitness

studies conducted by Rodolfo Margaria and Joseph Barcroft suggested it was not possible for
humans (96).
At the time of writing, only 208 people have been able to climb Mt. Everest without sup-
plemental oxygen since 1978. This suggests a special ability possessed by a limited number of
successful Mt. Everest climbers (>4800 people), which could have a genetic component, although
this is poorly understood. Certainly, a high CRF is important for successful mountaineering,
which is well known to possess a genetic influence (see earlier in this section) (97).

High-​altitude versus sea-​level dwellers

Several populations live at altitude and therefore hypobaric hypoxia.This provides an interesting
opportunity to understand the resulting chronic adaptations. An estimated 140 million people
reside at altitudes above 2500 m (98) and for thousands of years, several populations have
inhabited altitudes exceeding 3000 m (99).
Nahua, Quechua, and Ayamara in the Andes, Ethiopians in Northern Africa, Sherpa,
Tibetans, and Ladakhi on the Tibetan Plateau, and the Kyrgyz in the Asian Tien-​Shan and
Pamir mountains are examples of populations living in high-​altitude regions (99, 100). Natural
selection appears to have influenced survival and hence reproductive success in relation to
pulmonary function in a hypobaric hypoxic environment and interestingly the adaptations
are divergent between high-​altitude dwellers in different regions. As an example, heritability
estimates for resting ventilation, ventilator response to hypobaric hypoxia, hemoglobin concen-
tration, and PaO2 differed between Tibetan and Andean high-​altitude populations (101, 102).
Specifically, Tibetan populations had a higher resting ventilation (19.7 l/​min; males) compared
to Andean Aymara populations (13.4 l/​min; males) despite lifelong adaptation at a similar alti-
tude (101). However, the variation in resting ventilation measurements taken from the Tibetan
populations was higher than the Andean population for both males and females, which could
have contributed to the differences between the populations (101).
Other differences in pulmonary function between Tibetan and Andean populations include
the hypoxic ventilator response (HVR), which is a standard test for ventilatory reactivity, where
the Tibetan HVR was higher. Interestingly, HVR heritability was estimated at 13% higher in the
Tibetan population (101), although the genetic mechanisms associated with the differences in
pulmonary function between these populations are poorly understood.
Similar genetic variation between Tibetan and sea-​level populations have been reported
(103–​106), which may be due to insufficient selection pressure for different HVRs in Tibet
(107). However, HVRs in Andean populations were significantly different from both Tibetan
and sea-​ level populations. The strategy of higher resting ventilation and HVR could be
considered beneficial under conditions of low ambient PO2 to ensure sufficient gas exchange
and blood oxygenation. However, blood oxygen saturation was lower in the Tibetan, compared
to the Andean populations (101). Although speculative, taken together, these studies do suggest
that natural selection played a role in the success of these populations in their hypoxic environ-
ments, even though the phenotypic outcomes differed. However, one important consideration
that must not be overlooked when discussing the impact of natural selection on gas exchange
adaptations is whether there has been sufficient time for high-​altitude dwellers to be affected by
the process of natural selection or whether each individual born in to these populations have to
adapt to ambient conditions after birth.
The Andean populations in question have resided at high altitude for approximately
11,000 years and Tibetan populations for twice as long. Using calculations for assessing the
required length of time for adaptations to occur by natural selection (108), these time spans

193
194

R. Howden et al.

equate to approximately 550 and 1100 generations respectively, which should provide enough
opportunity for sufficient meiotic events for significant adaptation by natural selection to
improve gas exchange ability at high altitude. This suggests permanent genomic differences
between these high-​altitude dwellers and sea-​level inhabitants.

Candidate genes for pulmonary function responses and

adaptations to hypoxia
Although the genetic regulation of pulmonary responses to hypoxia is not well understood,
a number of potential candidate genes have been identified. For example, HIF transcrip-
tion factors are protective against hypoxic injury (109–​111). During prolonged hypoxia, HIF
increases EPO expression, increasing erythrocyte number (polycythemia) and improving blood
oxygenation (112), although this response may take weeks of exposure to occur (95). HIF is
involved in many other physiological and developmental processes which have been reviewed
elsewhere (113), but this does highlight the genetic influences on hypoxia-​induced responses
and adaptations. Moreover, HIF has been implicated in the regulation of angiogenesis involving
vascular endothelial growth factor and EPO in response to exercise training and therefore
improved CRF (for review, see 114).
The proinflammatory cytokine tumor necrosis factor-​ α (TNF-​ α) production increases
during hypoxia.This effect is especially important in lung disease patients, increasing pulmonary
inflammation and reducing pulmonary function. However, the precise role of TNF-​α, in add-
ition to interleukin-​1 (IL-​1; both α and β) during hypoxia is not clear because it has been
reported to attenuate EPO gene expression (115, 116), as well as increase HIF-​1 expression
(117). Interestingly, TNF-​α is associated with endothelial damage, which could be detrimental
to CRF. However, chronic exercise training has been shown to attenuate TNF-​α expression
at least in heart failure patients (118), further illustrating the importance of habitual physical
activity and improved CRF.

Summary
CRF can be defined by the capacity of the cardiorespiratory system to deliver oxygenated blood
to muscles contributing to prolonged exercise. Physical activity designed to improve CRF is an
important component of reducing health risks. This chapter provides a brief overview of the
kinds of genetic influences on innate CRF, responsiveness to exercise training to improve CRF,
and examples of candidate genes that may be involved.
Importantly, human evolution has involved running and therefore humans are endowed with
numerous and complex genetic mechanisms associated with exercise capacity (i.e., running
capacity improved survival). However, there is considerable between-​individual variation in the
extent to which genetics influence CRF and individual responsiveness to training.This suggests
that some individuals possess a genetic advantage, not just in terms of athletic performance, but
also in their capacity to reduce health risks by engaging in habitual physical activity. On the
other hand, the accepted concept of larger CRF gains with exercise training in those with a low
innate CRF may confer an advantage in terms of reductions in CRF-​related health risks. It is
not clear, however, if innate CRF is the crucial component to reduced health risk or engaging
in habitual physical activity that results in improved CRF.
Both cardiovascular and pulmonary function are the key components of CRF and major
genetic influences have been reported, including estimates of genetic contributions to
interindividual differences, QTL analyses, and candidate gene proposals. In the future, it will be

194
195

Cardiorespiratory fitness

challenging to dissect the complexity of genetic influences on CRF because gene expression
combinations are not consistent under differing environmental conditions (e.g., sea level vs.
altitude (hypoxia). However, it may be possible to improve our understanding of these genetics
mechanisms to design individualized training programs based on genotype, to maximize exer-
cise training responsiveness and therefore the accompanying health benefits.

References
1. Bramble, D.M. and D.E. Lieberman, Endurance running and the evolution of Homo. Nature,
2004. 432(7015): p. 345–​52.
2. Hill, A.V. and H. Lupton, Muscular exercise, lactic acid, and the supply and utilization of oxygen.
QJM, 1923. 16(62): p. 135–​71.
3. Laukkanen, J.A., et al., The predictive value of cardiorespiratory fitness for cardiovascular events
in men with various risk profiles: a prospective population-​based cohort study. Eur Heart J, 2004.
25(16): p. 1428–​37.
4. Kodama, S., et al., Cardiorespiratory fitness as a quantitative predictor of all-​cause mortality and car-
diovascular events in healthy men and women: a meta-​analysis. JAMA, 2009. 301(19): p. 2024–​35.
5. Myers, J., et al., Exercise capacity and mortality among men referred for exercise testing. N Engl J
Med, 2002. 346(11): p. 793–​801.
6. Church, T.S., et al., Effects of different doses of physical activity on cardiorespiratory fitness among
sedentary, overweight or obese postmenopausal women with elevated blood pressure: a randomized
controlled trial. JAMA, 2007. 297(19): p. 2081–​91.
7. Kelley, G.A. and K.S. Kelley, Efficacy of aerobic exercise on coronary heart disease risk factors. Prev
Cardiol, 2008. 11(2): p. 71–​5.
8. Church, T.S., et al., Effects of aerobic and resistance training on hemoglobin A1c levels in patients
with type 2 diabetes: a randomized controlled trial. JAMA, 2010. 304(20): p. 2253–​62.
9. Boulay, M.R., et al., A test of aerobic capacity: description and reliability. Can J Appl Sport Sci, 1984.
9(3): p. 122–​6.
10. Prud’Homme, D., et al., Reliability of assessments of ventilatory thresholds. J Sports Sci, 1984.
2(1): p. 13–​24.
11. Lee, D.C., et al., Mortality trends in the general population: the importance of cardiorespiratory
fitness. J Psychopharmacol, 2010. 24(4 Suppl): p. 27–​35.
12. Ericsson, K.A., R.T. Krampe, and S. Heizmann, Can we create gifted people? Ciba Found Symp,
1993. 178: p. 222–​31.
13. Ericsson, K.A., Deliberate practice and acquisition of expert performance: a general overview. Acad
Emerg Med, 2008. 15(11): p. 988–​94.
14. Gibbons, T., The Path to Excellence: A Comprehensive View of Development of US Olympians Who
Competed from 1984–​1998. Colorado Springs, CO: United States Olympic Committee, 2002.
15. Baker, J., J. Côté, and J. Deakin, Expertise in ultra-​endurance triathletes early sport involvement,
training structure, and the theory of deliberate practice. J Appl Sport Psychol, 2005. 17(1): p. 64–​78.
16. Macnamara, B.N., D.Z. Hambrick, and F.L. Oswald, Deliberate practice and performance in
music, games, sports, education, and professions: a meta-​analysis. Psychol Sci, 2014. 25(8): p. 1608–​18.
17. Watson, J.B., Behaviorism, revised edn. New York: WW Norton & Company, 1930.
18. Bouchard, C. and T. Rankinen, Individual differences in response to regular physical activity. Med
Sci Sports Exerc, 2001. 33(6 Suppl): p. S446–​51.
19. Swallow, J.G., et al., Effects of voluntary activity and genetic selection on aerobic capacity in house
mice (Mus domesticus). J Appl Physiol (1985), 1998. 84(1): p. 69–​76.
20. Pate, R.R. and A. Kriska, Physiological basis of the sex difference in cardiorespiratory endurance.
Sports Med, 1984. 1(2): p. 87–​89.
21. Prud’homme, D., et al., Sensitivity of maximal aerobic power to training is genotype-​dependent.
Med Sci Sports Exerc, 1984. 16(5): p. 489–​93.
22. Zhang, K., et al., Human heart rate: heritability of resting and stress values in twin pairs, and influ-
ence of genetic variation in the adrenergic pathway at a microribonucleic acid (microrna) motif in
the 3’-​UTR of cytochrome b561 [corrected]. J Am Coll Cardiol, 2014. 63(4): p. 358–​68.
23. Howden, R., et al., The genetic contribution to heart rate and heart rate variability in quiescent
mice. Am J Physiol Heart Circ Physiol, 2008. 295(1): p. H59–​68.

195
196

R. Howden et al.

24. Lightfoot, J.T., et al., Genetic influence on daily wheel running activity level. Physiol Genomics, 2004.
19(3): p. 270–​6.
25. Ditto, B., Familial influences on heart rate, blood pressure, and self-​report anxiety responses to
stress: results from 100 twin pairs. Psychophysiology, 1993. 30(6): p. 635–​45.
26. De Geus, E.J., D.I. Boomsma, and H. Snieder, Genetic correlation of exercise with heart rate and
respiratory sinus arrhythmia. Med Sci Sports Exerc, 2003. 35(8): p. 1287–​95.
27. Dalageorgou, C., et al., Heritability of QT interval: how much is explained by genes for resting
heart rate? J Cardiovasc Electrophysiol, 2008. 19(4): p. 386–​91.
28. Wu, T., F.A. Treiber, and H. Snieder, Genetic influence on blood pressure and underlying hemo-
dynamics measured at rest and during stress. Psychosom Med, 2013. 75(4): p. 404–​12.
29. An, P., et al., Familial aggregation of stroke volume and cardiac output during submaximal exer-
cise: the HERITAGE Family Study. Int J Sports Med, 2000. 21(8): p. 566–​72.
30. An, P., et al., Evidence of major genes for exercise heart rate and blood pressure at baseline and in
response to 20 weeks of endurance training: the HERITAGE family study. Int J Sports Med, 2003.
24(7): p. 492–​8.
31. Edington, D.W., A.C. Cosmas, and W.B. McCafferty, Exercise and longevity: evidence for a
threshold age. J Gerontol, 1972. 27(3): p. 341–​3.
32. Goodrick, C.L., et al., Effects of intermittent feeding upon growth and life span in rats. Gerontology,
1982. 28(4): p. 233–​41.
33. Kujala, U.M., J. Kaprio, and M. Koskenvuo, Modifiable risk factors as predictors of all-​cause
mortality: the roles of genetics and childhood environment. Am J Epidemiol, 2002. 156(11): p.
985–​93.
34. Xing, S., et al., Genetic influence on electrocardiogram time intervals and heart rate in aging mice.
Am J Physiol Heart Circ Physiol, 2009. 296(6): p. H1907–​13.
35. Sugiyama, F., et al., QTL associated with blood pressure, heart rate, and heart weight in CBA/​CaJ
and BALB/​cJ mice. Physiol Genomics, 2002. 10(1): p. 5–​12.
36. Alemayehu, A., et al., Reciprocal rat chromosome 2 congenic strains reveal contrasting blood
pressure and heart rate QTL. Physiol Genomics, 2002. 10(3): p. 199–​210.
37. Kreutz, R., et al., Evidence for primary genetic determination of heart rate regulation: chromosomal
mapping of a genetic locus in the rat. Circulation, 1997. 96(4): p. 1078–​81.
38. Silva, G.J., et al., Genetic mapping of a new heart rate QTL on chromosome 8 of spontaneously
hypertensive rats. BMC Med Genet, 2007. 8: p. 17.
39. Martin, L.J., et al., Major quantitative trait locus for resting heart rate maps to a region on chromo-
some 4. Hypertension, 2004. 43(5): p. 1146–​51.
40. Stiedl, O. and M. Meyer, Cardiac dynamics in corticotropin-​releasing factor receptor subtype-​2
deficient mice. Neuropeptides, 2003. 37(1): p. 3–​16.
41. Tovote, P., et al., Central NPY receptor-​mediated alteration of heart rate dynamics in mice during
expression of fear conditioned to an auditory cue. Regul Pept, 2004. 120(1–​3): p. 205–​14.
42. Warner, M.R. and M.N. Levy, Neuropeptide Y as a putative modulator of the vagal effects on heart
rate. Circ Res, 1989. 64(5): p. 882–​9.
43. Kleeberger, S.R. and D.A. Schwartz, From quantitative trait locus to gene: a work in progress. Am
J Respir Crit Care Med, 2005. 171(8): p. 804–​5.
44. Kloosterman, B., et al., From QTL to candidate gene: genetical genomics of simple and complex
traits in potato using a pooling strategy. BMC Genomics, 2010. 11: p. 158.
45. Drinkwater, N.R. and M.N. Gould, The long path from QTL to gene. PLoS Genet, 2012. 8(9): p.
e1002975.
46. De Burgh Daly, M., Interactions between respiration and circulation, in: Cherniack NS,Widdicombe
JG, eds. Handbook of Physiology, Section 3: The Respiratory System, Vol II: Control of Breathing, Part 2.
Bethesda, MD: American Physiological Society; 1986, p. 529–​94.
47. Iemitsu, M., et al., Exercise training improves cardiac function-​related gene levels through thyroid
hormone receptor signaling in aged rats. Am J Physiol Heart Circ Physiol, 2004. 286(5): p. H1696–​705.
48. Jin, H., et al., Effects of exercise training on cardiac function, gene expression, and apoptosis in rats.
Am J Physiol Heart Circ Physiol, 2000. 279(6): p. H2994–​3002.
49. Das, H., et al., Stem cell therapy with overexpressed VEGF and PDGF genes improves cardiac
function in a rat infarct model. PLoS One, 2009. 4(10): p. e7325.
50. Jamshidi, Y., et al., Peroxisome proliferator–​activated receptor α gene regulates left ventricular
growth in response to exercise and hypertension. Circulation, 2002. 105(8): p. 950–​5.

196
197

Cardiorespiratory fitness

51. Boström, P., et al., C/​EBPβ controls exercise-​induced cardiac growth and protects against patho-
logical cardiac remodeling. Cell, 2010. 143(7): p. 1072–​1083.
52. Levy, W.C., et al., Effect of endurance exercise training on heart rate variability at rest in healthy
young and older men. Am J Cardiol, 1998. 82(10): p. 1236–​1241.
53. Li, Y., et al., Age-​related differences in effect of exercise training on cardiac muscle function in rats.
Am J Physiol, 1986. 251(1 Pt 2): p. H12–​18.
54. Wei, J.Y., et al., Chronic exercise training protects aged cardiac muscle against hypoxia. J Clin Invest,
1989. 83(3): p. 778–​84.
55. McNally, E.M., et al., Full-​ length rat alpha and beta cardiac myosin heavy chain sequences.
Comparisons suggest a molecular basis for functional differences. J Mol Biol, 1989. 210(3): p. 665–​71.
56. Tate, C.A., et al., SERCA2a and mitochondrial cytochrome oxidase expression are increased in
hearts of exercise-​trained old rats. Am J Physiol, 1996. 271(1 Pt 2): p. H68–​72.
57. Kinugawa, K., et al., Regulation of thyroid hormone receptor isoforms in physiological and patho-
logical cardiac hypertrophy. Circ Res, 2001. 89(7): p. 591–​8.
58. Xu, X., et al., Post-​myocardial infarction exercise training beneficially regulates thyroid hormone
receptor isoforms. J Physiol Sci, 2018. 68(6): p. 743–​8.
59. Schunkert, H., et al., Association between a deletion polymorphism of the angiotensin-​converting-​
enzyme gene and left ventricular hypertrophy. N Engl J Med, 1994. 330(23): p. 1634–​8.
60. Celentano, A., et al., Cardiovascular risk factors, angiotensin-​converting enzyme gene I/​D poly-
morphism, and left ventricular mass in systemic hypertension. Am J Cardiol, 1999. 83(8): p. 1196–​200.
61. Kuznetsova, T., et al., Left ventricular mass in relation to genetic variation in angiotensin II receptors,
renin system genes, and sodium excretion. Circulation, 2004. 110(17): p. 2644–​50.
62. Jamshidi, Y., et al., Peroxisome proliferator –​activated receptor alpha gene regulates left ventricular
growth in response to exercise and hypertension. Circulation, 2002. 105(8): p. 950–​5.
63. Poch, E., et al., Genetic variation of the gamma subunit of the epithelial Na+ channel and essential
hypertension. Relationship with salt sensitivity. Am J Hypertens, 2000. 13(6 Pt 1): p. 648–​53.
64. Kupari, M., et al., Left ventricular size, mass, and function in relation to angiotensin-​converting
enzyme gene polymorphism in humans. Am J Physiol, 1994. 267(3 Pt 2): p. H1107–​11.
65. Puthucheary, Z., et al., The ACE gene and human performance: 12 years on. Sports Med, 2011.
41(6): p. 433–​48.
66. Coates, D., The angiotensin converting enzyme (ACE). Int J Biochem Cell Biol, 2003. 35(6): p.
769–​73.
67. Gunel, T., et al., Effect of angiotensin I-​converting enzyme and alpha-​actinin-​3 gene polymorphisms
on sport performance. Mol Med Rep, 2014. 9(4): p. 1422–​6.
68. Molkentin, J.D., et al., A calcineurin-​dependent transcriptional pathway for cardiac hypertrophy.
Cell, 1998. 93(2): p. 215–​28.
69. Bueno, O.F., et al., Impaired cardiac hypertrophic response in calcineurin Abeta-​deficient mice. Proc
Natl Acad Sci U S A, 2002. 99(7): p. 4586–​91.
70. He, Z.H., et al., Polymorphisms in the calcineurin genes are associated with the training responsive-
ness of cardiac phenotypes in Chinese young adults. Eur J Appl Physiol, 2010. 110(4): p. 761–​7.
71. He, Z., et al., Association between HMOX-​1 genotype and cardiac function during exercise. Appl
Physiol Nutr Metab, 2008. 33(3): p. 450–​60.
72. Howden, R., et al., Cardiac physiologic and genetic predictors of hyperoxia-​induced acute lung
injury in mice. Am J Respir Cell Mol Biol, 2012. 46(4): p. 470–​8.
73. American College of Sports Medicine, et al., American College of Sports Medicine position
stand. Exercise and physical activity for older adults. Med Sci Sports Exerc, 2009. 41(7): p. 1510–​30.
74. Lewitter, F.I., et al., Genetic and environmental determinants of level of pulmonary function. Am J
Epidemiol, 1984. 120(4): p. 518–​30.
75. McClearn, G.E., et al., Genetic and environmental influences on pulmonary function in aging
Swedish twins. J Gerontol, 1994. 49(6): p. 264–​8.
76. Hubert, H.B., et al., Genetic and environmental influences on pulmonary function in adult twins.
Am Rev Respir Dis, 1982. 125(4): p. 409–​15.
77. Wilk, J.B., et al., Linkage and association with pulmonary function measures on chromosome 6q27
in the Framingham Heart Study. Hum Mol Genet, 2003. 12(21): p. 2745–​51.
78. Levine, B.D. and J. Stray-​Gundersen, Point: positive effects of intermittent hypoxia (live high:train
low) on exercise performance are mediated primarily by augmented red cell volume. J Appl Physiol
(1985), 2005. 99(5): p. 2053–​5.

197
198

R. Howden et al.

79. Wehrlin, J.P., et al., Live high-​train low for 24 days increases hemoglobin mass and red cell volume
in elite endurance athletes. J Appl Physiol (1985), 2006. 100(6): p. 1938–​45.
80. Stray-​Gundersen, J., R.F. Chapman, and B.D. Levine, “Living high-​ training low” altitude
training improves sea level performance in male and female elite runners. J Appl Physiol (1985), 2001.
91(3): p. 1113–​20.
81. Ashenden, M.J., et al., “Live high, train low” does not change the total haemoglobin mass of male
endurance athletes sleeping at a simulated altitude of 3000 m for 23 nights. Eur J Appl Physiol Occup
Physiol, 1999. 80(5): p. 479–​84.
82. Ashenden, M.J., et al., Effects of a 12-​day “live high, train low” camp on reticulocyte production and
haemoglobin mass in elite female road cyclists. Eur J Appl Physiol Occup Physiol, 1999. 80(5): p. 472–​8.
83. Gore, C.J., et al., Increased serum erythropoietin but not red cell production after 4 wk of intermit-
tent hypobaric hypoxia (4,000–​5,500 m). J Appl Physiol (1985), 2006. 101(5): p. 1386–​93.
84. Friedmann, B., et al., Individual variation in the reduction of heart rate and performance at lactate
thresholds in acute normobaric hypoxia. Int J Sports Med, 2005. 26(7): p. 531–​6.
85. Lewis, B.S., et al., Angiogenesis by gene therapy: a new horizon for myocardial revascularization?
Cardiovasc Res, 1997. 35(3): p. 490–​7.
86. Schraufnagel, D.E., Monocrotaline-​induced angiogenesis. Differences in the bronchial and pul-
monary vasculature. Am J Pathol, 1990. 137(5): p. 1083–​90.
87. Peao, M.N., et al., Neoformation of blood vessels in association with rat lung fibrosis induced by
bleomycin. Anat Rec, 1994. 238(1): p. 57–​67.
88. West, J.B., Human physiology at extreme altitudes on Mount Everest. Science, 1984. 223(4638): p. 784–​8.
89. Powers, S.K. and E.T. Howley, Exercise Physiology: Theory and Application to Fitness and Performance.
New York: McGraw-​Hill Humanities/​Social Sciences/​Languages, 2014.
90. Scroop, G.C. and N.J. Shipp, Exercise-​induced hypoxemia: fact or fallacy? Med Sci Sports Exerc,
2010. 42(1): p. 120–​6.
91. West, J.B., et al., Maximal exercise at extreme altitudes on Mount Everest. J Appl Physiol Respir
Environ Exerc Physiol, 1983. 55(3): p. 688–​98.
92. Peacock, A.J. and P.L. Jones, Gas exchange at extreme altitude: results from the British 40th
Anniversary Everest Expedition. Eur Respir J, 1997. 10(7): p. 1439–​44.
93. Grocott, M.P., et al., Arterial blood gases and oxygen content in climbers on Mount Everest. N Engl
J Med, 2009. 360(2): p. 140–​9.
94. West, J.B., Arterial blood measurements in climbers on Mount Everest. Lancet, 2009. 373(9675): p.
1589–​90.
95. Pugh, L.G., et al., Muscular exercise at great altitudes. J Appl Physiol, 1964. 19: p. 431–​40.
96. Bailey, D.M., The last “oxygenless” ascent of Mt. Everest. Br J Sports Med, 2001. 35(5): p. 294–​6.
97. Burtscher, M., H. Gatterer, and A. Kleinsasser, Cardiorespiratory fitness of high altitude
mountaineers: the underestimated prerequisite. High Alt Med Biol, 2015. 16(2): p. 169–​70.
98. Moore, L.G., S. Niermeyer, and S. Zamudio, Human adaptation to high altitude: regional and
life-​cycle perspectives. Am J Phys Anthropol, 1998. Suppl 27: p. 25–​64.
99. Stobdan, T., J. Karar, and M.A. Pasha, High altitude adaptation: genetic perspectives. High Alt Med
Biol, 2008. 9(2): p. 140–​7.
100. Shriver, M.D., et al., Finding the genes underlying adaptation to hypoxia using genomic scans for
genetic adaptation and admixture mapping. Adv Exp Med Biol, 2006. 588: p. 89–​100.
101. Beall, C.M., et al.,Ventilation and hypoxic ventilatory response of Tibetan and Aymara high altitude
natives. Am J Phys Anthropol, 1997. 104(4): p. 427–​47.
102. Beall, C.M., et al., Quantitative genetic analysis of arterial oxygen saturation in Tibetan highlanders.
Hum Biol, 1997. 69(5): p. 597–​604.
103. Kawakami, Y., et al., Relationship between hypoxic and hypercapnic ventilatory responses in man.
Jpn J Physiol, 1981. 31(3): p. 357–​68.
104. Collins, D.D., et al., Hereditary aspects of decreased hypoxic response. J Clin Invest, 1978. 62(1): p.
105–​10.
105. Moore, G.C., et al., Respiratory failure associated with familial depression of ventilatory response to
hypoxia and hypercapnia. N Engl J Med, 1976. 295(16): p. 861–​5.
106. Saunders, N.A., S.R. Leeder, and A.S. Rebuck,Ventilatory response to carbon dioxide in young
athletes: a family study. Am Rev Respir Dis, 1976. 113(4): p. 497–​502.
107. Beall, C.M.,Tibetan and Andean contrasts in adaptation to high-​altitude hypoxia. Adv Exp Med Biol,
2000. 475: p. 63–​74.

198
199

Cardiorespiratory fitness

108. Aldenderfer, M.S., Moving up in the world: archaeologists seek to understand how and when
people came to occupy the Andean and Tibetan plateaus. Am Sci, 2003. 91(6): p. 542–​549.
109. Smith, T.G., P.A. Robbins, and P.J. Ratcliffe, The human side of hypoxia-​inducible factor. Br J
Haematol, 2008. 141(3): p. 325–​34.
110. Maxwell, P.H., Hypoxia-​ inducible factor as a physiological regulator. Exp Physiol, 2005.
90(6): p. 791–​7.
111. Yi, X., et al., Sequencing of 50 human exomes reveals adaptation to high altitude. Science, 2010.
329(5987): p. 75–​78.
112. West, J.B., American College of Physicians, and American Physiological Society, The
physiologic basis of high-​altitude diseases. Ann Intern Med, 2004. 141(10): p. 789–​800.
113. Semenza, G.L., Hydroxylation of HIF-​1: oxygen sensing at the molecular level. Physiology (Bethesda),
2004. 19: p. 176–​82.
114. Ohno, H., et al., Effect of exercise on HIF-​1 and VEGF signaling. J Phys Fitness Sports Med, 2012.
1(1): p. 5–​16.
115. Faquin, W.C., T.J. Schneider, and M.A. Goldberg, Effect of inflammatory cytokines on hypoxia-​
induced erythropoietin production. Blood, 1992. 79(8): p. 1987–​94.
116. Jelkmann, W., et al., Monokines inhibiting erythropoietin production in human hepatoma cultures
and in isolated perfused rat kidneys. Life Sci, 1992. 50(4): p. 301–​8.
117. Hellwig-​Burgel, T., et al., Interleukin-​ 1beta and tumor necrosis factor-​ alpha stimulate DNA
binding of hypoxia-​inducible factor-​1. Blood, 1999. 94(5): p. 1561–​7.
118. Niebauer, J., et al., Exercise training in chronic heart failure: effects on pro-​inflammatory markers.
Eur J Heart Fail, 2005. 7(2): p. 189–​93.

199
200

16
GENETIC CONTRIBUTIONS TO
MITOCHONDRIAL TRAITS
Mark Tarnopolsky

Introduction
Given the importance of mitochondrial biogenesis to endurance exercise, the aim of this
chapter is to evaluate the potential for variants in mitochondrial DNA (mtDNA), nuclear
genes encoding mitochondrial proteins (NuGEMPs), or nuclear genes that control mitochon-
drial biogenesis to influence mitochondrial function/​adaptations. It has long been known that
increased mitochondrial mass and function are associated with endurance exercise training (33,
35). It is not surprising that maximal oxygen consumption at the whole body level (VO2max)
increases co-​temporal with total mitochondrial mass (12, 53), and mitochondrial cristae density
(64), given that complex IV of the mitochondrial respiratory chain is the terminal oxygen
acceptor for cellular metabolism.
It is important to consider that mitochondria can respond to different stimuli by increasing
the total mass and/​or increasing the efficiency of individual mitochondria. In general, most
studies use citrate synthase activity/​ protein, mitochondrial mRNA transcript abundance,
voltage-​dependent anion channel (VDAC) protein, mtDNA content, or mitochondrial volume
density (electron microscopy) to measure total mitochondrial mass in response to endurance
exercise training (52, 85, 96). A crude functional measure is the ratio of a respiratory chain
enzyme activity (i.e., cytochrome c oxidase (COX))/​total mitochondrial volume (i.e., citrate
synthase); however, classical respirometry or high-​resolution respirometry can be expressed rela-
tive to total mitochondrial or muscle mass and provides a mitochondrial functional metric
(70). It is far beyond this chapter to review every paper evaluating the mitochondrial response
to exercise; however, a general increase in mitochondrial mass is usually seen in response to
training volume (12, 52, 53), with additional small increases in function (capacity/​mitochon-
dria) seen with higher-​intensity interval-​type training (for a review, see 7) or nitrates (42). The
dramatic adaptability of skeletal muscle mitochondrial mass in response to stimuli can also be
seen with hypodynamia, reducing mitochondrial mass by about 25% after only 2 weeks of leg
immobilization (1).
Although it is generally accepted that increased red blood cell mass and stroke volume are
the main determinants of exercise-​induced increases in VO2max (45), the importance of per-
ipheral mitochondrial adaptations is reflected in the severe reduction of VO2max in patients
with mtDNA mutations (94), and impaired peripheral oxygen extraction (91). Consequently,

200
201

Mitochondrial traits

exercise intolerance is a defining feature for many mitochondrial disorders that usually manifests
during endurance type activities (93). A major component of this review will be exploring the
potential for mtDNA, mitochondrial biogenesis-​associated genes, or NuGEMP DNA variants
to influence the increase in VO2peak with endurance exercise training and/​or peak endurance
performance in top sport athletes. I will first provide a background of mitochondrial biology
followed by an exploration of the genetic contributions to mitochondrial traits. Finally, I will
suggest avenues for future research.

Mitochondrial biology
Origin and structure of mitochondria and mitochondrial DNA
Mitochondria are bilayer organelles that exist in nearly every mammalian cell, with the exception
of mature red blood cells. They are thought to have origins as α-​protobacteria that established a
symbiotic relationship with a proto-​eukaryotic cell approximately 1.5 billion years ago to create
eukaryotic cells. With evolution, the majority of the genes encoding for mitochondrial proteins
have been transferred to the nucleus, and these are referred to as NuGEMPs. As a reflection
of their bacterial origin, mitochondria retain their own 16,569 bp intronless circular DNA
(mtDNA) that undergoes polycistronic replication as opposed to gene-​specific expression in
nuclear DNA. mtDNA encodes for 37 gene products including 22 tRNAs, 2 rRNAs, and 13
protein subunits; yet a fully functional mitochondrion requires about 1300 gene products for
proper structure and function with 1158 of these catalogued by the Broad Institute (MitoCarta;
www.broadinstitute.org/​scientific-​community/​science/​programs/​metabolic-​disease-​program/​
publications/​mitocarta/​mitocarta-​in-​0).
Why the few remaining mtDNA gene products are so conserved across the vertebrates and
were not transferred to nuclear DNA control is unclear; however, it is postulated that they
are highly hydrophobic and/​or subtle differences in the genetic code between the genomes
constrained transfer (20), and/​or they are necessary to form the template core of the respira-
tory chain that functions as series circuit (personal communication, Doug Wallace, 2017). The
NuGEMPs that are transcribed in the nucleus and translated in the cytosol are then imported
into the mitochondria through a series of proteins that include the transporter of the outer
mitochondrial membrane proteins (TOMM; i.e., TOMM20, TOMM70), the transporter of the
inner mitochondrial membrane proteins (TIMM; i.e., TIMM23, TIMM22), and mitochondrial
chaperones (i.e., hsp60, hsp72). Mutations in the TIMM8A protein can lead to a mitochondrial
disorder with dystonia and deafness called the Mohr–​Tranebjaerg syndrome (78).
Another unique aspect of mitochondrial genetics is that mtDNA is exclusively inherited
through maternal transmission. This fact allowed for the tracking of Homo sapiens evolution
from an ancestral mitochondrial Eve or matrilinear most recent common ancestor who was
thought to have existed about 100,000–​150,000 years ago in South Africa or Botswana (72), and
allowed for tracking of human evolution out of Africa (15). In fact, any two humans can deter-
mine when they last shared a common mother by analyzing mtDNA sequence variants. Specific
mtDNA sequences that define specific branches of human evolution are called haplogroups.
The first mtDNA haplogroup that defined Homo sapiens is called L and the emergence of L0
defined the mitochondrial Eve (44, 77). Random mutations in mtDNA likely played a role in
evolution, as reflected in the fact that haplogroups found in Siberia render the mitochondria
more uncoupled and this leads to greater heat production that may have promoted survival in
colder environments (81). Haplogroups also influence the penetrance of pathogenic mitochon-
drial DNA mutations (9). Given that mtDNA polymorphisms that define the haplogroups can

201
202

M. Tarnopolsky

Figure 16.1 Mitochondrial ultrastructure. Transmission electron micrograph of mitochondria (25,000×)

before (PRE) and after (POST) 2 months of endurance exercise training. IMF=intermyofibrillar
mitochondria; SS=sub-​sarcolemmal mitochondria; N=nucleus.

also lead to differences in coupled versus uncoupled mitochondrial respiration (39, 81), it is
conceptually attractive to hypothesize that these polymorphisms may partially determine the
evolution of the endurance trait (see Chapter 11). mtDNA also exhibits heteroplasmy within a
cell and tissue with a varying proportion of two different mtDNA variants at the same nucleo-
tide position. This is particularly important for pathogenic mutations where a higher propor-
tion of a mutant versus wild-​type mtDNA variant in skeletal muscle is associated with a greater
deleterious effect on function (37).
Mitochondria have an outer membrane (OMM) and a highly convoluted inner mem-
brane (IMM) that envelopes the matrix and interacts with the OMM through contact sites.
Mitochondria are present in proportion to the energy needs of the tissue, with cardiac cells
having up to 20%, and skeletal muscle about 4% mitochondrial volume density. Mitochondria
are traditionally considered to be static cylindrical tube-​like structures; however, mitochon-
dria form complex dynamic networks within cells (48). Mitochondrial fusion and fission are
important processes to allow for intercellular movement and optimal function (14, 58). This
dynamic movement of mitochondria is likely important in mtDNA complementation, allowing
for regional alterations in mitochondrial abundance in response to intracellular signals, and
altering mitochondrial abundance in response to changes in need for energy. The importance
of fusion and fission is reflected in that mutations within such genes (i.e., OPA1, Mfn2, and
DLP1) result in neurological disease (27, 107). Mitochondria are distributed in sub-​sarcolemmal
and intermyofibrillar regions in skeletal muscle, with increased density after endurance exercise
training (Figure 16.1).

Mitochondrial biogenesis and regulation

Mitochondrial biogenesis is the process of increasing mitochondrial mass in response to
physiological stimuli such as growth or exercise. This is a complex process that involves

202
203

Mitochondrial traits

mitochondrial fusion and fission and mtDNA replication. Even a cursory discussion of these
processes is far beyond the scope of this focused chapter; however, it is important to dis-
cuss the basics of mtDNA replication and transcriptional control during exercise. Seminal
work by Dr. Bruce Spiegelman’s laboratory led to the identification of the “master regu-
lator of mitochondrial biogenesis” called peroxisome proliferator-​activated receptor gamma
coactivator 1-​alpha (PGC1α, gene=PPARGC1A) (74). PGC1α functions as a coactivator of
nuclear transcription factors that upregulate NuGEMPs with histone-​mediated transcrip-
tion factor docking (73). It appears that PGC1α can also migrate to the mitochondria in
response to physiological stimuli to function as an activator of mtDNA replication in con-
junction with mitochondrial transcription factor A (Tfam) and sirtuin 1 (SIRT1) (5). Several
splice variants have been discovered with evidence suggesting that the alternate promotor
derived PGC1α4 mediates skeletal muscle hypertrophy (80). Others have found that both
the truncated PGC1α4 and the canonical nontruncated PGC1α splice variants are activated
by both acute resistance and endurance exercise and the adenosine monophosphate kinase
(AMPK) activator, AICAR (103). There are several cellular signals that trigger mitochon-
drial biogenesis in response to acute exercise including oxidative stress, calcium transients/​
MAP kinase, and alterations in energy charge (i.e., increases in AMP) (11, 23, 36, 54). These
alterations in cellular homeostasis activate several protein kinases that lead to transcription
factor or coactivator activation.
AMPK is considered the main sensor of the exercise signal leading to mitochondrial bio-
genesis (36, 49, 67, 68).The protein is activated by phosphorylation in response to lower energy
charge (higher AMP) during exercise and this leads to phosphorylation and/​or nuclear trans-
location of downstream proteins and kinases that leads to an upregulation of mitochondrial
biogenesis (i.e., PGC1α), activation of autophagy (i.e., ULK1), lipid oxidation (i.e., PPARα/​δ),
peroxisomal biogenesis, and antioxidant defences (10, 67). Important PGC1α-​mediated tran-
scription factors that upregulate NuGEMPs include NRF1, NRF2, MEF2, PPARα/​δ, and
ERRα. The newly transcribed NuGEMP mRNAs are exported to the cytosol for translation
and import into the mitochondria. Oxidative stress can also be an activator of mitochondrial
biogenesis through PGC1α activation (2, 4, 23), and calcium transients can also activate PGC1α
through MAP kinase activation (11).
It is not clear how the exercise signal is transduced to the mitochondria to coordinately
initiate mtDNA transcription but translocation of PGC1α to the mitochondria is a likely can-
didate (5). Given that mtDNA copy number increases with endurance exercise-​mediated mito-
chondrial biogenesis, there must be mtDNA replication (i.e., making new copies of mtDNA).
The replication of mtDNA requires many proteins and Tfam is important in mtDNA stability.
Since mtDNA does not have true chromatin or histones, Tfam is thought to have a histone-​
like function in the mitochondria (40). Replication of mtDNA requires polymerase gamma 1
(POLG1) and POLG2. Mutations in the POLG1 gene and to a lesser extent, POLG2, lead to
a wide spectrum of disorders from fatal infantile hepatoencephalopathy to later-​onset ataxia,
neuropathy, and Parkinsonian features (63, 83).The DNA helicase C10orf2 (twinkle) is required
to unwind the highly coiled mtDNA prior to DNA replication. In contrast to the documenta-
tion of many dozens of pathogenic mutations in the POLG1 gene (83), it may appear unusual
that there are no known mutations responsible for human neurological disease in PPARGC1A
(PGC1α gene) in spite of animal evidence for this possibility (90). The lack of demonstration
of mutations in the PPARGC1A gene in patients with mitochondrial disease may be due to
compensatory upregulation of PGC1β and PRC genes that are also involved in mitochondrial
biogenesis coactivation (79, 88).

203
204

M. Tarnopolsky

ATP > ADP + PCr + H+ > Cr + ATP

CYTOSOL

ACUTE
OUTER
MITOCHONDRIAL EXERCISE VDAC
MEMBRANE

INTER-MEMBRANE PCr Cr
SPACE
mtCK
H+ H+ H+ H+
ADP ATP

INNER e- e-
I CoQ10 e- IV
MITOCHONDRIAL III V ANT
MEMBRANE II e-

NADH+H+ NAD FADH2 FAD O2 H2O ADP ATP

ADP ATP
MATRIX

Figure 16.2 Mitochondrial respiratory chain. Carbohydrate, fat, and some amino acids can be
oxidized and lead to the production of reduced co-​factors, NADH+H+ and FADH2. These reduced
cofactors are oxidized by linked multienzyme complexes embedded in the inner mitochondrial
membrane (respiratory chain) with complex I and II gaining electrons from NADH+H+ and FADH2,
respectively. The electrons are transferred to coenzyme Q10 which becomes reduced and the electrons
are transferred to complex III (ubiquinol–​cytochrome c reductase) and the electrons are then used to
reduce cytochrome c. Cytochrome c is a mobile carrier protein that delivers its electrons to cytochrome
c oxidase (COX, complex IV) and the electrons are used to reduce molecular oxygen to water. The
potential energy from intermediary metabolism is used to pump protons (H+) from the matrix of the
mitochondria to the intermembrane space at complexes I, III, and IV. This causes an increase in proton
concentration (lower pH) and this potential energy is captured as they flow through complex V back into
the matrix to rephosphorylate ADP to ATP. ATP and ADP are translocated across the intermitochondrial
membrane through the adenine nucleotide translocase (ANT) where through the catalytic activity of
mitochondrial creatine kinase (mtCK), ATP is used to rephosphorylate creatine (Cr) to phosphocreatine
(PCr). The phosphocreatine and creatine are exchanged across the outer mitochondrial membrane
through the porin/​voltage-​dependent ion channel (VDAC) and the mitochondrial-​derived
phosphocreatine is then used to rephosphorylate ADP back to ATP in the cytosol (101).

Functions of mitochondria
The classical role of mitochondria is energy metabolism, where reducing equivalents from carbo-
hydrate fat, and to some extent amino acids, can be used to generate ATP in the presence of
oxygen (Figure 16.2). Mitochondria are also involved in a variety of other physiological and
pathophysiologic processes including oxidative stress, inflammasome activation, telomere mainten-
ance, and activation of apoptosis. Given that the respiratory chain depends on the flow of electrons,
it is not surprising that unpaired electrons (free radicals) can be formed by the respiratory chain
with complex I and III being traditionally considered as the main sites of free radical generation
(100). Although an increase in oxidative stress is felt to be a component of neurodegenerative
disorders and even aging (62), free radicals are thought to be signaling molecules in physiological
adaptation and their suppression can attenuate exercise-​induced mitochondrial adaptations (71) and
attenuate adaptations to endurance exercise (29, 69, 76). Some of the negative effects of antioxidants
could be through the inhibition of free radical-​induced stimulation of AMPK, CaMKII (60), and
PGC1α (23). In contrast, others have found no attenuation of exercise adaptations with antioxidants

204
205

Mitochondrial traits

(32, 104–​106) or reported mixed results (61). Dysfunction of the mitochondria can lead to a release
of danger-​associated molecular pattern (DAMP) molecules such as mtDNA and formyl proteins
that can activate an NLRP3-​mediated increase in caspase-​1 activation and the release of charac-
teristic cytokines (IL-​1β, IL-​18) (86). Telomeres function to protect the ends of chromosomes and
are a “replicometer” of aging. Telomere shortening is associated with mitochondrial dysfunction,
partially through an increase in oxidative stress (82). Dysfunctional mitochondria can also release
cytochrome c into the cytosol, activating caspase 9 that, in turn, activates caspases 3 and 7 to trigger
ATP-​dependent preprogrammed cell death (apoptosis).
Given the complexity of the mitochondrion, it is not surprising that there are many mtDNA
and NuGEMP DNA mutations that can impair the ability to generate ATP (3, 34, 37, 99, 107).
The emergence of mitochondrial medicine in the late 1980s focused on the identification of
mtDNA mutations and characterization of many mitochondrial diseases by their impact upon
the maximal enzyme activity (34, 99, 108).The majority of exercise physiology studies focusing
on mitochondrial adaptation have also relied on the measurement of maximal enzyme activity
(usually COX) as a reflection of the respiratory chain capacity with citrate synthase as a measure
of total mitochondrial mass (12, 52, 85). Other proteins such as VDAC and measurements of
mitochondrial DNA can also be used to measure the increase in total mitochondria in response
to endurance exercise training. Of course, these assays determine the Vmax of the enzymes
complexes in isolation, while newer techniques such as high-​resolution respirometry allow for
linked assays, especially in skinned muscle fibers versus isolated mitochondria (70).

Exercise adaptations and mtDNA variants

Mitochondrial disorders
The most dramatic influence of mtDNA variants upon exercise capacity are the deleterious
effects of specific mutations (3, 93, 97). The most striking influence of a single mtDNA point
mutation we have seen in the laboratory is the severe reduction of VO2max in patients with
the mitochondrial encephalomyopathy, lactic acidosis, and stroke-​ like episodes (MELAS)
m.3271T>C and MELAS m.3243A>G transition variants of <10 ml/​kg/​min (97). Numerous
publications have reported exercise intolerance, low aerobic capacity, and lifelong exercise
intolerance in patients with a variety of mtDNA point mutations (18, 19, 37, 93, 95), with cyto-
chrome b mutations being rather muscle symptom specific (3). Encouragingly, even with such a
severe exercise intolerance, endurance exercise training can improve exercise capacity,VO2max,
and quality of life in mitochondrial disease patients through an increase in mitochondrial mass
(38, 94).

Healthy individuals
There have been several summaries regarding the influence of mtDNA variants upon ath-
letic performance (8, 25), and here I will highlight some of the main studies. A study of 114
healthy Spanish men found that those with haplogroup J had lower VO2max as compared
with non-​J haplogroups (59). The lower VO2max was thought to be related to the known
lower oxidative stress (reactive oxygen species (ROS) production) and ATP generation effi-
ciency of mitochondrial of haplogroup J (59). In another study, the enrichment of mtDNA
haplogroup T (m.13368A) was less in elite male endurance athletes versus sedentary controls
and thus, haplogroup T was considered a negative predictor of elite athletic performance (13).
Finally, another Spanish study confirmed the lower VO2max association with haplogroup J in
70 healthy sedentary men, but not in 50 aerobically trained men (50).

205
206

M. Tarnopolsky

In contrast, is there evidence that specific mtDNA haplogroups (as defined by single or
several transition or transversion variants) or single nucleotide mtDNA polymorphisms can
positively influence mitochondrial traits? The first attempt to study the influence of mtDNA
variants upon VO2max or the adaptation to endurance training was by Dr. Claude Bouchard’s
group in 1991 and 1993 where polymerase chain reaction restriction fragment length poly-
morphism (PCR-​RFLP) analysis found that men who carried unique patterns in two regions
of the ND5 mtDNA subunit and one region in the tRNAthr gene had higher VO2max levels
at baseline, while those with an ND2 restriction pattern had lower VO2max levels at baseline
(21, 22). After endurance exercise training, those with a HincII restriction pattern in ND5
showed lower VO2max responses (22). Another Spanish group looked at haplogroups that were
enriched in elite male Caucasian endurance athletes and found that 15.7% of the elite endur-
ance athletes were from haplogroup V, with fewer in the sedentary cohort (7.5%) (66). Another
Spanish study looked at the mtDNA haplotype of 81 healthy men and found that those with J
had lower VO2max values than non-​J, with H accounting for most of the difference (51). The
finding of a lower VO2max in haplogroup J was confirmed in 114 healthy male Spanish subjects.
Interestingly, haplogroup H also showed higher oxidative damage versus J (50, 51). It would
have been of interest to perform skeletal muscle electron microscopy and PGC1α analysis in
the latter study to see if the association was mediated by ROS induction of mitochondrial bio-
genesis through a PGC1α-​mediated mechanism.
A Japanese study evaluated the mtDNA sequence of 100 elite male endurance athletes and
672 controls and found that the endurance athletes showed a slight but significant enrich-
ment of sub-​haplogroups D4e2 and D4g (55). This same Japanese group studied 79 Olympic-​
level endurance/​middle-​distance athletes versus 672 controls and found an enrichment of 8.9%
versus 3.7%, respectively for haplogroup G1 (57). They examined 100 elite male athletes in
the first study and the 672 controls from the second study and found that the athletes had an
enrichment of m.152C>T, m.514(CA)n (CA ≥5) and a poly-​A stretch at m.568–​573 (C ≥7)
in the noncoding control region (56). Finally, a Korean study compared 75 endurance/​middle-​
distance athletes to 265 nonathletic controls and found an enrichment for haplogroups M* and
N9 and a paucity of B compared to the control group (41).

Exercise adaptations, NuGEMP variants, and mitochondrial

biogenesis genes
Mitochondrial disorders
As with mtDNA variants, there are several NuGEMP mutations associated with severe endur-
ance exercise intolerance including YARS2, ISCU, TMEM126B, and TRMT5 (43, 87). To my
knowledge, there have not been reports of a significant relationship between nuclear-​encoded
subunits of the respiratory chain and VO2max or endurance exercise adaptability. Given that
the mitochondrial adaptations to endurance exercise are predominantly through changes in
mitochondrial mass mediated by mitochondrial biogenesis and that mitochondrial biogenesis
has PGC1α as a “master regulator,” it is not surprising that most of the described genetic associ-
ations for mitochondrial traits are found in this gene and associated biogenesis genes.

Healthy individuals
A large study looked at maximal treadmill exercise duration in 3783 young, white and black
men and women at baseline and again after a 20-​year period (N=2335) (84). They found that

206
207

Mitochondrial traits

a favorable single nucleotide polymorphism (SNP) in PPARGC1A (the PGC1α gene) and
three other genes was associated with baseline treadmill performance in black individuals
and specific SNPs in PPARGC1A (rs3774909) and HIF1A (rs1957757) were associated with
baseline treadmill performance in white individuals. Four favorable SNPs in different genes
(including PPARGC1A) were associated with less decline in treadmill performance after
20 years (84). Given the role of PGC1α as the master regulator of mitochondrial biogenesis
(74), it is not surprising that the PPARGC1A gene came up on an unbiased gene discovery
platform.
There is a common PPARGC1A SNP termed Gly482Ser (rs8192678) seen in about 35–​
50% of humans (28). It has been shown that humans with this specific SNP show an attenuated
binding of PGC1α to MEF2 (109), and impaired TFAM expression (16).This SNP impairs skel-
etal muscle slow-​twitch fiber-​type transition following endurance exercise training (89). Not
surprisingly the Gly482Ser SNP was slightly, but significantly, less prevalent in elite endurance
athletes (29.1%) as compared to sedentary controls with a 2.5-​fold lower VO2max (40.0%) (47).
Others have found a lower enrichment of homozygous 482Ser in elite Caucasian endurance
athletes (46). Somewhat paradoxically, it has been shown that homozygosity for 482Ser (~11%
of the population) has the greatest propensity to adapt to endurance exercise (28). Furthermore,
ethnicity may play a role in these associations as a Japanese cohort found that homozygosity for
482Ser was associated with a higher lactate threshold with no influence on VO2max (65), and
there was no relationship between the Gly482Ser polymorphism and VO2max adaptation to
endurance training (65). Another study in Turkish elite endurance athletes found that Gly482
was more enriched in controls (60%) versus endurance athletes (38.3%) (98). Finally, a study in
elite Chinese endurance athletes found no enrichment for any of 133 SNPs within 11 genes
associated with PGC1α (PPARGC1A, PPARGC1B, PPRC1, TFAM, TFB1M, TFB2M, NRF1,
GABPA, GABPB1, ERRα, and SIRT1) (30).
Independent associations between PPARα/​δ SNPs have also been evaluated in the context
of endurance capacity. PPARδ is encoded by the PPARD gene and this protein is involved in
mitochondrial function and lipid oxidation. A large study evaluated the enrichment of three
specific SNPs (rs2267668, rs2016520, and rs1053049) in the PPARD gene in 704 controls and
120 endurance athletes and found a significantly higher enrichment of the A/​C/​C haplotype
in the controls (7.25%) versus the athletes (0.83%) (47). Another study found that the specific
PPARD DNA variant (c.294T>C) associated with the rs2016520 SNP was not more enriched
in top sport endurance athletes; however, the presence of homozygosity for PPARGC1A
Gly482 and PPARD CC genotype was more prevalent in the athletes (24). PPARα is encoded
for by the PPARA and has similar functions to PPARδ. One study found that a specific PPARA
polymorphism G/​G (rs4253778) was slightly more enriched (P=0.051) associated in 155 Israeli
endurance athletes (N=155) versus controls (N=240), and they confirmed a lower enrichment
of the PPARGC1A Gly482Ser SNP (47). Another study looked at a PPARA intron 7 variant
(G>C; rs4253778) and found that G/​G was enriched in endurance athletes (N=60) versus
controls (N=110) (98).
Compared to single SNP analysis methods, a more complex analysis involves the potential
interacting effects of multiple SNPs in genes involved in pathways related to aerobic capacity.
As one can imagine, there are likely dozens to hundreds of potential SNPs in any given gene
and hundreds of genes involved in aspects of aerobic capacity (cardiac output, capillary density,
energy storage and mobilization, and mitochondrial biogenesis). Although one study found an
enrichment of PPARGC1A Gly482 and PPARD CC in endurance athletes (24), when this
group evaluated an “endurance genotype score” based upon the sum of these scores from six
polymorphisms within the PPARGC1A-​NRF-​TFAM axis they found that none of the elite

207
208

M. Tarnopolsky

endurance athletes (N=81) had the ideal (100) genotype score; however, the score was higher
(38.9%) for the athletes versus controls (30.6%) (26).

Conclusion
Given the complexities of biological systems (as highlighted by Dr. Bourchard’s closing
comments), and the fact that 1.5 billion years of mitochondrial evolution would have placed
enormous selective pressure on achieving the most efficient mitochondria, it is not surprising
that a given haplogroup/​mtDNA variants does not strongly associate with VO2max or the
adaptive propensity for mitochondrial adaptation. When an enrichment of a haplogroup is
reported to segregate with VO2max or endurance adaptation, the relative difference between
the athletes and controls or the responders and nonresponders can appear large by relative terms
(up to 100% different); however, most studies show that the absolute enrichment and differences
between the groups is in the 5–​15% range. Consequently, mtDNA variants should not be used
to encourage or dissuade an athlete from sport participation. Furthermore, even with severe
mtDNA mutations that massively lower VO2max, endurance exercise training can still lead to a
significant increase in VO2max, exercise performance, and quality of life (92).
A big issue when looking at mtDNA haplogroups is that within a given country the enrich-
ment of the haplogroups will be somewhat homogenous based upon evolutionary migration
patterns (i.e., the haplogroups in Korea and Japan will be very different than those in Spain). In
addition, for strong statistical association studies even with common diseases (31), the sample
size will need to be much larger and include many ethnicities as compared to the rather small
and country/​ethnicity-​centric studies published to date. Another serious issue is that most of
the studies exclusively used men and it is not appropriate to extrapolate the findings from
men to women. The latter issue is particularly apparent given that women, likely through 17-​
β-​estradiol, have lower oxidative stress; and a haplogroup such as H that likely works through
higher ROS-​inducing mitochondrial biogenesis would be less likely to associate in women.
By the same rationale of evolutionary pressure on mitochondrial efficiency, it would be
surprising if NuGEMPs that encode for mitochondrial structural proteins would be strongly
associated with mitochondrial adaptation. Even DNA variants in mitochondrial biogenesis
proteins including PGC1α are not strongly associated with VO2max or exercise capacity and
these associations can be opposite depending upon the ethnicity. It is highly likely that the lack
of a strong association between a SNP in any single gene and exercise performance or VO2max
is due to many factors, not the least of which is that the effect of any SNP is likely influenced
by genetic factors such as a permissive haplotype, SNPs in biogenesis-​associated genes, SNPs in
mitochondrial structural genes, and SNPs in other related mitochondrial proteins such as mem-
brane transporters, chaperones, nucleotide transporters, etc.The fact that about 30% of top sport
endurance athletes have an “unfavorable” PPARGC1A SNP (24, 46) shows that these variants
do not exert an overriding influence on endurance athletic performance. Finally, it is known
that epigenetic influences (including gut microbiota) can strongly influence gene expression
(6), even in the mitochondria (17), and these effects can hinder the ability to make genotype–​
phenotype correlations.
Another issue to consider in future studies is the influence of new technology upon the ability
to discover genetic determinants of exercise capacity. For example, the first study that I am aware
of that looked at mtDNA variants and endurance adaptations (21), used PCR-​RFLP to look at
a few mtDNA variants detected by the creation or removal of a restriction site; whereas, next-​
generation sequencing (NGS) can now provide up to 5000× reads of every base pair (16) of the
mtDNA. Many of the genotype/​phenotype association studies have used SNP-​chip methods

208
209

Mitochondrial traits

to look at specified SNPs, yet NGS-​based whole-​exome sequencing can provide an unbiased
evaluation of all the nucleotides within the exons (~1% of all genomic material) of the entire
genome (~22,000 genes). To explore the potential for variants in the intronic regions (~99% of
all genomic material), it is now possible to use whole-​genome sequencing. The latter method
is particularly powerful when used in conjunction with RNA sequencing that can look at
sequence variants in the transcribed exons and alternative splicing variants. Consequently,
determining the contribution of DNA variants to exercise capacity will require large numbers
of subjects and a combination of mtDNA sequencing, whole-​genome sequencing, methylation
arrays, and RNA sequencing in skeletal muscle. Initiatives such as the National Institutes of
Health-​sponsored Molecular Transducers of Physical Activity in Humans (www.commonfund.
nih.gov/​moleculartransducers/​overview), and the Gene SMART Study (102), will likely pro-
vide important answers to many of the questions concerning genetic contributions to mito-
chondrial traits. To further illustrate the complexity of the issue, a large study (GAMES) with
1520 endurance athletes and 2760 controls from Australia, Ethiopia, Japan, Kenya, Poland,
Russia, and Spain failed to find a marker or series of genetic markers that strongly segregated
with the endurance phenotype (75).

Acknowledgements: Dr. Tarnopolsky’s research in the area of exercise biochemistry and

mitochondrial medicine is funded by the Canadian Institute for Health Research and kind
donations from Dan Wright and family, Warren Lammert and family, and Giant Tiger Stores.
Dr Lauren Skelly kindly provided the electron microscopy images in Figure 16.1.

References
1. Abadi A, Glover EI, Isfort RJ, Raha S, Safdar A, Yasuda N, Kaczor JJ, Melov S, Hubbard
A, Qu X, Phillips SM, and Tarnopolsky M. Limb immobilization induces a coordinate down-​
regulation of mitochondrial and other metabolic pathways in men and women. PLoS One
4: e6518, 2009.
2. Adamovich Y, Shlomai A, Tsvetkov P, Umansky KB, Reuven N, Estall JL, Spiegelman BM,
and Shaul Y. The protein level of PGC-​1alpha, a key metabolic regulator, is controlled by NADH-​
NQO1. Mol Cell Biol 33: 2603–​2613, 2013.
3. Andreu AL, Hanna MG, Reichmann H, Bruno C, Penn AS, Tanji K, Pallotti F, Iwata S,
Bonilla E, Lach B, Morgan-​Hughes J, and DiMauro S. Exercise intolerance due to mutations in
the cytochrome b gene of mitochondrial DNA. N Engl J Med 341: 1037–​1044, 1999.
4. Aquilano K, Baldelli S, Pagliei B, Cannata SM, Rotilio G, and Ciriolo MR. p53 orchestrates
the PGC-​1alpha-​mediated antioxidant response upon mild redox and metabolic imbalance. Antioxid
Redox Signal 18: 386–​399, 2013.
5. Aquilano K, Vigilanza P, Baldelli S, Pagliei B, Rotilio G, and Ciriolo MR. Peroxisome
proliferator-​
activated receptor gamma co-​ activator 1alpha (PGC-​1alpha) and sirtuin 1 (SIRT1)
reside in mitochondria: possible direct function in mitochondrial biogenesis. J Biol Chem 285:
21590–​21599, 2010.
6. Barres R, Osler ME, Yan J, Rune A, Fritz T, Caidahl K, Krook A, and Zierath JR. Non-​CpG
methylation of the PGC-​1alpha promoter through DNMT3B controls mitochondrial density. Cell
Metab 10: 189–​198, 2009.
7. Bishop DJ, Granata C, and Eynon N. Can we optimise the exercise training prescription to maximise
improvements in mitochondria function and content? Biochim Biophys Acta 1840: 1266–​1275, 2014.
8. Bray MS, Hagberg JM, Perusse L, Rankinen T, Roth SM, Wolfarth B, and Bouchard C. The
human gene map for performance and health-​related fitness phenotypes: the 2006–​2007 update. Med
Sci Sports Exerc 41: 35–​73, 2009.
9. Brown MD, Starikovskaya E, Derbeneva O, Hosseini S, Allen JC, Mikhailovskaya IE, Sukernik
RI, and Wallace DC. The role of mtDNA background in disease expression: a new primary LHON
mutation associated with Western Eurasian haplogroup J. Hum Genet 110: 130–​138, 2002.

209
210

M. Tarnopolsky

10. Bujak AL, Crane JD, Lally JS, Ford RJ, Kang SJ, Rebalka IA, Green AE, Kemp BE, Hawke TJ,
Schertzer JD, and Steinberg GR. AMPK activation of muscle autophagy prevents fasting-​induced
hypoglycemia and myopathy during aging. Cell Metab 21: 883–​890, 2015.
11. Cao W, Daniel KW, Robidoux J, Puigserver P, Medvedev AV, Bai X, Floering LM,
Spiegelman BM, and Collins S. p38 mitogen-​activated protein kinase is the central regulator
of cyclic AMP-​dependent transcription of the brown fat uncoupling protein 1 gene. Mol Cell Biol
24: 3057–​3067, 2004.
12. Carter SL, Rennie CD, Hamilton SJ, and Tarnopolsky. Changes in skeletal muscle in males and
females following endurance training. Can J Physiol Pharmacol 79: 386–​392, 2001.
13. Castro MG, Terrados N, Reguero JR, Alvarez V, and Coto E. Mitochondrial haplogroup T is
negatively associated with the status of elite endurance athlete. Mitochondrion 7: 354–​357, 2007.
14. Chen H, Vermulst M, Wang YE, Chomyn A, Prolla TA, McCaffery JM, and Chan DC.
Mitochondrial fusion is required for mtDNA stability in skeletal muscle and tolerance of mtDNA
mutations. Cell 141: 280–​289, 2010.
15. Chen YS, Torroni A, Excoffier L, Santachiara-​Benerecetti AS, and Wallace DC. Analysis of
mtDNA variation in African populations reveals the most ancient of all human continent-​specific
haplogroups. Am J Hum Genet 57: 133–​149, 1995.
16. Choi YS, Hong JM, Lim S, Ko KS, and Pak YK. Impaired coactivator activity of the Gly482 variant
of peroxisome proliferator-​activated receptor gamma coactivator-​1alpha (PGC-​1alpha) on mito-
chondrial transcription factor A (Tfam) promoter. Biochem Biophys Res Commun 344: 708–​712, 2006.
17. Clark A and Mach N. The crosstalk between the gut microbiota and mitochondria during exercise.
Front Physiol 8: 319, 2017.
18. Connolly BS, Feigenbaum AS, Robinson BH, Dipchand AI, Simon DK, and Tarnopolsky
MA. MELAS syndrome, cardiomyopathy, rhabdomyolysis, and autism associated with the A3260G
mitochondrial DNA mutation. Biochem Biophys Res Commun 402: 443–​447, 2010.
19. Darin N, Hedberg-​Oldfors C, Kroksmark AK, Moslemi AR, Kollberg G, and Oldfors A.
Benign mitochondrial myopathy with exercise intolerance in a large multigeneration family due to a
homoplasmic m.3250T>C mutation in MTTL1. Eur J Neurol 24: 587–​593, 2017.
20. de Grey AD. Forces maintaining organellar genomes: is any as strong as genetic code disparity or
hydrophobicity? Bioessays 27: 436–​446, 2005.
21. Dionne FT, Turcotte L, Thibault MC, Boulay MR, Skinner JS, and Bouchard C. Mitochondrial
DNA sequence polymorphism, VO2max, and response to endurance training. Med Sci Sports Exerc
23: 177–​185, 1991.
22. Dionne FT, Turcotte L, Thibault MC, Boulay MR, Skinner JS, and Bouchard C. Mitochondrial
DNA sequence polymorphism, VO2max, and response to endurance training. Med Sci Sports Exerc
25: 766–​774, 1993.
23. Espinoza MB, Aedo JE, Zuloaga R, Valenzuela C, Molina A, and Valdes JA. Cortisol induces
reactive oxygen species through a membrane glucocorticoid receptor in rainbow trout myotubes. J
Cell Biochem 118: 718–​725, 2017.
24. Eynon N, Meckel Y, Alves AJ, Yamin C, Sagiv M, Goldhammer E, and Sagiv M. Is there
an interaction between PPARD T294C and PPARGC1A Gly482Ser polymorphisms and human
endurance performance? Exp Physiol 94: 1147–​1152, 2009.
25. Eynon N, Moran M, Birk R, and Lucia A. The champions’ mitochondria: is it genetically
determined? A review on mitochondrial DNA and elite athletic performance. Physiol Genomics
43: 789–​798, 2011.
26. Eynon N, Ruiz JR, Meckel Y, Moran M, and Lucia A. Mitochondrial biogenesis related endur-
ance genotype score and sports performance in athletes. Mitochondrion 11: 64–​69, 2011.
27. Feely SM, Laura M, Siskind CE, Sottile S, Davis M, Gibbons VS, Reilly MM, and Shy
ME. MFN2 mutations cause severe phenotypes in most patients with CMT2A. Neurology
76: 1690–​1696, 2011.
28. Franks PW, Barroso I, Luan J, Ekelund U, Crowley VE, Brage S, Sandhu MS, Jakes RW,
Middelberg RP, Harding AH, Schafer AJ, O’Rahilly S, and Wareham NJ. PGC-​1alpha geno-
type modifies the association of volitional energy expenditure with [OV0312]O2max. Med Sci Sports
Exerc 35: 1998–​2004, 2003.
29. Gomez-​Cabrera MC, Domenech E, Romagnoli M, Arduini A, Borras C, Pallardo FV, Sastre
J, and Vina J. Oral administration of vitamin C decreases muscle mitochondrial biogenesis and
hampers training-​induced adaptations in endurance performance. Am J Clin Nutr 87: 142–​149, 2008.

210
211

Mitochondrial traits

30. He ZH, Hu Y, Li YC, Gong LJ, Cieszczyk P, Maciejewska-​Karlowska A, Leonska-​Duniec A,

Muniesa CA, Marin-​Peiro M, Santiago C, Garatachea N, Eynon N, and Lucia A. PGC-​related
gene variants and elite endurance athletic status in a Chinese cohort: a functional study. Scand J Med
Sci Sports 25: 184–​195, 2015.
31. Herrnstadt C and Howell N. An evolutionary perspective on pathogenic mtDNA
mutations: haplogroup associations of clinical disorders. Mitochondrion 4: 791–​798, 2004.
32. Higashida K, Kim SH, Higuchi M, Holloszy JO, and Han DH. Normal adaptations to exercise
despite protection against oxidative stress. Am J Physiol Endocrinol Metab 301: E779–​784, 2011.
33. Holloszy JO, Oscai LB, Don IJ, and Mole PA. Mitochondrial citric acid cycle and related
enzymes: adaptive response to exercise. Biochem Biophys Res Commun 40: 1368–​1373, 1970.
34. Holt IJ, Harding AE, and Morgan-​Hughes JA. Deletions of muscle mitochondrial DNA in
patients with mitochondrial myopathies. Nature 331: 717–​719, 1988.
35. Hoppeler H, Luthi P, Claassen H, Weibel ER, and Howald H. The ultrastructure of the normal
human skeletal muscle. A morphometric analysis on untrained men, women and well-​ trained
orienteers. Pflugers Arch 344: 217–​232, 1973.
36. Jager S, Handschin C, St-​Pierre J, and Spiegelman BM. AMP-​activated protein kinase (AMPK)
action in skeletal muscle via direct phosphorylation of PGC-​1alpha. Proc Natl Acad Sci U S A
104: 12017–​12022, 2007.
37. Jeppesen TD, Schwartz M, Frederiksen AL, Wibrand F, Olsen DB, and Vissing J. Muscle
phenotype and mutation load in 51 persons with the 3243A>G mitochondrial DNA mutation. Arch
Neurol 63: 1701–​1706, 2006.
38. Jeppesen TD, Schwartz M, Olsen DB, Wibrand F, Krag T, Duno M, Hauerslev S, and Vissing
J. Aerobic training is safe and improves exercise capacity in patients with mitochondrial myopathy.
Brain 129: 3402–​3412, 2006.
39. Ji F, Sharpley MS, Derbeneva O, Alves LS, Qian P, Wang Y, Chalkia D, Lvova M, Xu J, Yao
W, Simon M, Platt J, Xu S, Angelin A, Davila A, Huang T, Wang PH, Chuang LM, Moore
LG, Qian G, and Wallace DC. Mitochondrial DNA variant associated with Leber hereditary optic
neuropathy and high-​altitude Tibetans. Proc Natl Acad Sci U S A 109: 7391–​7396, 2012.
40. Kanki T, Nakayama H, Sasaki N, Takio K, Alam TI, Hamasaki N, and Kang D. Mitochondrial
nucleoid and transcription factor A. Ann N Y Acad Sci 1011: 61–​68, 2004.
41. Kim KC, Cho HI, and Kim W. MtDNA haplogroups and elite Korean athlete status. Int J Sports
Med 33: 76–​80, 2012.
42. Larsen FJ, Schiffer TA, Borniquel S, Sahlin K, Ekblom B, Lundberg JO, and Weitzberg
E. Dietary inorganic nitrate improves mitochondrial efficiency in humans. Cell Metab
13: 149–​159, 2011.
43. Legati A, Reyes A, Ceccatelli Berti C, Stehling O, Marchet S, Lamperti C, Ferrari A,
Robinson AJ, Muhlenhoff U, Lill R, Zeviani M, Goffrini P, and Ghezzi D. A novel de novo
dominant mutation in ISCU associated with mitochondrial myopathy. J Med Genet 54: 815–​824, 2017.
44. Lewin R. The unmasking of mitochondrial Eve. Science 238: 24–​26, 1987.
45. Lundby C, Montero D, and Joyner M. Biology of VO2 max: looking under the physiology lamp.
Acta Physiol (Oxf) 220: 218–​228, 2017.
46. Maciejewska A, Sawczuk M, Cieszczyk P, Mozhayskaya IA, and Ahmetov, II.The PPARGC1A
gene Gly482Ser in Polish and Russian athletes. J Sports Sci 30: 101–​113, 2012.
47. Maciejewska-​Karlowska A, Hanson ED, Sawczuk M, Cieszczyk P, and Eynon N. Genomic
haplotype within the peroxisome proliferator-​activated receptor delta (PPARD) gene is associated
with elite athletic status. Scand J Med Sci Sports 24: e148–​155, 2014.
48. Mannella CA, Buttle K, Rath BK, and Marko M. Electron microscopic tomography of rat-​liver
mitochondria and their interaction with the endoplasmic reticulum. Biofactors 8: 225–​228, 1998.
49. Marcinko K, Sikkema SR, Samaan MC, Kemp BE, Fullerton MD, and Steinberg GR.
High intensity interval training improves liver and adipose tissue insulin sensitivity. Mol Metab
4: 903–​915, 2015.
50. Marcuello A, Martinez-​Redondo D, Dahmani Y, Casajus JA, Ruiz-​Pesini E, Montoya J,
Lopez-​Perez MJ, and Diez-​Sanchez C. Human mitochondrial variants influence on oxygen con-
sumption. Mitochondrion 9: 27–​30, 2009.
51. Martinez-​Redondo D, Marcuello A, Casajus JA, Ara I, Dahmani Y, Montoya J, Ruiz-​Pesini
E, Lopez-​Perez MJ, and Diez-​Sanchez C. Human mitochondrial haplogroup H: the highest
VO2max consumer –​is it a paradox? Mitochondrion 10: 102–​107, 2010.

211
212

M. Tarnopolsky

52. McKenzie S, Phillips SM, Carter SL, Lowther S, Gibala MJ, and Tarnopolsky MA. Endurance
exercise training attenuates leucine oxidation and BCOAD activation during exercise in humans. Am
J Physiol Endocrinol Metab 278: E580–​587, 2000.
53. Meinild Lundby AK, Jacobs RA, Gehrig S, de Leur J, Hauser M, Bonne TC, Fluck D,
Dandanell S, Kirk N, Kaech A, Ziegler U, Larsen S, and Lundby C. Exercise training increases
skeletal muscle mitochondrial volume density by enlargement of existing mitochondria and not de
novo biogenesis. Acta Physiol (Oxf) 222, 2018.
54. Merry TL and Ristow M. Nuclear factor erythroid-​derived 2-​like 2 (NFE2L2, Nrf2) mediates
exercise-​induced mitochondrial biogenesis and the anti-​ oxidant response in mice. J Physiol
594: 5195–​5207, 2016.
55. Mikami E, Fuku N, Kong QP, Takahashi H, Ohiwa N, Murakami H, Miyachi M, Higuchi M,
Tanaka M, Pitsiladis YP, and Kawahara T. Comprehensive analysis of common and rare mitochon-
drial DNA variants in elite Japanese athletes: a case-​control study. J Hum Genet 58: 780–​787, 2013.
56. Mikami E, Fuku N, Takahashi H, Ohiwa N, Pitsiladis YP, Higuchi M, Kawahara T, and
Tanaka M. Polymorphisms in the control region of mitochondrial DNA associated with elite
Japanese athlete status. Scand J Med Sci Sports 23: 593–​599, 2013.
57. Mikami E, Fuku N, Takahashi H, Ohiwa N, Scott RA, Pitsiladis YP, Higuchi M, Kawahara
T, and Tanaka M. Mitochondrial haplogroups associated with elite Japanese athlete status. Br J Sports
Med 45: 1179–​1183, 2011.
58. Mishra P, Varuzhanyan G, Pham AH, and Chan DC. Mitochondrial dynamics is a distinguishing
feature of skeletal muscle fiber types and regulates organellar compartmentalization. Cell Metab
22: 1033–​1044, 2015.
59. Monaco C, Whitfield J, Jain SS, Spriet LL, Bonen A, and Holloway GP. Activation of
AMPKalpha2 is not required for mitochondrial FAT/​CD36 accumulation during exercise. PLoS
One 10: e0126122, 2015.
60. Morales-​Alamo D, Ponce-​Gonzalez JG, Guadalupe-​Grau A, Rodriguez-​Garcia L, Santana
A, Cusso R, Guerrero M, Dorado C, Guerra B, and Calbet JA. Critical role for free radicals on
sprint exercise-​induced CaMKII and AMPKalpha phosphorylation in human skeletal muscle. J Appl
Physiol (1985) 114: 566–​577, 2013.
61. Morrison D, Hughes J, Della Gatta PA, Mason S, Lamon S, Russell AP, and Wadley GD.
Vitamin C and E supplementation prevents some of the cellular adaptations to endurance-​training in
humans. Free Radic Biol Med 89: 852–​862, 2015.
62. Muhammad MH and Allam MM. Resveratrol and/​or exercise training counteract aging-​associated
decline of physical endurance in aged mice; targeting mitochondrial biogenesis and function. J Physiol
Sci, 2017.
63. Naviaux RK and Nguyen KV. POLG mutations associated with Alpers’ syndrome and mitochon-
drial DNA depletion. Ann Neurol 55: 706–​712, 2004.
64. Nielsen J, Gejl KD, Hey-​Mogensen M, Holmberg HC, Suetta C, Krustrup P, Elemans CPH,
and Ortenblad N. Plasticity in mitochondrial cristae density allows metabolic capacity modulation
in human skeletal muscle. J Physiol 595: 2839–​2847, 2017.
65. Nishida Y, Iyadomi M, Higaki Y, Tanaka H, Kondo Y, Otsubo H, Horita M, Hara M, and
Tanaka K. Association between the PPARGC1A polymorphism and aerobic capacity in Japanese
middle-​aged men. Intern Med 54: 359–​366, 2015.
66. Nogales-​Gadea G, Pinos T, Ruiz JR, Marzo PF, Fiuza-​Luces C, Lopez-​Gallardo E,
Ruiz-​Pesini E, Martin MA, Arenas J, Moran M, Andreu AL, and Lucia A. Are mitochon-
drial haplogroups associated with elite athletic status? A study on a Spanish cohort. Mitochondrion
11: 905–​908, 2011.
67. O’Neill HM, Holloway GP, and Steinberg GR. AMPK regulation of fatty acid metabolism and
mitochondrial biogenesis: implications for obesity. Mol Cell Endocrinol 366: 135–​151, 2013.
68. O’Neill HM, Maarbjerg SJ, Crane JD, Jeppesen J, Jorgensen SB, Schertzer JD, Shyroka O,
Kiens B, van Denderen BJ, Tarnopolsky MA, Kemp BE, Richter EA, and Steinberg GR.
AMP-​activated protein kinase (AMPK) beta1beta2 muscle null mice reveal an essential role for
AMPK in maintaining mitochondrial content and glucose uptake during exercise. Proc Natl Acad Sci
U S A 108: 16092–​16097, 2011.
69. Paulsen G, Cumming KT, Holden G, Hallen J, Ronnestad BR, Sveen O, Skaug A, Paur I,
Bastani NE, Ostgaard HN, Buer C, Midttun M, Freuchen F, Wiig H, Ulseth ET, Garthe I,
Blomhoff R, Benestad HB, and Raastad T. Vitamin C and E supplementation hampers cellular

212
213

Mitochondrial traits

adaptation to endurance training in humans: a double-​blind, randomised, controlled trial. J Physiol

592: 1887–​1901, 2014.
70. Picard M, Taivassalo T, Ritchie D, Wright KJ, Thomas MM, Romestaing C, and Hepple
RT. Mitochondrial structure and function are disrupted by standard isolation methods. PLoS One
6: e18317, 2011.
71. Picklo MJ and Thyfault JP. Vitamin E and vitamin C do not reduce insulin sensitivity but
inhibit mitochondrial protein expression in exercising obese rats. Appl Physiol Nutr Metab
40: 343–​352, 2015.
72. Poznik GD, Henn BM, Yee MC, Sliwerska E, Euskirchen GM, Lin AA, Snyder M, Quintana-​
Murci L, Kidd JM, Underhill PA, and Bustamante CD. Sequencing Y chromosomes resolves
discrepancy in time to common ancestor of males versus females. Science 341: 562–​565, 2013.
73. Puigserver P, Adelmant G, Wu Z, Fan M, Xu J, O’Malley B, and Spiegelman BM. Activation
of PPARgamma coactivator-​1 through transcription factor docking. Science 286: 1368–​1371, 1999.
74. Puigserver P, Wu Z, Park CW, Graves R, Wright M, and Spiegelman BM. A cold-​inducible
coactivator of nuclear receptors linked to adaptive thermogenesis. Cell 92: 829–​839, 1998.
75. Rankinen T, Fuku N, Wolfarth B, Wang G, Sarzynski MA, Alexeev DG, Ahmetov, II,
Boulay MR, Cieszczyk P, Eynon N, Filipenko ML, Garton FC, Generozov EV, Govorun
VM, Houweling PJ, Kawahara T, Kostryukova ES, Kulemin NA, Larin AK, Maciejewska-​
Karlowska A, Miyachi M, Muniesa CA, Murakami H, Ospanova EA, Padmanabhan
S, Pavlenko AV, Pyankova ON, Santiago C, Sawczuk M, Scott RA, Uyba VV, Yvert T,
Perusse L, Ghosh S, Rauramaa R, North KN, Lucia A, Pitsiladis Y, and Bouchard C. No
evidence of a common DNA variant profile specific to world class endurance athletes. PLoS One
11: e0147330, 2016.
76. Ristow M, Zarse K, Oberbach A, Kloting N, Birringer M, Kiehntopf M, Stumvoll M, Kahn
CR, and Bluher M. Antioxidants prevent health-​promoting effects of physical exercise in humans.
Proc Natl Acad Sci U S A 106: 8665–​8670, 2009.
77. Rito T, Richards MB, Fernandes V, Alshamali F, Cerny V, Pereira L, and Soares P. The first
modern human dispersals across Africa. PLoS One 8: e80031, 2013.
78. Roesch K, Curran SP, Tranebjaerg L, and Koehler CM. Human deafness dystonia syndrome
is caused by a defect in assembly of the DDP1/​TIMM8a-​TIMM13 complex. Hum Mol Genet
11: 477–​486, 2002.
79. Rowe GC, El-​Khoury R, Patten IS, Rustin P, and Arany Z. PGC-​1alpha is dispensable for
exercise-​induced mitochondrial biogenesis in skeletal muscle. PLoS One 7: e41817, 2012.
80. Ruas JL, White JP, Rao RR, Kleiner S, Brannan KT, Harrison BC, Greene NP, Wu J,
Estall JL, Irving BA, Lanza IR, Rasbach KA, Okutsu M, Nair KS, Yan Z, Leinwand LA, and
Spiegelman BM. A PGC-​1alpha isoform induced by resistance training regulates skeletal muscle
hypertrophy. Cell 151: 1319–​1331, 2012.
81. Ruiz-​Pesini E, Mishmar D, Brandon M, Procaccio V, and Wallace DC. Effects of purifying and
adaptive selection on regional variation in human mtDNA. Science 303: 223–​226, 2004.
82. Sahin E, Colla S, Liesa M, Moslehi J, Muller FL, Guo M, Cooper M, Kotton D, Fabian AJ,
Walkey C, Maser RS, Tonon G, Foerster F, Xiong R, Wang YA, Shukla SA, Jaskelioff M,
Martin ES, Heffernan TP, Protopopov A, Ivanova E, Mahoney JE, Kost-​Alimova M, Perry
SR, Bronson R, Liao R, Mulligan R, Shirihai OS, Chin L, and DePinho RA. Telomere dys-
function induces metabolic and mitochondrial compromise. Nature 470: 359–​365, 2011.
83. Saneto RP and Naviaux RK. Polymerase gamma disease through the ages. Dev Disabil Res Rev
16: 163–​174, 2010.
84. Sarzynski MA, Rankinen T, Sternfeld B, Grove ML, Fornage M, Jacobs DR, Jr., Sidney
S, and Bouchard C. Association of single-​nucleotide polymorphisms from 17 candidate genes
with baseline symptom-​limited exercise test duration and decrease in duration over 20 years: the
Coronary Artery Risk Development in Young Adults (CARDIA) fitness study. Circ Cardiovasc Genet
3: 531–​538, 2010.
85. Short KR, Vittone JL, Bigelow ML, Proctor DN, Rizza RA, Coenen-​Schimke JM, and Nair
KS. Impact of aerobic exercise training on age-​related changes in insulin sensitivity and muscle oxi-
dative capacity. Diabetes 52: 1888–​1896, 2003.
86. Simmons JD, Lee YL, Mulekar S, Kuck JL, Brevard SB, Gonzalez RP, Gillespie MN, and
Richards WO. Elevated levels of plasma mitochondrial DNA DAMPs are linked to clinical outcome
in severely injured human subjects. Ann Surg 258: 591–​596, 2013.

213
214

M. Tarnopolsky

87. Sommerville EW, Ng YS, Alston CL, Dallabona C, Gilberti M, He L, Knowles C, Chin SL,
Schaefer AM, Falkous G, Murdoch D, Longman C, de Visser M, Bindoff LA, Rawles JM,
Dean JCS, Petty RK, Farrugia ME, Haack TB, Prokisch H, McFarland R, Turnbull DM,
Donnini C, Taylor RW, and Gorman GS. Clinical features, molecular heterogeneity, and prog-
nostic implications in YARS2-​related mitochondrial myopathy. JAMA Neurol 74: 686–​694, 2017.
88. Srivastava S, Diaz F, Iommarini L, Aure K, Lombes A, and Moraes CT. PGC-​1alpha/​beta
induced expression partially compensates for respiratory chain defects in cells from patients with
mitochondrial disorders. Hum Mol Genet 18: 1805–​1812, 2009.
89. Steinbacher P, Feichtinger RG, Kedenko L, Kedenko I, Reinhardt S, Schonauer AL, Leitner
I, Sanger AM, Stoiber W, Kofler B, Forster H, Paulweber B, and Ring-​Dimitriou S. The
single nucleotide polymorphism Gly482Ser in the PGC-​1alpha gene impairs exercise-​induced slow-​
twitch muscle fibre transformation in humans. PLoS One 10: e0123881, 2015.
90. Szalardy L, Zadori D, Plangar I, Vecsei L, Weydt P, Ludolph AC, Klivenyi P, and Kovacs GG.
Neuropathology of partial PGC-​1alpha deficiency recapitulates features of mitochondrial encephal-
opathies but not of neurodegenerative diseases. Neurodegener Dis 12: 177–​188, 2013.
91. Taivassalo T, Abbott A, Wyrick P, and Haller RG. Venous oxygen levels during aerobic forearm
exercise: an index of impaired oxidative metabolism in mitochondrial myopathy. Ann Neurol
51: 38–​44, 2002.
92. Taivassalo T, Gardner JL, Taylor RW, Schaefer AM, Newman J, Barron MJ, Haller RG, and
Turnbull DM. Endurance training and detraining in mitochondrial myopathies due to single large-​
scale mtDNA deletions. Brain 129: 3391–​3401, 2006.
93. Taivassalo T, Jensen TD, Kennaway N, DiMauro S, Vissing J, and Haller RG. The spectrum of
exercise tolerance in mitochondrial myopathies: a study of 40 patients. Brain 126: 413–​423, 2003.
94. Taivassalo T, Shoubridge EA, Chen J, Kennaway NG, DiMauro S, Arnold DL, and Haller
RG. Aerobic conditioning in patients with mitochondrial myopathies: physiological, biochemical,
and genetic effects. Ann Neurol 50: 133–​141, 2001.
95. Tarnopolsky MA, Maguire J, Myint T, Applegarth D, and Robinson BH. Clinical, physiological,
and histological features in a kindred with the T3271C melas mutation. Muscle Nerve 21: 25–​33, 1998.
96. Tarnopolsky MA, Rennie CD, Robertshaw HA, Fedak-​Tarnopolsky SN, Devries MC, and
Hamadeh MJ. Influence of endurance exercise training and sex on intramyocellular lipid and mito-
chondrial ultrastructure, substrate use, and mitochondrial enzyme activity. Am J Physiol Regul Integr
Comp Physiol 292: R1271–​1278, 2007.
97. Tarnopolsky MA, Roy BD, and MacDonald JR. A randomized, controlled trial of creatine
monohydrate in patients with mitochondrial cytopathies. Muscle Nerve 20: 1502–​1509, 1997.
98. Tural E, Kara N, Agaoglu SA, Elbistan M, Tasmektepligil MY, and Imamoglu O. PPAR-​alpha
and PPARGC1A gene variants have strong effects on aerobic performance of Turkish elite endurance
athletes. Mol Biol Rep 41: 5799–​5804, 2014.
99. Wallace DC, Brown MD, and Lott MT. Mitochondrial DNA variation in human evolution and
disease. Gene 238: 211–​230, 1999.
100. Wong HS, Dighe PA, Mezera V, Monternier PA, and Brand MD. Production of superoxide and
hydrogen peroxide from specific mitochondrial sites under different bioenergetic conditions. J Biol
Chem 292: 16804–​16809, 2017.
101. Wyss M, Smeitink J, Wevers RA, and Wallimann T. Mitochondrial creatine kinase: a key enzyme
of aerobic energy metabolism. Biochim Biophys Acta 1102: 119–​166, 1992.
102. Yan X, Eynon N, Papadimitriou ID, Kuang J, Munson F, Tirosh O, O’Keefe L, Griffiths LR,
Ashton KJ, Byrne N, Pitsiladis YP, and Bishop DJ. The gene SMART study: method, study
design, and preliminary findings. BMC Genomics 18: 821, 2017.
103. Ydfors M, Fischer H, Mascher H, Blomstrand E, Norrbom J, and Gustafsson T. The truncated
splice variants, NT-​PGC-​1alpha and PGC-​1alpha4, increase with both endurance and resistance
exercise in human skeletal muscle. Physiol Rep 1: e00140, 2013.
104. Yfanti C, Akerstrom T, Nielsen S, Nielsen AR, Mounier R, Mortensen OH, Lykkesfeldt J,
Rose AJ, Fischer CP, and Pedersen BK. Antioxidant supplementation does not alter endurance
training adaptation. Med Sci Sports Exerc 42: 1388–​1395, 2010.
105. Yfanti C, Fischer CP, Nielsen S, Akerstrom T, Nielsen AR, Veskoukis AS, Kouretas D,
Lykkesfeldt J, Pilegaard H, and Pedersen BK. Role of vitamin C and E supplementation on IL-​6
in response to training. J Appl Physiol (1985) 112: 990–​1000, 2012.

214
215

Mitochondrial traits

106. Yfanti C, Nielsen AR, Akerstrom T, Nielsen S, Rose AJ, Richter EA, Lykkesfeldt J, Fischer
CP, and Pedersen BK. Effect of antioxidant supplementation on insulin sensitivity in response to
endurance exercise training. Am J Physiol Endocrinol Metab 300: E761–​770, 2011.
107. Yu-​Wai-​Man P, Trenell MI, Hollingsworth KG, Griffiths PG, and Chinnery PF. OPA1
mutations impair mitochondrial function in both pure and complicated dominant optic atrophy.
Brain 134: e164, 2011.
108. Zeviani M and Antozzi C. Defects of mitochondrial DNA. Brain Pathol 2: 121–​132, 1992.
109. Zhang SL, Lu WS, Yan L, Wu MC, Xu MT, Chen LH, and Cheng H. Association between per-
oxisome proliferator-​activated receptor-​gamma coactivator-​1alpha gene polymorphisms and type 2
diabetes in southern Chinese population: role of altered interaction with myocyte enhancer factor
2C. Chin Med J (Engl) 120: 1878–​1885, 2007.

215
216

17
ANGIOTENSIN-​C ONVERTING
ENZYME AND THE
GENOMICS OF ENDURANCE
PERFORMANCE
Linda S. Pescatello, Lauren M.L. Corso, Lucas P. Santos,
Jill Livingston, and Beth A. Taylor

Introduction
The preceding Chapters 15 and 16 focused on the overall genetic influence on cardiorespira-
tory fitness and mitochondrial function during physical performance, respectively. This chapter
will focus specifically on the influence of the angiotensin-​converting enzyme gene (ACE) on
human physical performance that was first studied by Montgomery and colleagues in 1998 (39).
ACE has now become the most investigated gene in relation to human physical performance
(57). Our search in PubMed located 316 citations that involved ACE and all types of physical
performance. The number of yearly citations in PubMed from 1998 to 2017 involving ACE
and all types of physical performance is displayed in Figure 17.1. Please note the search terms
for this search differed slightly from the search conducted for this chapter on ACE and endur-
ance performance, as we wished to emphasize the extent to which ACE is cited in the exercise
performance literature.
ACE (OMIM 106180) is located on the long arm of chromosome 17 band 17q23.3,
has the genomic coordinates 17:63,477,060–​ 63,498,379 (GRCh38/​ hg38:CM000679.2), is
21,319 bp long, and consists of 26 exons and 25 introns (25). The ACE gene has 22 transcripts
(splice variants), is a member of three Ensembl protein families, and is associated with five
phenotypes (http://​grch37.ensembl.org/​index.html). These phenotypes include diabetes mel-
litus, renal disease, Alzheimer’s disease, cancer and tumor burden, and athletic performance.
ACE encodes an enzyme (ACE) that is a peptidyldipeptide hydrolase belonging to the class of
zinc metallopeptidase (11, 25, 51). ACE is involved in catalyzing the conversion of angiotensin
I into a physiologically active peptide, angiotensin II. Angiotensin II is a potent vasopressor
and aldosterone-​stimulating peptide that regulates blood pressure (BP) and fluid–​electrolyte
balance and is also involved with muscle function (45, 57). ACE also inactivates bradykinin, a
potent vasodilator. ACE plays a key regulatory role in the renin–​angiotensin–​aldosterone system
(RAAS), as will be discussed later in this chapter.
Rigat et al. found that an insertion (I)/​deletion (D) of an Alu repetitive element poly-
morphism in intron 16 of the ACE gene involving about 250 bp was strongly associated with
circulating plasma ACE levels, with the mean plasma ACE level of DD genotype subjects about

216
217

Angiotensin-converting enzyme

Figure 17.1 Angiotensin-​converting enzyme gene (ACE) and physical performance number of citations
per year in PubMed.

two times that of subjects with II genotype, with ID subjects having intermediate levels between
the two homozygous genotypes (51, 52).This polymorphism is now called ACE rs4340, a struc-
tural variant (30), with the genomic coordinates ch17:63488531–​63488532 (GRCh38/​hg38).
There are several dbSNP entries all tagging ACE rs4340 that include rs4343, rs4341, rs1799752
(named as present on 23andMe chips), rs4646994, and rs13447447. The structural variant ACE
rs4340 has become the most investigated variant in ACE in relation to physical performance.
Therefore, ACE rs4340 will be the focus of this chapter in relation to endurance exercise
performance. In both common and scientific usage, endurance exercise has come to mean
structured and planned physical activity using large muscle groups that could be continued
for more than a few minutes that would be expected to maintain or improve cardiorespiratory
fitness or endurance capacity (44). See Chapter 12 of this book for a primer on endurance
exercise-​related phenotypes.
The purpose of this chapter is to present a systematic review of ACE in relation to endur-
ance exercise performance and other endurance exercise health-​related outcomes as well as
to provide a discussion on the systems through which ACE may affect endurance exercise
performance. Specially, we will: 1) overview the systematic review methodology; 2) briefly
discuss the regulation of the classical and nonclassical RAAS in relation to ACE, endurance
exercise, and other endurance exercise health-​related outcomes; 3) summarize the influence
of ACE on endurance exercise performance and other endurance exercise health-​related
outcomes and the systems by which ACE may operate; and 4) conclude with take-​home
messages from this systematic review. As first originally postulated by Montgomery and
colleagues (39), the working hypothesis of this chapter is that adults with the ACE I allele
will exhibit superior endurance exercise performance compared with adults with the ACE
D allele.

Systematic review methods

Trial selection process
A comprehensive Boolean search was run in PubMed (including Medline) and Scopus from
earliest coverage to December 28, 2017 to locate all qualifying systematic reviews, meta-​analyses,

217
218

L.S. Pescatello et al.

and primary level trials that examined the influence of ACE polymorphisms on endurance per-
formance and/​or endurance exercise health-​related outcomes. The full search strategy for ACE
and endurance performance and endurance performance health-​related outcomes for potential
qualifying trials was: (“exercise”[majr] OR exercise[ti] OR exercises[ti] OR exercising[ti] OR
postexercise[ti] OR running[MeSH] OR running[ti] OR bicycling[MeSH] OR bicycling OR
bicycle* OR cycling[ti] OR treadmill* OR ergometer OR “endurance training” OR “endur-
ance exercise” OR “endurance athlete” OR “endurance athletes” OR “speed training” OR
“circuit training” OR “training duration” OR “training frequency” OR “training intensity” OR
“aerobic endurance” OR “aerobic training” OR “interval training” OR “­combination training”
OR “combined training” OR plyometric* OR “HIIT” OR walking[MeSH] OR walking[ti]
OR swimming OR swimmer* OR hiker* OR hiking OR mountaineer* OR climber OR
climbers OR climbing OR “military training” OR walking OR ironman OR triathlon* OR
triathlete* OR marathon* OR ultramarathon* OR rowing OR rower* OR skiing OR skier*
OR “endurance sport” OR “endurance sports” OR soccer OR football OR rugby OR hockey
OR tennis OR basketball OR lacrosse OR athlete* OR “endurance performance” OR “phys-
ical endurance” OR “physical performance” OR endurance[ti]) AND (Genetics[majr] OR
genetics[ti] OR gene[ti] OR genes[ti] OR “genotype” OR “genotypes” OR chromosome[ti]
OR chromosomes[ti] OR “SNP” OR “SNPS” OR allele OR alleles OR genotype*[ti] OR
phenotype*[ti] OR genome[ti] OR polymorphism OR polymorphisms OR polymorphic OR
“trait loci” OR “gene map” OR “autosomal gene” OR “trait locus” OR “trait loci” OR “can-
didate genes” OR “candidate gene” OR “quantitative trait loci” OR “mitochondrial genome”
OR “nuclear genome” OR autosome* OR microrna[ti] OR micrornas[ti] OR mirna[ti]
OR mirnas[ti] OR loci[ti] OR “DNA” OR genetic* OR “Insertion/​deletion” OR “I/​D”
OR genetics[sh] OR “Physical Endurance/​genetics”[MAJR]) AND (ace[ti] OR “angiotensin
converting enzyme” OR “angiotensin converting enzymes” OR “angiotensin 1 converting
enzyme” OR “angiotensin 2 converting enzyme” OR “Peptidyl-​Dipeptidase A”[nm]) NOT
(“ace inhibitor”[ti] OR “ace inhibitors”[ti] OR “angiotensin converting enzyme inhibitor”[ti]
OR “angiotensin converting enzyme inhibitors”[ti] OR “Angiotensin-​Converting Enzyme
Inhibitors”[MeSH]).
At minimum, qualifying trials met the following criteria: 1) included adult samples ≥18 years
that were generally healthy and free from chronic disease; 2) performed ACE genotyping;
3) disclosed the Frequency Intensity, Time, and Type or FITT of the endurance exercise
intervention; 4) reported a measure of endurance performance or endurance exercise health-​
related outcome pre-​and post-​acute (i.e., immediate or short-​term) or chronic (i.e., long-​term
or training) endurance exercise intervention; and 5) were published in English. Trials were
excluded if they contained dietary or weight loss interventions, purposely induced dehydration,
involved substance abuse, or only reported ACE associations at baseline and not the change of
the measure of performance post-​versus pre-​exercise intervention to minimize sample selec-
tion bias (7, 8, 36). Furthermore, any study whose ACE allelic distributions were not in Hardy–​
Weinberg equilibrium was excluded.
Potential reports were screened by LMLC for title, abstract, and full text review to determine
if they qualified. In addition, reference lists of 13 qualifying narrative and systematic reviews
and meta-​analyses (8, 16, 34, 35, 42, 45, 46, 68–​72, 76) were manually searched for primary level
trials and assessed for possible inclusion. The search yielded 383 reports, of which 29 primary
level trials qualified. Details of the trial selection process appear in Figure 17.2. Of these pri-
mary level qualifying trials, 20 involved ACE and endurance performance (1, 2, 4, 9, 10, 13, 15,
18, 22, 23, 29, 60, 62–​66, 73, 74, 77), five involved ACE and endurance exercise health-​related

218
219

Angiotensin-converting enzyme

Identification
Records Located in Electronic Database
Cross-referencing
Search (PubMed N=297; Scopus N=306)
k=2
k=603

Records Screened by TITLE

(after removing duplicates) Excluded by TITLE
k=216
k=383
Screening

Records Screened by ABSTRACT Excluded by ABSTRACT

k=167 k=73

Records Screened for FULL TEXT REVIEW Excluded by FULL TEXT REVIEW
(exclusion rationale)
Eligibility

Total (k=94)
ACE and Endurance Performance
ACE and Endurance Performance (k=50)
(k=30) (NR/SR/MA k=10)
(NR/SR/MA* k=23)
ACE and Health Outcomes (k=21) Chronic Disease (k=3)
No Exercise Stimulus (k=20)
No ACE Genotyping (k=3)
Not Published in English (k=3)
Not in HWE (k=4)
Other (k=7)
ACE and Health Outcomes
(k=16)
Included

No Exercise Intervention (k=3)

Resistance Exercise Trial (k=3)
Qualifying Reports Non-adult Population (k=3)
Total Primary Level Trials (k=29) No ACE Genotyping (k=1)
ACE and Endurance Performance (k=20) Not Published in English (k=1)
ACE and Health Outcomes (k=5) Not in HWE (k=2)
ACE Endurance and Health Outcomes (k=4) Other (k=3)

Figure 17.2 Flow chart detailing systematic search of potentially qualifying primary level trials and
systematic reviews and meta-​analysis that examined ACE and endurance and/​or endurance performance
related outcomes. ACE=angiotensin-​converting enzyme; HWE=Hardy–​Weinberg equilibrium;
k=number of trials; MA=meta-​analysis; NR=narrative review; SR=systematic review. *k=13 qualifying
NR/​SR/​MAs were disaggregated and individual trials were assessed for inclusion. The final sample
reflects all qualifying primary level trials.

outcomes (6, 20, 26, 43, 54), and four trials reported both endurance performance and health-​
related outcomes (17, 28, 48, 49, 67); please note references 48 and 49 are from the same trial.
Relevant data pertaining to the sample characteristics, endurance performance intervention,
and endurance performance or endurance exercise health-​related outcomes were extracted by
LMLC for each respective trial and appear in Tables 17.1 and 17.2, displayed in their respective
sections of this chapter.

219
220

L.S. Pescatello et al.

Classical and nonclassical renin–​angiotensin–​aldosterone system in

relation to cardiovascular control and endurance exercise performance
and health-​related outcomes
The classical RAAS, as it pertains to overall cardiovascular control, is stimulated by the release of
renin in response to low BP, which catalyzes the conversion of angiotensinogen to angiotensin
I. The subsequent conversion of angiotensin I to angiotensin II by ACE then evokes a cascade of
actions to raise BP that are largely regulated by the binding of angiotensin II to the angiotensin
1 (AT1) receptor (24, 53). This pivotal activity by ACE can be increased or decreased according
to classical endocrine control (i.e., alterations in circulating levels of renin, angiotensinogen,
angiotensin I and II, and ACE), or as more recent evidence suggests, by intracrine, autocrine, and
paracrine control in which the components of the RAAS are produced by local tissues and the
brain (32, 37, 41, 50, 53, 59). The ensuing downstream pressor actions of angiotensin II include
increased sodium and water retention through aldosterone and antidiuretic hormone release;
direct vasoconstriction through smooth muscle contraction; stimulation of the sympathetic ner-
vous system to increase norepinephrine release and augment vasoconstriction; and an increase
in reactive oxygen species (primarily nicotinamide adenine dinucleotide phosphate (NAD(P)
H oxidase)) to reduce nitric oxide (NO) levels and blunt vasodilation. ACE also inhibits the
vasodilators bradykinin and kallidin (14, 53), enhancing its role as a pressor enzyme. In addition,
RAAS activation is associated with other adverse vascular effects such as cell hypertrophy and
proliferation (55).
This classical view of the RAAS has been supplemented by identification of two nonclassical
axes. The first nonclassical axis is the involvement of the angiotensin 2 (AT2) receptor found
predominantly in fetal tissue that decreases after birth such that relatively low amounts are typ-
ically expressed in adult tissue (50). The AT2 receptor is associated with antagonistic depressor
actions to those regulated by AT1 binding. These opposing depressor actions to the classical
RAAS pressor effects include an increased vasodilatory response through stimulation of endo-
thelial NO synthase to produce NO as well as augmented bradykinin release (14, 53). The
second nonclassical axis is the more recent identification of the ACE homolog, ACE2, which
binds to a mitochondrial assembly receptor and stimulates a variety of largely antagonistic
actions to the classical RAAS pressor actions (14, 31).The actions of ACE2 have not been com-
prehensively identified and characterized in human models, but appear to be associated with
various cardioprotective phenotypes with actions such as vasodilation and antiproliferation (27,
55). An overview of the classical and nonclassical axes of ACE as they relate to cardiorespiratory
control is depicted in Figure 17.3.
ACE is a rate-​limiting enzyme of the RAAS.Variations in ACE concentrations that are gen-
etically mediated by the ACE rs4340 I and D alleles (18, 51, 52) have been hypothesized to
play an important role in endurance performance as well as endurance exercise health-​related
outcomes (8, 16, 39, 40, 45, 57, 69, 73). More recent evidence characterizes the RAAS as a “dual
function system” in which the balance between vasoconstrictor and proliferative actions versus
vasodilator and antiproliferative actions are influenced by ACE/​ACE2 balance (50). For example,
individuals with greater ACE concentrations at rest may also shunt more angiotensin II to the
ACE homolog, ACE2, for conversion to angiotensin-​(1–​7), resulting in a lower-​than-​expected
impact on resting BP and sympathetically mediated vasoconstriction (31). Furthermore, the BP
response to acute and chronic endurance exercise has been shown to be influenced by the struc-
tural variant ACE rs4340 and indel ACE2 rs59272 (6, 20, 21, 26, 30, 48, 54).
The impact of the ACE rs4340 genotype on the short-​and long-​term endurance per-
formance responses is likely mediated by both chronic circulating levels of ACE, as well as

220
 newgenrtpdf
221
Brain, systemic, and local

ACE ACE2

Angiotensin I Angiotensin II Angiotensin-(1-7)

Non-ACE

Classical Nonclassical Nonclassical

AT1 receptor AT2 receptor MAS receptor

Renal effects: Renal effects:

↑ water and sodium retention ↑ water and sodium excretion
through ↑ ADH and aldosterone Vascular effects:
Vascular effects:
Vascular effects: Antagonist to angiotensin II
↑ vasoconstriction through
antagonistic ↑ vasodilation through
↑ vasodilation through ↑NO
↑ bradykinin, eNOS, and NO
↑ smooth muscle contraction, CNS effects:
↑ endothelin-1
↓ vasoconstriction through ↓SNS
↓ vasodilation through ↑ ROS activation and norepinephrine
(Most significant: ↑NAD(P)H antagonistic Other effects:
oxidase) and ↓NO, bradykinin,
↑ cardioprotection
kallidin
↑ baroreflex sensitivity
↑ hypertrophy and proliferation
↑ glucose and lipid homeostasis
CNS effects:
↑ vasoconstriction through ↑ SNS
activation and norepinephrine

Figure 17.3 Angiotensin-​converting enzyme (ACE) and cardiovascular control: classical and nonclassical axes. This figure demonstrates the enzymatic
role that ACE, by stimulating the conversion of angiotensin I to angiotensin II, initiates through varied mechanisms that influence cardiovascular
control. In the classical axis, the pressor actions result from angiotensin II binding to the AT (angiotensin) 1 receptor (or via the direct inhibitory
actions of ACE). In the first nonclassical axis, angiotensin II may bind to AT2 receptors to evoke antagonistic actions to those stimulated by AT1
binding. Alternatively, in the second nonclassical axis, ACE2 may convert angiotensin II to angiotensin-​(1–​7) to bind to the mitochondrial assembly
(MAS) receptor, evoking a series of additional antagonistic pathways to those regulated by AT1 receptor binding and lowering the concentration
of angiotensin II available for the classic axis (non-​ACE). ACE=angiotensin-​converting enzyme; ADH=antidiuretic hormone; AT1=angiotensin
type 1 receptor; AT2=angiotensin type 2 receptor; eNOS=endothelial nitric oxide synthase; MAS=mitochondrial assembly receptor; NAD(P)H
oxidase=nicotinamide adenine dinucleotide phosphate; NO=nitric oxide; ROS=reactive oxygen species; SNS=sympathetic nervous system.
222

L.S. Pescatello et al.

acute exercise-​associated increases in ACE and angiotensin II, which are paradoxically higher
in individuals with lower resting levels of ACE (65). Under chronic resting conditions, higher
concentrations of circulating ACE, as would occur with the presence of the ACE rs4340 D
allele, are associated with the production of reactive oxygen species, reduced NO-​dependent
vasodilation, systemic inflammation, sympathetic nervous system activation, and cellular prolif-
eration –​responses which are proposed to reduce functional and/​or structural cardiorespiratory
capacity (5, 37, 53). In contrast, under acute exercise conditions, higher ACE concentrations as
associated with the ACE rs4340 D allele may serve to enhance sympathetically mediated redir-
ection of muscle blood flow to active muscle, increasing metabolic capacity through improved
local vasodilation, capillary perfusion, and muscle metabolism, actions that potentially would
improve cardiorespiratory capacity (64, 66). Muscle fiber type composition (i.e., type I vs. type
II) also differs by the presence or absence of the ACE rs4340 I or D allele (75), raising the possi-
bility of a “cascade” effect, whereby the metabolic outcomes regulated by ACE rs4340 are amp-
lified by associated muscle fiber distribution distinctions tied to allelic differences. Surprisingly,
there is a paucity of studies linking ACE rs4340 and ACE concentrations to measures of acute
and chronic endurance performance so that there is an incomplete understanding of how the
complexity of the RAAS may regulate endurance performance and health-​related outcomes.
A final consideration adding to the complexity of the RAAS is that the modulating influence
of ACE on endurance performance outcomes is likely variable over time, regulated additionally
by epigenetic modification through factors such as nutrition, climate, and environment, as well
as sex and race (37, 47, 61).

The influence of ACE on endurance exercise performance

In total, there were 24 qualifying primary level trials that investigated ACE and endurance exer-
cise performance that are detailed in Table 17.1.The subjects in these trials were predominantly
young men of various races/​ethnicities and levels of cardiorespiratory fitness, ranging from being
sedentary to recreational and elite endurance athletes.The types of endurance exercise included
submaximal and maximal treadmill walking, running, and cycling, and competitive events in
extreme environments such as climbing Mt. Kilimanjaro and the 9-​day ascent to Mt. Blanc.
The submaximal and maximal measures of endurance performance were varied and included
oxygen consumption (VO2); cardiac output (CO); heart rate (HR); stroke volume (SV); venti-
latory threshold; running economy and speed; power output; capillary density, recruitment, and
muscle fiber ratio; and mitochondrial volume; among others.

ACE rs4340, plasma ACE concentration/​activity, and endurance performance

Rigat et al. found ACE rs4340 was strongly associated with circulating plasma ACE levels,
with the mean plasma ACE level of DD subjects being about two times that of II subjects, and
with ID subjects having intermediate levels between the two homozygous genotypes (51, 52).
A long-​held supposition has been that varying levels of circulating ACE as mediated by ACE
rs4340 contribute to differences in physical exercise performance, with carriers of the ACE
I allele exhibiting lower circulating ACE levels and superior endurance performance compared
to those with the D allele (40). Two qualifying trials examined the relationship between plasma
ACE concentration/​activity and endurance performance by ACE rs4340 among different sub-
ject populations and endurance exercise performance perturbations (15, 18). Both trials found a
strong relationship with ACE and circulating ACE activity levels. However, Day et al. (15) found
no relationships among ACE rs4340, circulating ACE activity levels, and acute submaximal

222
223

Angiotensin-converting enzyme

mechanical efficiency and maximal aerobic capacity (VO2max), while Domingo et al. (18) did
not report triathlon finishing time by ACE genotype nor did they examine the relationships
among ACE rs4340, plasma ACE concentrations, and measures of endurance performance. Due
to limited and disparate findings, no conclusions can be made regarding the role of systemic
ACE as genetically modulated by ACE rs4340 in endurance performance.

ACE rs4340, endurance performance, and cardiorespiratory fitness

as assessed by maximal or peak oxygen consumption
A long-​held notion has been that the ACE I allele is associated with enhanced endurance
exercise performance, largely due to greater improvements in VO2max/​peak associated with
the ACE I allele than the D allele. There were 12 qualifying trials that examined associations
between the VO2max/​peak response to an endurance exercise intervention and ACE rs4340
(Table 17.1). Of these, two reported greater increases in VO2max/​peak levels with ACE II than
ID/​DD (1, 23), two reported greater increases in VO2max/​peak levels with ACE DD than
ID/​II (49, 77), and eight trials found no association between VO2max/​peak and ACE rs4340
genotype (2, 9, 15, 17, 60, 66, 67, 73). Furthermore, three acute (i.e., short-​term or immediate)
(1, 23, 77) and one chronic (long-​term or training) (49) trial reported associations with ACE
rs4340 and changes in VO2max/​peak, while four acute (9, 15, 60, 67) and four chronic (2, 17, 66,
73) trials did not find associations with ACE rs4340 and changes in VO2max/​peak (P >0.05).
Collectively, due to disparate findings, this literature does not support the working hypothesis
of this chapter that ACE rs4340, particularly the ACE I allele, plays a major role in endurance
exercise performance as assessed by VO2max/​peak.

ACE rs4340 and other measures of endurance performance

In addition to VO2max/​peak, other endurance performance phenotypes have been linked
to ACE and endurance performance (45). Our search resulted in eight qualifying trials that
examined associations among these other measures of endurance performance and ACE rs4340.
Of these, one trial reported greater endurance performance enhancement with moderate inten-
sity (longer duration running among those with ACE II than ID and DD (10)), one trial
reported greater endurance performance enhancement with higher intensity (shorter duration
running among those with ACE DD than ID and II) (13), and six trials found no associations
with a variety of endurance performances measures that included running efficiency, power
output, minute ventilation, and the respiratory exchange ratio, among others (9, 15, 22, 49, 60,
67). Consistent with the literature examining the associations of ACE rs4340 with VO2max/​
peak, the literature on ACE and these other measures of endurance performance does not
support the working hypothesis of this chapter that ACE rs4340, particularly the ACE I allele,
plays a major role in modifying endurance exercise performance.

ACE rs4340, endurance performance, and cardiac function

The RAAS plays a central role in left ventricular remodeling (19, 58). Montgomery and
colleagues were the first to note the ACE D allele was associated with greater increases in left
ventricular mass after endurance exercise training than the I allele (38). There were five trials
that examined associations among ACE rs4340, endurance performance, and various indices of
cardiac function (2, 4, 23, 49, 60). Of these, four trials found no association among ACE and
measures of cardiac function that included left ventricular mass, CO, SV, and HR, among others

223
224

Table 17.1 A summary of the qualifying trials that examined the influence of ACE rs4340 on endurance
performance*

Author, yr Sample characteristics

Valdivieso et al., Untrained Trained

2017 (63 N=24 M (Caucasian) N=28 M (Caucasian)
BIB-​063) Age=26.3±1.1 yr Age=28.1±1.2 yr
BMI=24.0±0.7 kg/​m2 BMI=24.1±0.7 kg/​m2
%BF=22.2±6.9% %BF=14.2±1.3%
BP=124.0±2.3/​74.0±2.1 mmHg BP=125.3±2.2/​75.3±2.0 mmHg
PPO=293.4±11.4 W PPO=361.1±11.6 W
VO2peak=48.8±1.7 ml/​kg/​min VO2peak=59.5±1.7 ml/​kg/​min

Bueno et al., ACE II ACE ID ACE DD

2016 (9 N=32 M N=77 M N=41 M
BIB-​009) Age=26±3 yr Age=25±3 yr Age=24±4 yr
Weight=73.9±8.3 kg Weight=78.1±9.7 kg Weight=75.8±8.9 kg
Height=175.2±4.8 cm Height=177.2±5.9 cm Height=177.4±5.1 cm
%BF=13±3.6% %BF=14±3.8% %BF=13.1±4.3%
VO2max=50±5 ml/​kg/​min VO2max=47±3 ml/​kg/​min VO2max=49±3 ml/​kg/​min
RCPspeed=14±1 km/​hr RCPspeed=13±2 kg/​hr RCPspeed=13.5±2 km/​hr
RE10=37±6 ml/​kg/​min RE10=36±4 ml/​kg/​min RE10=36.5±3 ml/​kg/​min
RE12=44±3 ml/​kg/​min RE12=42.5±3 ml/​kg/​min RE12=43±3 ml/​kg/​min
Vaughan et al., ACE II ACE ID ACE DD
2016 (65) N=7 N=13 N=11
Age=23.0±2.9 yr Age=31.0±1.4 yr Age=25.0±1.6 yr
BMI=23.1±0.9 kg/​m2 BMI=23.6±0.9 kg/​m2 BMI=22.1±0.6 kg/​m2
VO2max=50.13.3 ml/​kg/​min VO2max=49.0±2.2 ml/​kg/​min VO2max=55.6±2.2 ml/​kg/​min
225

ACE allele Intervention characteristics Results ACE association

distribution^

NR 1) Maximal two-​legged cycling They measured ACE transcript levels, capillary-​ Yes, ACE II/​ID >DD
exercise test to-​fiber ratio, fiber cross-​sectional area, and for PPO, capillary
2) Dominant leg, one-​legged decreased muscle glycogen in the m. vastus CSA and to fiber ratio,
cycling bout Exercise started lateralis. ACE II/​ID had higher PPO than DD and muscle glycogen
with 5-​min warm up, pedaling (P=0.014). Untrained ACE II/​ID carriers had among untrained but
at 80 rpm, followed by 25 min higher capillary CSA (22.6% higher) and fiber not men
of one-​legged cycling at 30% ratio than DD (P=0.040). Postexercise muscle
of two-​leg PPO. Intensity was glycogen was decreased to greater levels
increased 10 W every min until among those ACE II/​ID than DD (P=0.010).
fatigue There were other favorable adaptations
noted among the untrained subjects for
proangiogenic factors, ACE transcript levels,
and various metabolites with ACE II/​ID
that were not found for the trained subjects
with ACE II/​ID. They concluded that
regular participation in a training program
appeared to override the genetically mediated
ACE-​dependent angiogenic and metabolic
responses to acute endurance exercise they
found in the untrained men. Therefore, some
of the discrepancies in the ACE rs4340 and
endurance performance literature may be
related to the training status of the study
participants
II (21%) 1) Maximal incremental ACE did not influence VO2max, endurance No, for VO2max,
ID (52%) treadmill test with 3 min capacity, or RE at 10 km/​hr or 12 km/​hr endurance capacity,
DD (27%) warm up (8 km/​hr), (P >0.05) and RE
followed by an increase in
speed of 1 km/​hr every min
until failure
2) Two submaximal treadmill
tests were conducted at 10
km/​hr and 12 km/​hr, using
6 min intervals
II (22.6%) One-​leg cycling bout. Subjects They examined associations with ACE Yes, ACE II >ID/​
ID (41.9%) pedaled with the dominant rs4340 and capillary volume, mitochondrial DD capillary
DD (35.4) leg at 80 rpm. Warm up was density, the capillary fiber ratio, muscle to fiber ratio,
conducted at 15% of the leg glycogen, and various energy metabolites capillary volume,
max. The exercise portion in the m. vastus lateralis. Capillary to fiber and mitochondrial
of the bout was 25 min ratio, capillary volume, and mitochondrial density among
long at 30% of leg max. The density were lower in ACE DD than ACE DD >ID/​
intensity was changed by 10 ID/​DD (~24%, P <0.05), indicating DD for glucose
W until exhaustion followed impaired glucose utilization among aerobic metabolic
by a 3 min cool down ACE DD than ID/​DD. Subjects with perturbations
ACE DD showed increased glucose and
decreased glycogen levels post exercise
compared to ID/​D (P <0.05). Maximal
RER was lower among ACE DD post
exercise than ID/​II (P=0.04). Despite
reduced glucose oxidation in the working
muscle to an acute bout of one-​legged
cycling to exhaustion with ACE DD, the
authors did not find associations with
ACE and measures of endurance exercise
performance as assessed by VO2max or
marathon completion time
(continued )
226

Table 17.1 (Cont.)

Author, yr Sample characteristics

Van Ginkel N=10

et al., 2015 Age=29.6±2.1 yr
(64) Weight=76.1±5.6 kg
Height=180.9±7.6 cm
HR=65.8±9.0 bpm
ACE II ACE ID/​DD
Pre-​maximal test Pre-​maximal test
Capillary recruitment (%)=35.5±5.5% Capillary recruitment (%)=36.2±6.3%
Capillary recruitment=18.3±2.6 Capillary recruitment=15.9±2.7
Post-​maximal test Post-​maximal test
Capillary recruitment (%)=43.3±4.6% Capillary recruitment (%)=23.8±4.1%
Capillary recruitment=26.8±2.2 Capillary recruitment=12.9±2.4
Yauet al., ACE II ACE ID ACE DD
2014 (74) N=12 M N=21 M N=12 M
Age=28±7 yr Age=27±8 yr Age=30±9 yr
BMI=23.19±2.38 kg/​m2 BMI=25.87±3.94 kg/​m2 BMI=24.31±3.54 kg/​m2
VO2peak=56.30±8.24 ml/​ VO2peak=48.44±8.5 ml/​kg/​ VO2peak=49.98±12.83 ml/​kg/​min
kg/​min min Urine Osm=459±298 mOsmol/​kg
Urine Osm=472±311 Urine Osm=502±291
mOsmol/​kg mOsmol/​kg
Verlengia Sedentary Physically active
et al., N=58 W (Caucasian) N=59 W (Caucasian)
2014 (67) Age=23.7±4.5 yr Age=22.1±4.4 yr
BMI=21.9±2.0 kg/​m2 BMI=21.5±1.8 kg/​m2
Supine HR=68.3±8.8 bpm Supine HR=64.3±9.9 bpm
BP=110.6±9.4/​71.6±7.9 mmHg BP=110.4±7.8/​74±6.8 mmHg
Glucose=72±7.0 mg/​dl Glucose=70±5.0 mg/​dl
TChol=161±29.0 mg/​dl TChol=175±26.0 mg/​dl
LDL=109±24.0 mg/​dl LDL=75±23.0 mg/​dl
HDL=41±8.0 mg/​dl HDL=53±16.0 mg/​dl
TRIG=92±30.0 mg/​dl TRIG=77±15.0 mg/​dl
ACE II ACE ID ACE DD ACE II ACE ID ACE DD
VO2=29.2± VO2=23.9± VO2=24.6± VO2=33.7± VO2=32.6± VO2=31.0±
5.9 ml/​ 3.1 ml/ 2.4 ml/ 1.6 ml/ 4.0 ml/​ 3.9 ml/
kg/​min ​kg/​min ​kg/​min ​kg/​min kg/​min ​kg/​min
Domingo Fastest triathletes Slowest triathletes
et al., N=72 M (AA) N=73 M (AA)
2013 (18) Age=33.1±6.0 yr Age=32.9±7.9 yr
BMI=23.4±1.6 kg/​m 2
BMI=24.9±2.4 kg/​m2
Total Ironman=689±37 min Total Ironman=860±47 min
Swim=64±10 min Swim=78±13 min
Cycle=365±21 min Cycle=430±28 min
Run=253±26 min Run=332±33 min

Vaughn et al., ACE II/​ID ACE DD

2013 (66) N=21 M N=14 M
Age=33.7±1.9 yr Age=29.9±2.2 yr
BMI=25.5±1.0 kg/​m2 BMI=23.0±0.4 kg/​m2
VO2max=39.4±1.8 ml/​kg/​min VO2max=47.1±2.3 ml/​kg/​min
Fiber CSA=3586±226 um2 Fiber CSA=3408±242 um2
Slow Fiber=51.3±2.4% Slow Fiber=50.8±4.9%
Capillary density=438.9±16.8/​mm2 Capillary density=540.4±26.5/​mm2
227

ACE allele Intervention characteristics Results ACE association

distribution^

NR Three separate exercise sessions Angiotensin II increased to greater levels Yes, ACE II >ID/​
on a cycle ergometer: in ACE II than ID/​DD. Subjects with DD for increased
1) Maximal test, warm up for ACE II and ACE II had higher capillary angiotensin II and
3 min at 50 W and increased recruitment and perfusion than ID/​DD had higher capillary
25 W each min until failure (P <0.05) fiber recruitment
2) Cycle at 50 W for 3 min and perfusion
at RCP determined by
maximal test
3) Cycle for 12 min at VT
determined by maximal test

II (26.7%) 1) Maximal cycle ergometer There was no relationship among ACE No, for measures of
ID (46.7%) exercise test to determine and body weight, plasma volume, serum fluid balance and
DD (26.7) VO2peak osmolality, fluid intake, total sweat loss, or hydration
2) Cycling at 55% VO2peak, perceptions of thirst (P >0.05)
for 60 min

Sedentary Graded exercise test on a There was no relationship between ACE and No, for VO2,VCO2,VE,
II (15.5%) cycle ergometer. Subjects VO2 VCO2,VE, power output, HR, SBP, power output, HR,
ID (48.3%) had a 4 min warm up at 4 and DBP to peak exercise (P >0.05) SBP, and DBP to
DD (36.2%) W. Workload was increased peak exercise
Physically Active by 20–​25 W per min until
II (25.4%) failure or 60 rpm was not
ID (45.8%) sustained
DD (28.8%)

NR Compete in South African The fastest athletes (top 50%, finishing Yes, ACE II <ID/​
Ironman Triathlon 2000–​ time=689±37 min) tended to have lower DD for plasma ACE
2001 that consists of 3.8 km ACE levels (28.8±8.6 mU/​ml), whereas concentration
open swim, 180 km cycle, the slower athletes (bottom 50%, finishing
and 42.2 km run time=860±47 min) had higher ACE levels
(31.6±8.7 mU/​ml) (P=0.055). ACE II
athletes had lower plasma ACE levels than
ACE ID or ACE DD genotype athletes
(P <0.001). These findings suggest the
ACE I allele is associated with lower ACE
levels and faster triathlon finishing time
compared to the D allele
NR 1) Cycling, 5 d/​wk, 30 min/​ They examined associations between ACE No,VO2max;
session, at 90% HRmax for and VO2max, muscle capillary growth, and Yes, ACE II/​ID >DD
6 months mitochondrial volume in the m. vastus for mitochondria
OR lateralis.VO2max not different by ACE. volume and
2) Running, 4 d/​wk, 30 min/​ Mitochondrial volume and intramyocellular intramyocellular
session at 75% VO2max for lipids increased ~3× more among ACE II/​ lipids
6 months ID than DD after training (P <0.05). ACE
I allele carriers demonstrated upregulation
of 15 transcripts related to glucose and
lipid metabolism, among other metabolic
processes positively associated aerobic
energy utilization. The authors concluded
their findings support the hypothesis that
(continued)
228

Table 17.1 (Cont.)

Author, yr Sample characteristics

Almeidaet al., ACE II ACE ID ACE DD

2012 (1) N=17 M (Latino) N=25 M (Latino) N=15 M (Latino)
Age=22.5±3.8 yr Age=23.7±3.8 yr Age=22.3±1.3 yr
Weight=70.5±6.6 kg Weight=73.2±4.5 kg Weight=71.3±8.4 kg
Height=181±4 cm Height=178±4 cm Height=177±3 cm
BMI=21.5±2.2 kg/​m2 BMI=23.1 kg/​m2 BMI=22.8±2.5 kg/​m2
V1600=258.6±5.4 m/​min V1600=249.1±4.3 m/​min V1600=211.2±8.3 m/​min
VO2max=54.2±0.9 ml/​kg/​min VO2max=52.2±0.8 ml/​kg/​min VO2max=45.8±1.8 ml/​kg/​min
Alves et al., ACE II ACE ID/​DD
2009 (2) N=18 policemen N=65 policemen
Age=26±1 yr Age=26±1 yr
Pre-​training Post-​training Pre-​training Post-​training
BMI=24±0.5 kg/​m2 BMI=23±1 kg/​m2 BMI=24± BMI=24±
VO2peak=51±2 ml/​kg/​min VO2peak=55±2 ml/​kg/​min 0.5 kg/​m2 0.5 kg/​m2
MAP=91±1 mmHg MAP=90±2 mmHg VO2peak= VO2peak=52±1
HR=74±2 bpm HR=67±2 bpm 49±1 ml/ ml/​kg/​min
LVESD=50.2±0.8 mm LVESD=31.3±0.8 mm ​kg/​min MAP=91±1
LVEDD=31.5±0.8 mm LVEDD=50.6±0.8 mm MAP=91± mmHg
LVM=143±8 g LVM=149.5±7.5 g 1 mmHg HR=70±2 bpm
LVMI=77.7±4.4 g/​m2 LVMI=81.5±4.3 g/​m2 HR=74±2 bpm LVESD=31.2
LVESD=31.5 ±0.5 mm
±0.8 mm LVEDD=51.3
LVEDD=50.2 ±0.4 mm
±0.8 mm LVM=159.3±4 g
LVM=145±8 g LVMI=83.5±
LVMI=76±2 g/​m2 2 g/​m2
Gomes-​ ACE II ACE DD
Gallego N=13 N=19
et al., Age=28±1 yr Age=26±1 yr
2009 (22) Weight=72.3±1.9 kg Weight=67.0±1 kg
Height=183±1.8 cm Height=176.6±1.2 cm
VO2max=71.7±1.3 ml/​kg/​min VO2max=73.0±0.7 ml/​kg/​min
PowerVT=4.3±0.2 W/​kg PowerVT=4.5±0.1 W/​kg
GEVT=23.9±0.0% GEVT=23.8±0.0%
Kalson et al., N=173 (104 M/​69 W)
2009 (29) Age=35.8±11.6 yr
Weight=69.7±13.5 kg

Cam et al., N=55 W (Caucasian)

2007 (10) Age=20.7±2 yr
ACE II ACE ID ACE DD
N=12 N=23 N=20
HRbase=71.8±6.2 bpm HRbase=75.5±4.1 bpm HRbase=72.6±7.8 bpm
HRpost=67.8±4.9 bpm HRpost=71.9±4 bpm HRpost=69.7±6.4 bpm
HRRbase70=161.8±2 bpm HRRbase70=161.6±2 bpm HRRbase70=161.5±3 bpm
HRRpost70=160.6±1.7 bpm HRRpost70=160.5±2 bpm HRRpost70=160.6±2.5 bpm
V30base=9.61±0.78 km/​hr V30base=9.05±0.62 km/​hr V30base=9.4±0.89 km/​hr
V30post=11.57±0.8 km/​hr V30post=10.75±0.84 km/​hr V30post=10.96±1.04 km/​hr
229

ACE allele Intervention characteristics Results ACE association

distribution^

ACE mediated remodeling of aerobic

substrate pathways contributes to the
superior endurance performance of ACE
I allele carriers. However, they did not
directly compare the relationships among
the various metabolic adaptations to
endurance exercise training that they found,
ACE, and changes in VO2max
NR 1) Maximal incremental ACE II had higher VO2max and 1600 m Yes, ACE II >DD for
treadmill test, incline at 1%, performance than DD (P <0.05) VO2max and 1600
starting speed at 6 km/​hr. m performance
Increments of 0.75 km/​hr
per min until failure
2) 1600 m running test on a
standard track. Conditions
were considered neutral
II (21.6%) Supervised treadmill running There were no associations among ACE and No, for VO2peak,
ID/​DD (78.3%) was conducted 3 d/​wk endurance performance echocardiographic LVEDV, LVEDD,
for 60 min per session for cardiac outcomes that included;VO2peak, AD, LVEF, IVST,
4 months. Intensity for LVEDV, LVEDD, AD, LVEF, IVST, LVM, LVM, LVMI
months 1–​2 was moderate, LVMI (P >0.05). Contrary to the long-​
and the last 2 months held hypothesis that individuals with ACE
intensity was above the DD would exhibit increased LV mass
RCT compared to ACE I allele carriers, there
were no significant associations found
among changes in cardiac morphology and
ACE post vs. pre training

NR Maximal cycle ergometer For all performance outcomes that included No, for PPO,VT, RCT,
exercise test starting at 25 W PPO,VT, RCT, and GE, there was no and GE
starting and increasing by 25 association with ACE (P >0.05).
W each minute until failure
at 70–​90 rpm

II (22.0%) 5 day ascent of Mt. Subjects with ACE DD tended to ascent slower No, for summit ascent
ID (52.6%) Kilamanjaro. Exercise than II/​ID (P=0.09). Assessment scores for success or failure
DD (25.4%) was gauged as successful/​ AMS tended to be higher for those with and AMS
unsuccessful summit ascent ACE DD than ACE I allele carriers at an
altitude of 2700 m (P=0.07, but not 3700 m
(P=0.76) and 4700 m (P=0.14).There was
no difference in ACE for those who were
successful vs. unsuccessful (P=0.09)
II (21.8%) 1) Maximal treadmill exercise 30 min and 70% HRR running speed Yes, ACE II >ID >DD
ID (41.8%) test. Each 2 min stage improvements post training were greater for 30 min and
DD (36.4%) increased 0.8 km/​hr among those with ACE II >ID >DD than 70% HRR running
2) Training was conducted 3 d/​ DD (P=0.05). However, at 90% HRR speed; and ACE
wk. Intensity was ~65–​80% running speed improvements were greater DD >ID >II for
HRR for a continuous bout among those with ACE DD>ID>II 90% HRR running
of 88 min at the anaerobic (P=0.0001), perhaps due to muscle fiber speed
threshold, and 36 min of type
endurance power training at
85–​90% HRR. Modality of
endurance exercise ranged
from treadmill, cross-​trainer,
and cycle ergometer
(continued)
230

Table 17.1 (Cont.)

Author, yr Sample characteristics

Dayet al., N=62

2007 (15) Age=24±7 yr
BMI=23.8±3.2 kg/​m2
VO2max=37.0±6.2 ml/​kg/​min

Cerit et al., ACE II ACE ID ACE DD

2006 (13) N=31 M (Caucasian) N=86 M (Caucasian) N=69 M (Caucasian)
2400 m Performance 2400 m Performance 2400 m Performance
Baseline=599.1±33.0 sec Baseline=591.9±33.9 sec Baseline=601.0±40.3 sec
Post=541±25.4 sec Post=532.5±28.0 sec Post=529.6±28.7 sec

Ashley et al., N=85 (62 M/​23 W)

2006 (4) Age=34±5 yr
Weight=74.6±1.2 kg
Height=177.7±0.8 cm
Body Fat=19.9±0.56%
LV=322.53±9.43 g
LVMI=168.51±5.59 g/​m2
SV=52.4±1.9 ml
CO=5.7±0.21 L/​min
TPR=1536±109 dyns/​sec/​cm5
RR interval=1.08±0.03 ms
LF/​HF=1.92±0.37
Jones et al., ACE II ACE ID ACE DD
2006 (28) N=7 N=14 N=10
Age=53±2 yr Age=58±2 yr Age=52±2 yr
BMI=33±0.8 kg/​m2 BMI=30±6 kg/​m2 BMI=33±2 kg/​m2
BP (24-​hr)=141±2/​ BP (24-​hr)=141±2/​ BP (24-​hr)=135±3/​82±3 mmHg
90±1 mmHg 83±2 mmHg VO2max=20±2 ml/​kg/​min
VO2max=21±2 ml/​kg/​min VO2max=22±2 ml/​kg/​min CC=114±11 ml/​min
CC=89±10 ml/​min CC=95±10 ml/​min 24-​Na=117±17 mEq/​day
24-​Na=123±21 mEq/​day 24-​Na=101±8 mEq/​day 24-​K=53±7 mEq/​day
24-​K=52±8 mEq/​day 24-​K=48±7 mEq/​day

Tsianos et al., N=1284 (235 M/​1049 W)

2005 (62) Age=37.0±11.5 yr

Zhao et al., N=67 M (Asian)

2003 (77) Age=23.2±0.29 yr
BMI=22.0±0.26 kg/​m2
ACE II ACE ID ACE DD
N=31 N=27 N=9
VO2max=50.48±1.58 ml/​ VO2max=50.58±1.80 ml/​kg/​ VO2max=57.86±3.5 ml/​kg/​min
kg/​min min
231

ACE allele Intervention characteristics Results ACE association

distribution^

II (23%) Submaximal test on a cycle ACE was associated with plasma ACE activity, Yes, ACE II <ID <DD
ID (40%) ergometer to test for with ACE DD (33.8+8.9 nM His-​Leu/​ml) for ACE activity;
DD (37%) DE and GE40,60, & 80 having greater levels than ID (25.9+5.5 nM No, for VO2max, DE,
W mechanical efficiency His-​Leu/​ml) followed by II (20.5+4.1 nM and GE
Maximal exercise treadmill His-​Leu/​ml) (P <0.0005). No associations
test. Starting speed was 8–​10 were found between plasma ACE and
km/​hr at 1% grade. Each VO2max, DE, and GE40,60, & 80 W
stage was 4 min and the
speed was increased by 1
km/​hr each stage
II (16.7%) 1) 2400 m middle-​distance run Post-​training 2400 m performance time Yes, ACE DD
ID (46.2%) at baseline and post-​training, improvements were greatest for ACE > ID >II for
DD (37.1%) measured in sec DD>ID>II (P=0.001), perhaps due to improvements
2) Exercise training consisted muscle fiber type in 2400 m
of 2 sessions/​d, 6 d/​wk for performance time
6 months. The program
was military-​style flexibility,
circuit training and 2400 or
3000 m run on the track
II (23%) Adrenalin Rush Adventure ACE predicted LV dysfunction after Yes, ACE II >ID >DD
ID (46%) Race. The race is run in extreme endurance exercise measured for decreases in
DD (31%) teams of 4, consists of 300 by echocardiography and impedance fractional shortening
miles of trekking, mountain cardiography. ACE II had a greater decrease time; ACE DD> Yes,
biking, kayaking, climbing in fractional shortening than ID >DD. ACE ID >DD for
fixed ropes, and swimming. ACE DD had greater LF:HF ratio than increases in LF:HF
The race can last for ID>II after the race measured by HR ratio
multiple days without rest variability. The authors concluded that
increased sympathetic system dominance
contributed to less LV decline post-​race
among athletes with ACE DD than I allele
carriers (56)
II (23%) Supervised treadmill walking Levels of sodium excreted in mEq/​day, after Yes, ACE II >ID/​
ID (45%) at 65% HRR with a 10 exercise was greater for ACE II >ID/​DD DD for sodium
DD (32%) min warm up, 30 min (II 114±22 vs. 169±39 mEq/​day, ID 100±8 excretion;
walk, 5 min rest, then an vs. 133±17 mEq/​day, DD 113±18 vs. No, for ambulatory
additional 20 min walk and 138±11 mEq/​day). ACE did not associate SBP, DBP, and MAP
10 min cool down for 7 or 8 with changes in SBP, DBP, or MAP. For
consecutive days ACE II the increase in sodium excretion
was inversely correlated with decreases in
ambulatory 24 hr DBP (r=−0.88, P=0.02)
and MAP (r=−0.95, P=0.004). They
postulated that ACE dependent effects in
sodium excretion may be due to gene to
gene interactions, termed epistasis, with
ACE and other renal variants that impact
sodium resorption in the kidney
II (19.7%) 9 day ascent of Mt. Blanc. Frequency of those who made the summit Yes, ACE II/​ID >
ID (50.0%) Exercise was gauged as with the ACE I allele was higher than DD for summit
DD (30.3%) successful/​unsuccessful those who failed to reach the summit (47% completion;
summit ascent vs. 21%, P=0.01). No relationship was No, for AMS
noted for ACE and development of AMS
II (46.3%) Graded exercise treadmill test. VO2max was higher among those with ACE Yes, ACE DD>ID/​II
ID (40.3%) 5 min warm up at 9 km/​hr DD than ID or II (P=0.04). for VO2max
DD (13.4%) and 1% incline. The speed
was increased 1 km/​min
until failure

(continued)
232

Table 17.1 (Cont.)

Author, yr Sample characteristics

Dengel et al., ACE II ACE ID ACE DD

2002 (17) N=8 3 M/​5 W (White) N=20 8 M/​12 W N=7 3 M/​4 W
Age=63±2 yr N=4 AA, N=16 white N=1 AA, N=4 white
BMI=28.5±1.8 kg/​m2 Age=62±2 yr Age=62±2 yr
%BF=37.8±1.5 BMI=28.1±0.9 kg/​m2 BMI=32.4±1.8 kg/​m2
WHR=0.88±0.03 %BF=38.4±2.3 %BF=41.7±3.2
BP=155±4/​89±4 mmHg WHR=0.87±0.03 WHR=0.85±0.02
VO2max=19.1±1.2 ml/​kg/​min BP=149±2/​88±2 mmHg BP=155±5/​88±2 mmHg
VO2max=18.1±0.9 ml/​kg/​min VO2max=16.3±1.0 ml/​kg/​min
Woods et al., N=50
2002 (73) ACE II (N=24), ACE DD (N=26)
Age=18.9±0.4 yr
Weight=73.4±1.3 kg
Height=178±10 cm

Hagberget al., ACE II ACE ID ACE DD

2002 (23) N=12 W N=33 W N=17 W
Age=62±1 yr Age=63±1 yr Age=66±1 yr
Body fat=33±2% Body fat=33±1% Body fat=30±2%
VO2max=32±1 ml/​kg/​min VO2max=29±1 ml/​kg/​min VO2max=27±1 ml/​kg/​min
HRmax=172±4 bpm HRmax=162±2 bpm HRmax=162±3 bpm
submaximal Hemodynamics Submaximal hemodynamics Submaximal hemodynamics
BP=160±4/​82±2 mmHg BP=151±2/​78±1 mmHg BP=156±3/​76±2 mmHg
HR=124±4 bpm HR=113±2 bpm HR=116±3 bpm
CO=9.1±0.4 L/​min CO=87±0.3 L/​min CO=8.3±0.4 L/​min
SV=74±3 ml SV=77±2 ml SV=72±3 ml
TPR=1030±60 dyn/​sec/​cm2 TPR=990±40 dyn/​sec/​cm5 TPR=1060±50 dyn/​sec/​cm5
Sonna et al., N=147 (62 M/​85 W)
2001 (60) Age=21.7±3.6 yr
BMI=23.1±3.1 kg/​m2
Body fat=27.9±6.1%
ACE II ACE ID ACE DD
Men Women Men Women Men Women
N=17 N=10 N=26 N=29 N=13 N=22
PreVO2peak PreVO2peak= PreVO2peak= PreVO2peak = PreVO2peak= PreVO2peak=
=50.5±1.5 39.4±1.0 ml/ 51.8±1.5 ml/​ 39.7±1.0 ml/​ 49.8±1.4 ml/ 40.2±1.3 ml/​
ml/​kg/​ ​kg/​min kg/​min kg/​min ​kg/​min kg/​min
min PostVO2peak PostVO2peak= PostVO2peak = PostVO2peak= PostVO2peak=
PostVO2peak =42.1±1.5 ml/​ 52.5±1.1 ml/​ 42.2±0.9 ml/​ 51.8±1.3 ml/​ 43.8±1.2 ml/
=52.3±1.3 kg/​min kg/​min kg/​min kg/​min ​kg/​min
ml/​kg/​
min
233

ACE allele Intervention characteristics Results ACE association

distribution^

II (23%) 24 wk of supervised treadmill There were no interactions in the training No, for VO2max, BMI,
ID (57%) walking, 3 d/​wk, at 75–​ response in VO2max, BMI, BF, fat mass, BF, fat mass, WHR,
DD (20%) 85% HRR for 40 min per WHR, SBP, and DBP by ACE (P >0.05). SBP, and DBP;Yes,
session. Adherence to the Individuals with ACE II (2.5+0.8 µU × for ACE II >ID/​
intervention was 91% 10–​4/​min/​ml) exhibited improved insulin DD improved
sensitivity than ID/​DD (0.7+0.2 µU × insulin sensitivity
10–​4/​min/​ml) as assessed with a GTT
(P=0.011)

II (23%) The endurance training No significant interaction between ACE and No,VO2max or HR;
ID (54%) program consisted of VO2max, or HR after training. However, Yes, for ACE II
DD (23%) running, marching, and the decrease in submaximal VO2 was >DD for reduced
circuit training in squads greater for ACE II than DD (P=0.02) submaximal VO2 at
for 11 wk. Maximal (40 W 80 W
increments every 3 min until
exhaustion) and submaximal
(pedaled for 3 min at 60
rpm in 20 W increments)
cycle ergometer exercise test
II (19.4%) Graded maximal and Subjects with ACE II has greater increases in Yes, ACE II > ID/​
ID (53.3%) submaximal (40%, 60%, and VO2max and submaximal (9–​11 bpm) and DD for VO2max
DD (27.4%) 80% VO2max) treadmill maximal HR (10 bpm) than ID/​DD, but and HR;
exercise test there were no overall ACE associations No for CO, SV, TPR,
with CO, SV, TPR, or the a–​VO2diff or a–​VO2diff

Men Before and after basic training, There was no relationship among ACE and No, for VO2peak, CO,
II (30.3%) peak treadmill exercise test VO2peak, CO,VE, RER, and %HRmax VE, RER, and
ID (46.4%) with a 5 min warm up at 0% and any APFT measure %HRmax
DD (23.3%) grade, speed was 2.68 m/​
Women sec (women), and 2.24 m/​
II (16.4%) sec (men) and performance
ID (47.5%) on the APFT. Speed was
DD (36.1%) increased by 0.45 m/​sec and
2% grade every 3 min until
failure

(continued)
234

Table 17.1 (Cont.)

Author, yr Sample characteristics

Rankinen Caucasian parents Caucasian AA parents (N=65) AA offspring (182)

et al., (N=177) offspring Age=48.4±7.1 yr Age=28.0±7.4 yr
2000 (49) Age=52.9±5.2 yr (294) BMI=28.6±5.2 kg/​m2 BMI=27.6±6.2 kg/​m2
BMI=27.9±4.6 kg/​m2 Age=25.4± HRmax=171.9±15.8 HRmax=185.4±12.2 bpm
HRmax=174.2± 6.2 yr bpm Wmax=156±49 W
12.9 bpm BMI=24.6±4.7 Wmax=127±36 watt VO2max=2.18±0.65 L/​min
Wmax=158±50 watt kg/​m2 VO2max=1.78±0.41
VO2max=2.15±0.62 HRmax=192.2± L/​min
L/​min 9.0 bpm
Wmax=199±
60 W
VO2max=2.64±
0.74 L/​min
235

ACE allele Intervention characteristics Results ACE association

distribution^

Caucasian The HEalth, RIsk factors, Measured 54 endurance performance Yes, for ACE DD>ID/​
(N=476) exercise Training And phenotypes in 4 groups of subjects, with II for VO2max, and
II/​ID (46.8%) GEnetics (HERITAGE) only 11 showing significant associations. submaximal VO2,
DD (53.2%) family training study Caucasian DD offspring > II for increased VE,VT, and HR;
AA (N=248) consisted of 20 wk of VO2max after training (+0.48±0.02 L/​ No, for maximal and
II/​ID (41.6%) supervised training on a min vs. +0.42±0.03 L/​min, P=0.042), submaximal CO, SV,
DD (58.4%) cycle ergometer 3 days per and submaximal VO2,VE (ACE II >ID/​ and blood lactate
wk at 55% VO2max (wk DD), and VT (ACE II >ID/​DD), and a
1–​2) and gradually increased greater decrease in submaximal HR at 50
to 75% VO2max (wk 14–​20) W. No associations for Caucasian parents
(monitored by HR) for 30 or AA parents or offspring. No associations
min per session. Pre-​and for maximal and submaximal CO, SV, and
post-​training endurance blood lactate
performance outcomes were
measured with submaximal
and maximal cycle
ergometer tests

Note: Data are presented as reported by authors. Data is reported as mean±SD unless otherwise noted.
AA=African American. ACE=angiotensin-​converting enzyme. AD=auricular diameter. AMS=acute
mountain sickness. APFT=Army Physical Fitness Test. a–​VO2diff=arterial venous oxygen difference.
Base=baseline. %BF=body fat percentage. BMI=body mass index. BP=blood pressure. bpm=beats per
minute. CC=creatinine clearance. CRF=cardiorespiratory fitness. CO=cardiac output. CSA=cross-​
sectional area. D=deletion. DE=delta efficiency. DBP=diastolic blood pressure. GE40,60 & 80 W=gross
mechanical efficiency. GTT=glucose tolerance test. HDL=high density lipoprotein. HR=heart rate.
HR0=HR at time 0. HR120=heart rate at 120 minutes after exercise. HRmax=maximum heart
rate. HRpeak=peak heart rate. HRR=heart rate reserve. I=insertion. IVST=intraventricular septum
thickness. K=potassium. Kg=kilogram. LF:HF=low and high frequency domains of heart rate variability.
LDL=low density lipoproteins. LVEDD=left ventricular end diastolic diameter. LVEF=left ventricular
ejection fraction. LVESD=left ventricular end systolic diameter. LVEDV=left ventricular end diastolic
volume. LVM=left ventricular mass. LVMI=left ventricular mass index. M=meter. MAP=mean arterial
pressure. mEq=milliequivalents. min=minute. mOsmol=milliosmols. Na=sodium. NO=nitric oxide.
Osm=osmolality. PEH=postexercise hypotension. PowerVT=power at ventilator threshold. PPO=peak
power output. PSI=physiological strain index. RCT=respiratory compensation threshold. RE=running
economy. RER=respiratory exchange rate. S=heat storage. SBP=systolic blood pressure. SD=standard
deviation. SE=standard error. Sec=second. SV=stroke volume. rpm=revolutions per minute. TChol=total
cholesterol. TG=triglycerides. TPR=total peripheral resistance.V=velocity.VE=minute ventilation.
VCO2=carbon dioxide output.VO2max=maximal oxygen consumption.VO2peak=peak oxygen
consumption.VT=ventilatory threshold. WC=waist circumference.
^ACE distribution appears when authors report the % of each genotype present in the sample and
whether the sample was in Hardy–​Weinberg equilibrium. Trials that did not provide both components
are listed as NR.
*Data are reported as mean±SD/​SE.
236

L.S. Pescatello et al.

(2, 23, 49, 60), and one trial (4) found a greater decrease in fractional shortening among athletes
with ACE II than ID and DD and an increase in sympathetic system dominance as assessed
by HR variability among athletes with ACE DD than ID and II after 120 min of continuous
extreme exercise. Collectively, these trials do not support the working hypothesis that the ACE
I allele plays a major role in endurance exercise performance as assessed by measures of cardiac
function.

ACE rs4340, endurance performance, and muscle metabolism

Angiotensin II elevates BP due to its vasoconstrictor properties, and restricts blood from
entering nonactive muscles, thereby allowing blood flow to be redirected from nonactive
to metabolically active tissues as occurs with enhanced energy turnover during exercise.
Furthermore, angiotensin II promotes angiogenesis, facilitating exercise-​induced capillary
growth in human skeletal muscle. Those with the ACE I allele have lower resting levels of
angiotensin II than carriers of the D allele. It has been hypothesized that the ACE I allele is
associated with superior endurance performance partially due to enhanced oxygen delivery
and extraction in the skeletal muscle capillary beds that may result from greater increases
in exercise-​induced serum angiotensin II levels than with the ACE D allele, effects that are
overridden at the local level during exercise (54, 64). There were four trials that examined
associations among ACE rs4340, endurance performance, and muscle metabolism (63–​66).
Of these, two trials from the same investigative team (65, 66) and two trials from different
groups (63, 64) all reported positive associations with the ACE I allele and increased capil-
lary density or volume, capillary to fiber ratio, capillary recruitment/​perfusion, and/​or mito-
chondrial density or volume compared to ACE DD, that varied by training status (63).
Furthermore,Vaughan et al. (66) found that ACE I allele carriers demonstrated upregulation
of 15 mRNA transcripts related to glucose and lipid metabolism, among other metabolic
processes positively associated with aerobic energy utilization. Yet, collectively these inves-
tigative teams did not comment on how these favorable aerobic metabolic adaptations that
appeared to be mediated by the ACE I allele related to measures of endurance exercise
performance. Therefore, no definitive conclusions can be made from this series of studies
regarding whether these favorable genetically mediated skeletal muscle metabolic and vas-
cular adaptations associated with the ACE I allele contribute to the purported superior
endurance exercise performance of ACE I allele carriers than those with the D allele.

ACE rs4340, endurance exercise performance, and fluid electrolyte balance

Adequate fluid consumption/​replacement during exercise is critical for successful performance
(3, 12). Water loss or body mass loss during or after exercise that reaches at least 3% is detri-
mental to endurance performance. Furthermore, fluid intake is tightly linked to sodium balance,
an electrolyte with a prominent role in BP regulation, nerve and muscle function, and various
other metabolic processes (e.g., glucose metabolism). The RAAS is a critical regulatory system
of fluid and electrolyte balance. There were two trials that examined the association among
ACE rs4340, endurance performance, and indicators of fluid electrolyte balance (28, 74). These
two investigative teams found conflicting findings regarding associations with biomarkers of
fluid and electrolyte balance and ACE rs4340 so that no conclusions can be made regarding
ACE, fluid and electrolyte balance, and endurance performance.

236
237

Angiotensin-converting enzyme

ACE rs4340 and endurance performance at high altitude

It has been postulated that the ACE I allele is associated with lower rates of acute mountain
sickness among successful mountaineering athletes (33, 35). Lower levels of circulating ACE
among ACE I allele carriers are thought ultimately to limit actions of the RAAS that are
implicated in the etiology of acute mountain sickness that include vasoconstriction of pul-
monary arteries via angiotensin II, dysregulation of fluids and electrolytes, and inhibition of
cardiac and pulmonary growth and remodeling. Two trials examined associations between
indicators of performance at high altitude that included successful or unsuccessful moun-
tain ascent and assessments of acute mountain sickness by ACE (29, 62). The authors reported
conflicting findings regarding successful or unsuccessful mountain ascent time, and both found
no associations with ACE and acute mountain sickness so that no conclusions can be made
regarding ACE and endurance performance at high altitude.

ACE rs4340 and endurance exercise health-​related outcomes

There were nine qualifying primary levels trials that investigated ACE rs4340 and endurance
exercise health-​related outcomes that are detailed in Table 17.2.The subjects in these trials were
predominantly middle-​aged to older men and women of various races/​ethnicities who were
sedentary at baseline and then completed either a moderate to vigorous intensity treadmill or
cycling exercise training program or acute treadmill, walking, or cycle ergometer exercise bout.
The endurance exercise health-​related outcomes were varied and included BP, measures of
body composition, and various hormone and inflammatory markers.

ACE rs4340 and the blood pressure response to endurance exercise

The RAAS is a major BP regulatory pathway due to its vasoactive actions, maintenance of fluid
and electrolyte balance, and angiogenic properties (Figure 17.3). ACE is a key enzyme in this
pathway, whose circulating levels vary by the ACE I and D alleles as discussed previously (15,
18, 51), that converts biologically inactive angiotensin I into angiotensin II, a potent vasopressor,
whose actions include vasoconstriction, renal sodium reabsorption, and aldosterone production.
For this reason, ACE is an ideal candidate gene to explore for associations with the BP response
to acute (short-​term or immediate, termed postexercise hypotension (PEH)) and chronic (long-​
term or training) endurance exercise. Five trials examined the influence of ACE on PEH (6, 20,
43, 54, 67) and four trials investigated the influence of ACE on the BP response to endurance
exercise training (17, 26, 28, 48).
Of the five trials that examined associations with ACE and PEH, four found associations
with ACE and BP, but with conflicting results (6, 20, 43, 54), and one found no association
(67). It should be noted that due to the small sample size, the results of the study by Pescatello
et al. (43) should be interpreted with caution. Nonetheless, these investigators utilized deep-​
exon sequencing of ACE and instituted other important methodological strategies to bolster
the statistical power to detect PEH–​renal genotype associations (7, 8, 36). These other strategies
included: 1) a randomized controlled repeated-​measure design with 19 hourly time points per
subject who served as their own control; 2) a focused inquiry of variants with a prioritized panel
of genes that reduced the search space within the genome; 3) high-​throughput exon sequencing
to focus on functional gene regions; 4) the same standardized protocols from their earlier studies
that used the clinical gold standard of BP assessment, ambulatory BP monitoring; 5) a closely
monitored, well-​controlled exercise exposure; and 6) adjustment for multiple testing based on

237
238

Table 17.2 A summary of the qualifying trials that examined the influence of ACE rs4340 on endurance
exercise health-​related outcomes*

Author, yr Sample characteristics

Pescatello et al., Caucasian African American/​black

2016 (43) N=9 M/​W N=14 M/​W
Age=45.1±7.8 yr Age=39.9±10.6 yr
BMI=30.5±1.8 kg/​m2 BMI=31.1±4.5 kg/​m2
Waist Circumference=98.0±7.2 cm Waist Circumference=88.3±9.4 cm
VO2peak=29.7±6.4 ml/​kg/​min VO2peak=25.3±5.7 ml/​kg/​min
Awake BP=139.3±7.0/​85.0±5.1 mmHg Awake BP=140.2±12.3/​84.3±6.9 mmHg
Glucose=96.4±12.2 mg/​dl Glucose=97.2±10.3 mg/​dl
Insulin=13.1±10.1 uIU/​ml Insulin=9.4±6.0 uIU/​ml
HOMA=3.1±2.2 HOMA=2.3±1.5
TChol=207.8±31.1 mg/​dl TChol=178.5±27.3 mg/​dl
LDL=129.4±20.3 mg/​dl LDL=53.9±14.8 mg/​dl
HDL=44.1±109 mg/​dl HDL=53.9±14.8 mg/​dl
Trig=170.8±88.8 mg/​dl Trig=83.7±35.8 mg/​dl
Nitrate/​nitrite=23.3±37.0 umol/​L Nitrate/​nitrite=10.9±13.1 umol/​L
CRP=1.1±1.0 mg/​dl CRP=2.8±3.5 mg/​dl
Endothelin=0.22±0.21 pmol/​L Endothelin=0.38±0.66 pmol/​L
PRA=1.7±1.1 ng/​ml/​hr PRA=0.95±0.84 ng/​ml/​hr
Goessler et al., ACE II/​ID ACE DD
2015 (20) N=24 8 M/​16 W N=10
Age=62.9±2.1 yr Age=58.2±3.1 yr
BMI=28.5±1.2 kg/​m2 BMI=31±0.9 kg/​m2
BP=133.8±3.0/​84.3±1.8 mmHg BP=128±5.3/​84.5±3.8 mmHg
(70.8% BP medications, 8.3% with diabetes) (70% BP medications, 10% with diabetes)

Verlengia Sedentary Physically Active

et al., 2014 (67) N=58 W (Caucasian) N=59 W (Caucasian)
Age=23.7±4.5 yr Age=22.1±4.4 yr
BMI=21.9±2.0 kg/​m2 BMI=21.5±1.8 kg/​m2
Supine HR=68.3±8.8 bpm Supine HR=64.3±9.9 bpm
BP=110.6±9.4/​71.6±7.9 mmHg BP=110.4±7.8/​74±6.8 mmHg
Glucose=72±7.0 mg/​dl Glucose=70±5.0 mg/​dl
TChol=161±29.0 mg/​dl TChol=175±26.0 mg/​dl
LDL=109±24.0 mg/​dl LDL=75±23.0 mg/​dl
HDL=41±8.0 mg/​dl HDL=53±16.0 mg/​dl
TRIG=92±30.0 mg/​dl TRIG=77±15.0 mg/​dl
ACE II ACE ID ACE DD ACE II ACE ID ACE DD
VO2=29.2±5. VO2=23.9±3.1 VO2=24.6±2. VO2=33.7±1. VO2=32.6±4.0 VO2=31.0±3.9
9 ml/​kg/​min ml/​kg/​min 4 ml/​kg/​ 6 ml/​kg/​ ml/​kg/​min ml/​kg/​min
min min
Izzicupo et al., ACE II/​ID ACE DD
2013 (26) N=21 W (postmenopausal) N=15 W (postmenopausal)
Age=56.0±4.6 yr Age=56.0±3.9 yr
BMI=25.9±4.5 kg/​m2 BMI=27.9±3.8 kg/​m2
%BF=33.2±6.7 %BF=36.8±4.8
WC=82.6±10.5 cm WC=88.7±9.2 cm
WHR=0.82±0.06 WHR=0.84±0.67
METsm/​die=1.54±0.25 ml/​kg/​min METsm/​die=1.41±0.23 ml/​kg/​min
BP=122.1±14/​78.1±9.0 mmHg BP=133.7±17.7/​81.7±7.2 mmHg
HR=64.7±8.9 bpm HR=69.6±6.3 bpm
TNF-​α=34.0±22.1 pg/​ml TNF-​α=60.0±36.2 pg/​ml
DHEA-​S=1.13±0.69 pg/​ml DHEA-​S=0.66±0.35 pg/​ml
Cortisol=123.7±54.2 pg/​ml Cortisol=122.5±58.9 pg/​ml
239

ACE allele distribution^ Intervention characteristics Results ACE association

NR Subjects completed a After vigorous intensity over 19 hr among ACE, Yes, ACE
cardiopulmonary cycle AGTR1, CYP11B2, and ADD1 variants minor<major
ergometer exercise test passing multiple testing thresholds, as the #MA alleles for 19 hr
to exhaustion (100% of increased, SBP and/​or DBP decreased ambulatory SBP
VO2peak), a 20 min 12 (P=4.5E-​05) to 30 mmHg (P=6.4E-​04) and DBP
session of cycling at 60% among African Americans only. In contrast,
VO2peak with a 5 min after moderate intensity over 19 hr among
warm up and cool down ACE and CYP11B2 variants passing multiple
to total 30 min, and a testing thresholds, as the #MA increased,
control session of seated SBP increased 21 (P=8.0E-​04) to 22 mmHg
rest. Deep-​sequenced (P=8.2E-​04) among African Americans only. Of
exons from ACE, note, although several ACE variants emerged
adducin 1 (ADD1), as significantly associated with PEH in this
angiotensin II type 1 study, ACE rs4340 did not, and our significant
receptor (AGTR1), and findings were observed only among African
the aldosterone synthase Americans and after vigorous intensity
(CYP11B2) genes

II/​ID (70.6%) An acute 45 min walking Ambulatory DBP was lowered to greater Yes, for ACE II/​
DD (29.4%) bout at 60–​75% HRR levels following the walking session ID <DD for
with a 5 min warm among individuals with ACE II/​ SBP and DBP
up and cool down to ID (−6.3+1.9 mmHg) than ACE DD
total 55 min (−2.9+1.9 mmHg) for 5 hr (P=0.002), with
similar trends for ambulatory SBP (P=0.02),
although the SBP data were not provided
Sedentary Graded exercise test on There was no relationship among ACE and VO2 No, for VO2,
II (15.5%) a cycle ergometer. VCO2,VE, power output, HR, SBP, and DBP VCO2,VE,
ID (48.3%) Subjects had a 4 at peak exercise (P >0.05) power output,
DD (36.2%) min warm up at 4 HR, SBP, and
Physically Active W. Workload was DBP at peak
II (25.4%) increased by 20–​25 W exercise
ID (45.8%) per min until failure
DD (28.8%) or 60 rpm was not
sustained

NR 13 wk 4 d/​wk for 40 After walking, SBP, MAP, and the double Yes, ACE DD
min/​d (wk 1–​4) product (i.e., HR × SBP) were lower by <II/​ID for
progressing to 50 min ACE DD <II/​ID. There were no differences SBP, MAP, and
per session (wk 5–​13) in the response of TNF-​α, DHEA-​S, or the double
of walking at a RPE cortisol by ACE (P <0.05) product
11 (wk 1–​9) and RPE
12/​13 (wk 10–​13)

(continued)
240

Table 17.2 (Cont.)

Author, yr Sample characteristics

Santana ACE II/​ID ACE DD

et al., 2011 (54) N=10/​10 N=10
Age=70.4±6.2 yr Age=70.6±5.8 yr
BMI=25.9±3.0 kg/​m2 BMI=25.4±3.2 kg/​m2
Glucose=86.9±13.1 mg/​dl Glucose=79.6±17.1 mg/​dl
BP=129.0±17.0/​77.0±8.0 mmHg BP=125.0±14.0/​77.0±7.0 mmHg
NO=294.4±176.5 uM NO=278.2±159.0 uM
Wpeak=62.3±21.4 watt Wpeak=57.0±19.7 watt
VO2peak=20.5±4.3 ml/​kg/​min VO2peak=20.1±3.0 ml/​kg/​min
HRpeak=148.0±21.0 bpm HRpeak=142.7±19.6 bpm
20% on diuretics 20% on diuretics
Jones et al., ACE II ACE ID ACE DD II (23%)
2006 (28) N=7 N=14 N=10 ID (45%)
Age=53±2 yr Age=58±2 yr Age=52±2 yr DD (32%)
BMI=33±0.8 kg/​m2 BMI=30±6 kg/​m2 BMI=33±2 kg/​m2
BP (24-​hr)=141±2/​ BP (24-​hr)=141±2/​ BP (24-​hr)=135±3/​
90±1 mmHg 83±2 mmHg 82±3 mmHg
VO2max=21±2 ml/​ VO2max=22±2 ml/​kg/​min VO2max=20±2 ml/​kg/​min
kg/​min CC=95±10 ml.min CC=114±11 ml/​min
CC=89±10 ml/​min 24-​Na=101±8 mEq/​day 24-​Na=117±17 mEq/​day
24-​Na=123±21 mEq/​day 24-​K=48±7 mEq/​day 24-​K=53±7 mEq/​day
24-​K=52±8 mEq/​day
Blanchard N=47 M (Caucasian)
et al., 2006 (6) Age=43.8±1.4 yr
BMI=29.4±0.7 kg/​m2
BP (awake)=145.0±1.5/​85.8±1.1 mmHg (hypertension)
WC=101.8±2.0 cm
VO2max=31.3±0.9 ml/​kg/​min

Dengel et al., ACE II ACE ID ACE DD

2002 (17) N=8 3 M/​5 W (White) N=20 8 M/​12 W N=7 3 M/​4 W
Age=63±2 yr N=4 AA, N=16 white N=1 AA, N=4 white
BMI=28.5±1.8 Age=62±2 yr Age=62±2 yr
kg/​m2 BMI=28.1±0.9 kg/​m2 BMI=32.4±1.8 kg/​m2
%BF=37.8±1.5 %BF=38.4±2.3 %BF=41.7±3.2
WHR=0.88±0.03 WHR=0.87±0.03 WHR=0.85±0.02
BP=155±4/​89±4 mmHg BP=149±2/​88±2 mmHg BP=155±5/​88±2 mmHg
VO2max=19.1±1.2 ml/​ VO2max=18.1±0.9 ml/​ VO2max=16.3±1.0 ml/​kg/​min
kg/​min kg/​min
241

ACE allele distribution^ Intervention characteristics Results ACE association

NR Acute maximal to After the maximal (−7.4+8.4/​ Yes, ACE II/​

exhaustion and a −4.2+6.0 mmHg) and submaximal ID>DD for
submaximal constant (2.2+5.9 mmHg/​−1.1+3.9) tests, SBP and SBP, MAP, and
load cycle ergometer MAP were reduced, respectively, and NO NO
exercise test at 90% was increased to greater levels for ACE II/​
AT and RPE for ID >DD
20 min vs. a 20
min control session
without exercise

Supervised treadmill Levels of sodium excreted in mEq/​day, after Yes, ACE II >ID/​
walking at 65% HRR exercise was greater for ACE II >ID/​ DD for sodium
with a 10 min warm DD (II 114±22 vs. 169±39 mEq/​day, ID excretion;
up, 30 min walk, 5 100±8 vs. 133±17 mEq/​day, DD 113±18 vs. No, for
min rest, then an 138±11 mEq/​day). ACE did not associate ambulatory
additional 20 min with changes in SBP, DBP, or MAP. For SBP, DBP, and
walk and 10 min ACE II the increase in sodium excretion MAP
cool down for 7 or 8 was inversely correlated with decreases in
consecutive days ambulatory 24 hr DBP (r=−0.88, P=0.02)
and MAP (r=−0.95, P=0.004)

NR 30 min acute bouts After the 40% VO2max bout, 14 hr ambulatory Yes, ACE DD
of cycle ergometer SBP was lower in ACE DD than ACE II/​ID <II/​ID for 14
exercise at (132.4 vs. 128.1 mmHg) (P=0.047). After the hr ambulatory
40% VO2max or 60% 60% VO2max bout, there were no differences SBP after 40%
VO2max with a 5 in the BP by ACE. After the 40% VO2max VO2max
min warm-​up and bout, men with 3 RAAS minor alleles
cool down to total 40 lowered SBP and DBP more than men with
min compared to a 0–​2 minor alleles
control session of rest.
Examined the epistatic
interactions of ACE,
and polymorphisms
in angiotensin II type
1 receptor (AGTR1)
and the aldosterone
synthase (CYP11B2)
genes
II (23%) 24 wk of supervised There were no interactions among the training No, for VO2max,
ID (57%) treadmill walking, 3 d/​ response of VO2max, BMI, BF, Fat Mass, BMI, BF, Fat
DD (20%) wk, at 75–​85% HRR WHR, SBP, and DBP by ACE (P >0.05). Mass, WHR,
for 40 min per session. They hypothesized ACE DD is associated SBP, and DBP;
Adherence to the with increased ACE activity, which would Yes, for ACE
intervention was 91% lead to increased production of angiotensin II >ID/​DD
II, and a decreased half-​life of bradykinin, a improved
vasodilator, that would result in decreased insulin
glucose delivery to the muscle. They sensitivity
found individuals with ACE II (2.5+0.8
µU ×10–​4/​min/​ml) exhibited improved
insulin sensitivity than ID/​DD (0.7+0.2
µU × 10–​4/​min/​ml) as assessed with a GTT
(P=0.011)
(continued)
242

Table 17.2 (Cont.)

Author, yr Sample characteristics

Rankinen ACE II ACE ID ACE DD

et al.,2000 Men Women Men Women Men Women
(48) N=54 N=59 N=101 N=117 N=69 N=66
Age=36.4± Age=35.0± Age=36.4± Age=35.0± Age=36.4± Age=35.0±0.9 yr
1.0 yr 0.9 yr 1.0 yr 0.9 yr 1.0 yr BMI=25.1±0.3
BMI=26.7± BMI=25.1± BMI=26.7± BMI=25.1± BMI=26.7± kg/​m2
0.3 kg/​m2 0.3 kg/​m2 0.3 kg/​m2 0.3 kg/​m2 0.3 kg/​m2 BP=144.5±1.9/
BP=148.4± BP=144.9± BP=145.9± BP=141.0± BP=145.2± 68.4±1.0 mmHg
2.1/​72.4± 2.0/​70.3±1.1 1.5/​71.7± 1.4/​69.3±0.8 1.9/​70.5±1.1 BP Response=
1.3 mmHg mmHg 0.9 mmHg mmHg mmHg −7.6±1.2/​−4.0±0.8
BP Response= BValdiviesP BP Response= BP Response= BP Response= mmHg
−5.7±1.3/​ Response= −5.8±0.9/​ −6.8±0.9/​− −7.1±1.1/​
−2.8±0.7 −9.7±1.3/​− −2.4±0.5 3.5±0.6 −4.4±0.6
mmHg 3.6±0.8 mmHg mmHg mmHg
mmHg
243

ACE allele distribution^ Intervention characteristics Results ACE association

Men 20 wk of supervised Men with ACE DD marginally lowered DBP Yes, for ACE
II (24.1%) training on a cycle to greater levels ~2 mmHg during the DD <II/​ID for
ID (45.1%) ergometer 3 days per submaximal test than II/​ID (2.8 (P=0.05) DBP
DD (30.8%) wk at 55% VO2max
Women (wk 1–​2) and
II (24.4%) gradually increased
ID (48.3%) to 75% VO2max (wk
DD (27.2%) 14–​20) (monitored
by HR) for 30 min
per session. BP was
measured pre-​and
post-​training with
submaximal and
maximal cycle
ergometer tests

Note: Data are presented as reported by authors. Data is reported as mean±SD unless otherwise noted.
ACE=angiotensin-​converting enzyme. %BF=body fat percentage. BMI=body mass index. BP=blood
pressure. bpm=beats per minute. CC=creatinine clearance. CRF=cardiorespiratory fitness. CRP=C-​
reactive protein. D=deletion. DBP=diastolic blood pressure. DHEA-​S=dehydroepiandrosterone-​sulfate
conjugate. HOMA=homeostatic model assessment. HR=heart rate. HRmax=maximum heart rate.
HRpeak=peak heart rate. HRR=heart rate reserve. I=insertion. METsm/​die=daily mean of metabolic
equivalent of task. min=minute. NO=nitric oxide. NR=not reported. #MA=number of minor alleles.
PEH=postexercise hypotension. PRA=plasma renin activity. SBP=systolic blood pressure. SD=standard
deviation. SE=standard error. TChol=total cholesterol. TG=triglycerides. TNF-​α=tumor necrosis factor
alpha.VE=expiratory minute ventilation.VO2max=maximal oxygen consumption.VO2peak=peak
oxygen consumption. WC=waist circumference. wk=week. WHR=waist to hip ratio. Wmax=maximum
power. Wpeak=peak power.Yr=year. ACE polymorphism distribution denotes the percentage (%) of
the sample with the following genotypes: II=homozygous for the insertion. ID=heterozygous.
DD=homozygous for the deletion.
*Data are reported as mean±SD/​SE.
^ACE distribution appears when authors report the % of each genotype present in the sample and
whether the sample was in Hardy–​Weinberg equilibrium. Trials that did not provide both components
are listed as NR.
+ACE and PEH associations only for rs1055086.
244

L.S. Pescatello et al.

genetic variants exhibiting variability in the number of minor alleles and with unique genotypic
values. Of the four trials that examined associations with ACE rs4340 and the BP response to
endurance exercise training, two found associations with ACE and BP (26, 48), and one of these
was marginal, and two found no association (17, 28). Collectively, these acute and chronic trials
tend to show BP is lowered to greater levels with ACE DD than I allele carriers. However, due
to the disparate nature of the findings, no definitive conclusions can be made regarding the BP
response to either acute or chronic endurance exercise.

ACE rs4340 and the response of other endurance exercise

health-​related outcomes
Pharmacological ACE inhibitors have been shown to improve insulin resistance, and the D
allele has been associated with alterations in exercise-​induced glucose metabolism that resemble
a prediabetic state (65). Dengal et al., in addition to BP mentioned previously, examined the
association with insulin resistance as assessed by an intravenous glucose tolerance test and ACE
rs4340 (17). They found subjects with ACE II showed greater improvements in insulin sen-
sitivity than D allele carriers (P=0.011). These findings are consistent with the more recent
findings of Vaughan et al. (65), showing the genetically mediated exercise-​induced adaptations
in the skeletal muscle by ACE lead to improved capillarization and glucose and lipid metab-
olism among I allele carriers.Yet, how these alterations would lead to ACE genetically mediated
differences in endurance performance are not clear.

Conclusion
For the many reasons discussed in this chapter, ACE rs4340 is a biologically plausible candidate
gene to explore for its associations with endurance exercise performance and health-​related
outcomes. To this end, the working hypothesis of this chapter is, adults with the ACE I allele
exhibit superior endurance exercise performance than adults with the D allele. In contrast to this
hypothesis, we conclude there is limited evidence (44) to suggest that ACE rs4340 does not play
a major role in endurance exercise performance regarding exercise-​induced changes in cardio-
respiratory fitness as assessed by VO2max/​peak and other measures of endurance performance,
cardiac morphology and function, fluid and electrolyte balance, and acute mountain sickness.
The evidence is moderate (44) indicating that ACE does play a role in the acute and chronic
exercise-​induced adaptations that occur in the skeletal muscle relating to improved capillary
perfusion and aerobic energy utilization. Yet, how these more favorable exercise-​induced skel-
etal muscle effects associated with the ACE I allele translate into superior endurance perform-
ance remains unclear due to the paucity of studies linking ACE rs4340 and ACE concentrations
to measures of acute and chronic endurance performance. Accordingly, this systematic review
does not support the working hypothesis of this chapter that adults with the ACE I allele exhibit
superior endurance exercise performance than adults with the D allele.
These evidence-​based conclusions are somewhat surprising due to the extensive volume of
literature on this topic that reflects the general level of scientific interest in ACE rs4340 and
human physical performance (Figure 17.1). Yet, this systematic review, as well as our recent
meta-​analysis (8) of candidate gene association studies examining the BP response to acute and
chronic endurance exercise, highlight the significant limitations of candidate gene association
studies. For, they often: 1) have small sample sizes; 2) examine a small number of polymorphisms,
and in the case of this systematic review, the focus was on ACE rs4340; 3) are subject to
sample selection bias; 4) are not sufficiently powered to examine possible confounding sample

244
245

Angiotensin-converting enzyme

features; and 5) do not employ statistical corrections for multiple comparisons. Other contrib-
uting factors include: 1) endurance exercise performance is a polygenic trait so that it is unlikely
that one structural variant such as ACE rs4340 would have a major effect; 2) several dbSNP
entries exist all tagging ACE rs4340 that include: rs4343, rs4341, rs1799752, rs4646994, and
rs13447447 suggesting the possibility that ACE rs4340 is in linkage disequilibrium with other
variants as well as possible epistatic interactions with other genes having stronger effects; 3) the
lack of utilization of a systems approach that integrates “omic” high-​throughput technology of
the genome, transcriptome, proteome, and/​or epigenome that would better capture genetically
mediated exercise-​induced adaptations over time; and 4) the major heterogeneity of this litera-
ture in terms of the types of endurance exercise interventions and subject populations.
When Montgomery and colleagues first began to examine the influence of ACE rs4340 on
human physical performance (39), the RAAS was thought to be a hormonal system occurring
exclusively in the circulation. The RAAS is now regarded as a complex “dual functioning
system” balancing agonist and antagonist pathways which have intracrine, autocrine, and para-
crine actions in virtually every organ in the body (32, 37, 41, 50, 53, 59). Because of the wide-​
reaching and counterregulatory effects of the RAAS, as well as the other reasons stated above,
it is not likely ACE rs4340 plays a major role in endurance exercise performance as we have
concluded in this chapter. Nonetheless, the ACE and endurance exercise performance literature
sets the stage for future exercise systems genetic studies that integrate “omic” high-​throughput
technology that will better inform the genetic basis for what constitutes superior endurance
performance.

References
1. Almeida JA, Boullosa DA, Pardono E, Lima RM, Morais PK, Denadai BS, Souza VC,
Nobrega OT, Campbell CS and Simoes HG. The influence of ACE genotype on cardiovascular
fitness of moderately active young men. Arquivos Brasileiros de Cardiologia 98: 315–​320, 2012.
2. Alves GB, Oliveira EM, Alves CR, Rached HR, Mota GF, Pereira AC, Rondon MU,
Hashimoto NY, Azevedo LF, Krieger JE and Negrao CE. Influence of angiotensinogen and
angiotensin-​converting enzyme polymorphisms on cardiac hypertrophy and improvement on max-
imal aerobic capacity caused by exercise training. European Journal of Cardiovascular Prevention and
Rehabilitation 16: 487–​492, 2009.
3. Armstrong LE, Casa DJ, Maresh CM and Ganio MS. Caffeine, fluid-​electrolyte balance,
temperature regulation, and exercise-​ heat tolerance. Exercise and Sport Sciences Reviews
35: 135–​140, 2007.
4. Ashley EA, Kardos A, Jack ES, Habenbacher W, Wheeler M, Kim YM, Froning J, Myers
J, Whyte G, Froelicher V and Douglas P. Angiotensin-​converting enzyme genotype predicts
cardiac and autonomic responses to prolonged exercise. Journal of the American College of Cardiology
48: 523–​531, 2006.
5. Azevedo ER, Mak S, Floras JS and Parker JD. Acute effects of angiotensin-​converting enzyme
inhibition versus angiotensin II receptor blockade on cardiac sympathetic activity in patients
with heart failure. American Journal of Physiology –​Regulatory, Integrative and Comparative Physiology
313: R410–​R417, 2017.
6. Blanchard BE, Tsongalis GJ, Guidry MA, LaBelle LA, Poulin M, Taylor AL, Maresh CM,
Devaney J, Thompson PD and Pescatello LS. RAAS polymorphisms alter the acute blood
pressure response to aerobic exercise among men with hypertension. European Journal of Applied
Physiology 97: 26–​33, 2006.
7. Bouchard C. Overcoming barriers to progress in exercise genomics. Exercise and Sport Sciences
Reviews 39: 212–​217, 2011.
8. Bruneau ML, Jr, Johnson BT, Huedo-​Medina TB, Larson KA, Ash GI and Pescatello LS. The
blood pressure response to acute and chronic aerobic exercise: a meta-​analysis of candidate gene asso-
ciation studies. Journal of Science and Medicine in Sport 19: 424–​431, 2016.

245
246

L.S. Pescatello et al.

9. Bueno S, Pasqua LA, de Araujo G, Eduardo Lima-​Silva A and Bertuzzi R. The association of
ACE genotypes on cardiorespiratory variables related to physical fitness in healthy men. PLoS One
11: e0165310, 2016.
10. Cam S, Colakoglu M, Colakoglu S, Sekuri C and Berdeli A. ACE I/​D gene polymorphism and
aerobic endurance development in response to training in a non-​elite female cohort. Journal of Sports
Medicine and Physical Fitness 47: 234–​238, 2007.
11. Cambien F, Alhenc-​Gelas F, Herbeth B, Andre JL, Rakotovao R, Gonzales MF, Allegrini
J and Bloch C. Familial resemblance of plasma angiotensin-​converting enzyme level: the Nancy
Study. American Journal of Human Genetics 43: 774–​780, 1988.
12. Casa DJ, Stearns RL, Lopez RM, Ganio MS, McDermott BP, Walker Yeargin S, Yamamoto
LM, Mazerolle SM, Roti MW, Armstrong LE and Maresh CM. Influence of hydration on
physiological function and performance during trail running in the heat. Journal of Athletic Training
45: 147–​156, 2010.
13. Cerit M, Colakoglu M, Erdogan M, Berdeli A and Cam FS. Relationship between ace geno-
type and short duration aerobic performance development. European Journal of Applied Physiology
98: 461–​465, 2006.
14. Chappell MC. Nonclassical renin-​angiotensin system and renal function. Comprehensive Physiology
2: 2733–​2752, 2012.
15. Day SH, Gohlke P, Dhamrait SS and Williams AG. No correlation between circulating ACE
activity and VO2max or mechanical efficiency in women. European Journal of Applied Physiology
99: 11–​18, 2007.
16. De MC and Slocombe R. The use of angiotensin-​I converting enzyme I/​D genetic polymorphism
as a biomarker of athletic performance in humans. Biosensors 2: 396–​404, 2012.
17. Dengel DR, Brown MD, Ferrell RE, Reynolds TH, 4th and Supiano MA. Exercise-​induced
changes in insulin action are associated with ACE gene polymorphisms in older adults. Physiological
Genomics 11: 73–​80, 2002.
18. Domingo R, Sturrock ED and Collins M. ACE activity and endurance performance during the
South African Ironman triathlons. International Journal of Sports Medicine 34: 402–​408, 2013.
19. Ferrario CM. Cardiac remodelling and RAS inhibition. Therapeutic Advances in Cardiovascular Disease
10: 162–​171, 2016.
20. Goessler KF, Cornelissen VA, de Oliveira EM, de F Mota G and Polito MD. ACE polymorphisms
and the acute response of blood pressure to a walk in medicated hypertensive patients. Journal of the
Renin-​Angiotensin-​Aldosterone System 16: 720–​729, 2015.
21. Goessler KF, Polito MD, Mota GF, de Oliveira EM and Cornelissen VA. Angiotensin
converting enzyme 2 polymorphisms and postexercise hypotension in hypertensive medicated indi-
viduals. Clinical Physiology and Functional Imaging 38: 206–​212, 2018.
22. Gomez-​Gallego F, Santiago C, Gonzalez-​Freire M, Muniesa CA, Fernandez Del Valle M,
Perez M, Foster C and Lucia A. Endurance performance: genes or gene combinations? International
Journal of Sports Medicine 30: 66–​72, 2009.
23. Hagberg JM, McCole SD, Brown MD, Ferrell RE, Wilund KR, Huberty A, Douglass LW
and Moore GE. ACE insertion/​deletion polymorphism and submaximal exercise hemodynamics in
postmenopausal women. Journal of Applied Physiology 92: 1083–​1088, 2002.
24. Herichova I and Szantoova K. Renin-​angiotensin system: upgrade of recent knowledge and
perspectives. Endocrine Regulations 47: 39, 2013.
25. Hubert C, Houot AM, Corvol P and Soubrier F. Structure of the angiotensin I-​converting
enzyme gene. Two alternate promoters correspond to evolutionary steps of a duplicated gene. Journal
of Biological Chemistry 266: 15377–​15383, 1991.
26. Izzicupo P, Ghinassi B, D’Amico MA, Di Blasio A, Gesi M, Napolitano G, Gallina S and Di
Baldassarre A. Effects of ACE I/​D polymorphism and aerobic training on the immune-​endocrine
network and cardiovascular parameters of postmenopausal women. Journal of Clinical Endocrinology &
Metabolism 98: 4187–​4194, 2013.
27. Jiang F, Yang J, Zhang Y, Dong M, Wang S, Zhang Q, Liu FF, Zhang K and Zhang C.
Angiotensin-​converting enzyme 2 and angiotensin 1–​7: novel therapeutic targets. Nature Reviews
Cardiology 11: 413–​426, 2014.
28. Jones JM, Park JJ, Johnson J, Vizcaino D, Hand B, Ferrell R, Weir M, Dowling T, Obisesan
T and Brown M. Renin-​angiotensin system genes and exercise training-​induced changes in sodium
excretion in African American hypertensives. Ethnicity & Disease 16: 666–​674, 2006.

246
247

Angiotensin-converting enzyme

29. Kalson NS, Thompson J, Davies AJ, Stokes S, Earl MD, Whitehead A, Tyrrell-​Marsh I, Frost
H and Montgomery H. The effect of angiotensin-​converting enzyme genotype on acute moun-
tain sickness and summit success in trekkers attempting the summit of Mt. Kilimanjaro (5,895 m).
European Journal of Applied Physiology 105: 373–​379, 2009.
30. Khurana E, Fu Y, Chakravarty D, Demichelis F, Rubin MA and Gerstein M. Role of non-​
coding sequence variants in cancer. Nature Reviews Genetics 17: 93–​108, 2016.
31. Kuczeriszka M, Kompanowska-​Jezierska E, Sadowski J, Prieto M and Navar LG. Modulating
role of Ang1–​7 in control of blood pressure and renal function in AngII-​infused hypertensive rats.
American Journal of Hypertension 31: 504–​511, 2018.
32. Leenen FH. Actions of circulating angiotensin II and aldosterone in the brain contributing to
hypertension. American Journal of Hypertension 27: 1024–​1032, 2014.
33. Luo Y, Chen Y, Zhang Y and Gao Y. The association of angiotensin-​converting enzyme gene
insertion/​deletion polymorphisms with acute mountain sickness susceptibility: a meta-​analysis. High
Altitude Medicine & Biology 13: 252–​257, 2012.
34. Ma F, Yang Y, Li X, Zhou F, Gao C, Li M and Gao L. The association of sport performance
with ACE and ACTN3 genetic polymorphisms: a systematic review and meta-​analysis. PLoS One
8: e54685, 2013.
35. MacInnis MJ, Koehle MS and Rupert JL. Evidence for a genetic basis for altitude illness: 2010
update. High Altitude Medicine & Biology 11: 349–​368, 2010.
36. Mattsson CM, Wheeler MT, Waggott D, Caleshu C and Ashley EA. Sports genetics moving
forward: lessons learned from medical research. Physioogical Genomics 48: 175–​182, 2016.
37. Mirabito Colafella KM and Danser AHJ. Recent advances in angiotensin research. Hypertension
69: 994–​999, 2017.
38. Montgomery H and Dhamrait S. ACE genotype and performance. Journal of Applied Physiology
92: 1774–​5, 2002.
39. Montgomery HE, Marshall R, Hemingway H, Myerson S, Clarkson P, Dollery C, Hayward
M, Holliman DE, Jubb M, World M, Thomas EL, Brynes AE, Saeed N, Barnard M, Bell JD,
Prasad K, Rayson M, Talmud PJ and Humphries SE. Human gene for physical performance.
Nature 393: 221–​222, 1998.
40. Montgomery H and Safari L. Genetic basis of physical fitness. Annual Review of Anthropology
36: 391–​405, 2007.
41. Nakagawa P and Sigmund CD. How is the brain renin-​angiotensin system regulated? Hypertension
70: 10–​18, 2017.
42. Pelliccia A and Thompson PD. The genetics of left ventricular remodeling in competitive athletes.
Journal of Cardiovascular Medicine 7: 267–​270, 2006.
43. Pescatello LS, Schifano ED, Ash GI, Panza GA, Lamberti L, Chen MH, Deshpande V,
Zaleski A, Farinatti P, Taylor BA and Thompson PD. Deep-​targeted exon sequencing reveals
renal polymorphisms associate with postexercise hypotension among African Americans. Physiology
Reports 4: e12992, 2016.
44. Physical Activity Guidelines Advisory Committee. 2018 Physical Activity Guidelines Advisory
Committee Scientific Report. Washington, DC: US Department of Health and Human Services, 2018.
45. Puthucheary Z, Skipworth JR, Rawal J, Loosemore M, Van Someren K and Montgomery
HE. The ACE gene and human performance: 12 years on. Sports Medicine 41: 433–​448, 2011.
46. Qi Y, Sun J, Zhu T, Wang W, Liu J, Zhou W, Qiu C and Zhao D. Association of angiotensin-​
converting enzyme gene insertion/​deletion polymorphism with high-​altitude pulmonary oedema: a
meta-​analysis. Journal of the Renin-​Angiotensin-​Aldosterone System 12: 617–​623, 2011.
47. Raleigh SM. Epigenetic regulation of the ACE gene might be more relevant to endurance physi-
ology than the I/​D polymorphism. Journal of Applied Physiology 112: 1082–​1083, 2012.
48. Rankinen T, Gagnon J, Perusse L, Chagnon YC, Rice T, Leon AS, Skinner JS, Wilmore JH,
Rao DC and Bouchard C. AGT M235T and ACE ID polymorphisms and exercise blood pressure
in the HERITAGE Family Study. American Journal of Physiology –​Heart and Circulatory Physiology
279: H368–​74, 2000.
49. Rankinen T, Perusse L, Gagnon J, Chagnon YC, Leon AS, Skinner JS, Wilmore JH, Rao DC
and Bouchard C. Angiotensin-​converting enzyme ID polymorphism and fitness phenotype in the
HERITAGE Family Study. Journal of Applied Physiology 88: 1029–​1035, 2000.

247
248

L.S. Pescatello et al.

50. Ribeiro-​Oliveira A, Jr, Nogueira AI, Pereira RM, Boas WW, Dos Santos RA and Simoes e
Silva AC. The renin-​angiotensin system and diabetes: an update. Vascular Health and Risk Management
4: 787–​803, 2008.
51. Rigat B, Hubert C, Alhenc-​Gelas F, Cambien F, Corvol P and Soubrier F. An insertion/​dele-
tion polymorphism in the angiotensin I-​converting enzyme gene accounting for half the variance of
serum enzyme levels. Journal of Clinical Investigation 86: 1343–​1346, 1990.
52. Rigat B, Hubert C, Corvol P and Soubrier F. PCR detection of the insertion/​deletion poly-
morphism of the human angiotensin converting enzyme gene (DCP1) (dipeptidyl carboxypeptidase
1). Nucleic Acids Research 20: 1433, 1992.
53. Rush JW and Aultman CD. Vascular biology of angiotensin and the impact of physical activity.
Applied Physiology, Nutrition, and Metabolism 33: 162–​172, 2008.
54. Santana HA, Moreira SR, Neto WB, Silva CB, Sales MM, Oliveira VN, Asano RY, Espindola
FS, Nobrega OT, Campbell CS and Simoes HG. The higher exercise intensity and the presence
of allele I of ACE gene elicit a higher post-​exercise blood pressure reduction and nitric oxide release
in elderly women: an experimental study. BMC Cardiovascular Disorders 11: 71, 2011.
55. Santos RA, Ferreira AJ, Verano-​Braga T and Bader M. Angiotensin-​converting enzyme 2,
angiotensin-​(1–​7) and Mas: new players of the renin-​angiotensin system. Journal of Endocrinology
216: R1–​R17, 2013.
56. Sassi R, Cerutti S, Lombardi F, Malik M, Huikuri HV, Peng CK, Schmidt G and Yamamoto
Y. Advances in heart rate variability signal analysis: joint position statement by the e-​Cardiology ESC
Working Group and the European Heart Rhythm Association co-​endorsed by the Asia Pacific Heart
Rhythm Society. Europace 17: 1341–​1353, 2015.
57. Šećerović A, Gurbeta L, Omanović-​Mikličanin E and Badnjević A. Genotype association
with sport activity: the impact of ACE and ACTN3 gene polymorphism on athletic performance.
International Journal of Engineering Research and Technology 6: 859–​863, 2017.
58. Silva SD, Jr, Zampieri TT, Ruggeri A, Ceroni A, Aragao DS, Fernandes FB, Casarini DE and
Michelini LC. Downregulation of the vascular renin-​angiotensin system by aerobic training –​focus
on the balance between vasoconstrictor and vasodilator axes. Circulation Journal 79: 1372–​1380, 2015.
59. Siragy HM and Carey RM. Protective role of the angiotensin AT2 receptor in a renal wrap hyper-
tension model. Hypertension 33: 1237–​1242, 1999.
60. Sonna LA, Sharp MA, Knapik JJ, Cullivan M, Angel KC, Patton JF and Lilly CM. Angiotensin-​
converting enzyme genotype and physical performance during US Army basic training. Journal of
Applied Physiology 91: 1355–​1363, 2001.
61. Touyz RM and Montezano AC. Angiotensin-​(1–​7) and vascular function: the clinical context.
Hypertension 71: 68–​69, 2018.
62. Tsianos G, Eleftheriou KI, Hawe E, Woolrich L, Watt M, Watt I, Peacock A, Montgomery
H and Grant S. Performance at altitude and angiotensin I-​converting enzyme genotype. European
Journal of Applied Physiology 93: 630–​633, 2005.
63. Valdivieso P, Vaughan D, Laczko E, Brogioli M, Waldron S, Rittweger J and Fluck M. The
metabolic response of skeletal muscle to endurance exercise is modified by the ACE-​I/​D gene poly-
morphism and training state. Frontiers of Physiology 8: 993, 2017.
64. van Ginkel S, Ruoss S, Valdivieso P, Degens H, Waldron S, de Haan A and Fluck M. ACE
inhibition modifies exercise-​induced pro-​angiogenic and mitochondrial gene transcript expression.
Scand J Med Sci Sports 26: 1180–​1187, 2016.
65. Vaughan D, Brogioli M, Maier T, White A, Waldron S, Rittweger J, Toigo M, Wettstein J,
Laczko E and Fluck M. The angiotensin converting enzyme insertion/​deletion polymorphism
modifies exercise-​induced muscle metabolism. PLoS One 11: e0149046, 2016.
66. Vaughan D, Huber-​Abel FA, Graber F, Hoppeler H and Fluck M. The angiotensin converting
enzyme insertion/​deletion polymorphism alters the response of muscle energy supply lines to exer-
cise. European Journal of Applied Physiology 113: 1719–​1729, 2013.
67. Verlengia R, Rebelo AC, Crisp AH, Kunz VC, Dos Santos Carneiro Cordeiro MA, Hirata
MH, Crespo Hirata RD and Silva E. Lack of association between ACE indel polymorphism
and cardiorespiratory fitness in physically active and sedentary young women. Asian Journal of Sports
Medicine 5: e22768, 2014.
68. Wang QQ, Yu L, Huang GR, Zhang L, Liu YQ, Wang TW, Lin H, Ren Q, Liu P, Huang L,
Qin J, Wu GM, Li QN, Li YF and Xiong HY. Polymorphisms of angiotensin converting enzyme

248
249

Angiotensin-converting enzyme

and nitric oxide synthase 3 genes as risk factors of high-​altitude pulmonary edema: a case-​control
study and meta-​analysis. Tohoku Journal of Experimental Medicine 229: 255–​266, 2013.
69. Woods D. Angiotensin-​converting enzyme, renin-​angiotensin system and human performance.
Medicine and Sport Science 54: 72–​87, 2009.
70. Woods DR, Brull D and Montgomery HE. Endurance and the ACE I/​D polymorphism. Science
Progress 83: 317–​336, 2000.
71. Woods DR, Humphries SE and Montgomery HE. The ACE I/​D polymorphism and human
physical performance. Trends in Endocrinology and Metabolism 11: 416–​420, 2000.
72. Woods DR and Montgomery HE. Angiotensin-​converting enzyme and genetics at high altitude.
High Altitude Medicine & Biology 2: 201–​210, 2001.
73. Woods DR, Pollard AJ, Collier DJ, Jamshidi Y, Vassiliou V, Hawe E, Humphries SE and
Montgomery HE. Insertion/​deletion polymorphism of the angiotensin I-​converting enzyme gene
and arterial oxygen saturation at high altitude. American Journal of Respiratory and Critical Care Medicine
166: 362–​366, 2002.
74. Yau AM, Moss AD, James LJ, Gilmore W, Ashworth JJ and Evans GH. The influence of
angiotensin converting enzyme and bradykinin receptor B2 gene variants on voluntary fluid intake
and fluid balance in healthy men during moderate-​intensity exercise in the heat. Applied Physiology,
Nutrition, and Metabolism 40: 184–​190, 2015.
75. Zhang B, Tanaka H, Shono N, Miura S, Kiyonaga A, Shindo M and Saku K. The I allele of the
angiotensin-​converting enzyme gene is associated with an increased percentage of slow-​twitch type
I fibers in human skeletal muscle. Clinical Genetics 63: 139–​144, 2003.
76. Zhang T, Zhang CF, Jin F and Wang L. Association between genetic factor and physical perform-
ance. Yi Chuan 26: 219–​226, 2004.
77. Zhao B, Moochhala SM, Tham S, Lu J, Chia M, Byrne C, Hu Q and Lee LK. Relationship
between angiotensin-​converting enzyme ID polymorphism and VO(2max) of Chinese males. Life
Sciences 73: 2625–​2630, 2003.

249
250
251

SECTION 4

Systems genetics of muscle mass,

strength, and trainability

Section 3 explored the systems genetics underlying endurance and endurance-​related traits.
Now we shift gears away from endurance and towards muscle-​related traits such as muscle size,
strength, and power. These traits not only contribute to performance in a myriad of sports, but
also heavily influence quality of life and health outside of sport. Again, each author was tasked
particularly to review traits at baseline (untrained state) and following acute and chronic exer-
cise training, to highlight gaps in our knowledge and to provide a foundation for future research
related to sport and exercise genetics research.
Chapter 18 kicks off Section 4 with a discussion of heritability of muscle strength and size from
Dr. Martine Thomis. Readers will quickly realize that “strength” is not a single trait –​and the com-
plexity of measuring and interpreting muscle functional data will become clear. Dr. Monica Hubal
and colleagues provide a primer for muscle strength in Chapter 19, and start to discuss molecular
and genetic drivers of strength variability. It is somewhat impossible to discuss strength without
considering muscle size, which is a key component of function. In Chapter 20, Dr. Philip Atherton
and colleagues continue the specific discussion of muscle size, again highlighting variation in size
at baseline and discussing changes in size following resistance training. It is important to notice
similarities and differences between acute bouts of resistance training and chronic exercise training.
Both acute and chronic training affect muscle in large part via modulation of neuroendocrine
factors such as various growth hormones. Chapter 21 reviews these neuroendocrine modulators
and is expertly crafted by Dr. William Kraemer (a pioneer in this field) and colleagues.
As noted in Section 3, a few genetic variations have been studied enough to warrant
their own chapters. In Section 4, we explore two such variations: genetic variations related
to the potent muscle regulator protein myostatin (MSTN; reviewed by Dr. Dustin Hittel in
Chapter 22) and a common variation in the fast-​twitch muscle gene alpha-​actinin-​3 (ACTN3),
reviewed by Dr. Peter Houweling and colleagues in Chapter 23. Each of these mutations is
quite interesting to the general public –​for example, natural mutations in myostatin have been
noted in nature as grossly hypertrophied animal breeds such as the bully whippet or the Belgian
Blue cow and myostatin has been under intense study by pharmaceutical companies to try to
develop therapies to combat muscle wasting.
Together, the chapters in Section 4 provide a comprehensive review of traits underlying strength
and size-​related traits both at baseline and following training, highlight gaps in our current know-
ledge, and provide a solid foundation for future work to inform systems models.

251
252
253

18
HERITABILITY OF MUSCLE
SIZE AND STRENGTH TRAITS
Martine Thomis

Muscle mass, strength, and power contribute to athletic performance of extraordinary levels.
In the general aging population, high levels of muscular fitness prolong the ability to per-
form activities of daily living (28) and are related to high levels of physical and psychological
well-​being (39). Low or weak muscular fitness is associated with higher prevalence of all-​cause
mortality (29), type 2 diabetes, cardiovascular disease, and osteoarthritis (51). Recent data show
grip strength as a stronger predictor of cardiovascular and all-​cause mortality than systolic blood
pressure (29). Dynapenia (age-​associated loss of muscle strength) and sarcopenia (age-​associated
loss of muscle mass) (36) have an increasing impact on healthcare costs in aging populations
(23). On the other hand, gene hunting for “elite power performance” shows high interest in
identifying genetic endowments related to strength and or power excellence in sports (43) (see
Chapters 19–​23).
Skeletal muscle mass, strength, and power are multifactorial traits influenced by a wide range
of genetic and environmental factors. This chapter aims to describe studies that have estimated
the heritability of skeletal muscle mass and skeletal muscle strength and power, both at baseline
and in response to exercise. The following chapters will focus on explaining specific biological
pathways that play a role in muscle mass and strength and their responses to progressive resist-
ance training (Chapters 19–​21).

Heritability of muscle mass

The increase in muscle mass with age due to fiber hypertrophy is fairly linear from young
childhood until puberty, with small but consistent advantages in boys. The sex difference
enlarges during and after puberty, driven primarily by the differences in sex steroids (5). Muscle
mass remains rather stable during the third and fourth decade of life; nonetheless, after about
the age of 50, muscle mass declines at a rate of approximately 1%/​year in men and 0.5%/​year
in women (36). In elite athletes, muscle mass and muscle fiber type composition is related to
the specific demands of the sport (63). Skeletal muscle mass is mainly indirectly measured or
estimated, by using whole-​body lean tissue (dual X-​ray absorptiometry (DXA) or magnetic res-
onance imaging (MRI)) scans to determine whole-​body muscle mass (11), or regional measures
using computed tomography (CT)/​ MRI, ultrasonography, or anthropometric measures

253
254

M. Thomis

(skinfold-​corrected circumferences) (17). Muscle fiber type distribution (types I, IIa, and IIx)
is, among other qualitative factors, related to specific strength, power, and muscular endurance
performances and shows large variation in elite performers (63) and in the general population.
Muscle biopsies from vastus lateralis muscle are most widely studied, although noninvasive
techniques have also emerged (4).

Heritability of regional muscle mass based on circumferences –​

anthropometric measures
Loos et al. (31) reported heritability estimates in the range of 87–​95% for circumferences at the
upper-​and forearm, and thigh and calf of 10–​14-​year-​old twins in the Leuven Longitudinal
Twin Study. Different multivariate models showed age-​and sex-​dependent genetic covari-
ation in an “arm–​leg” or “proximal–​distal” pattern. Other twin studies indicate strong evidence
for a genetic component of circumference measurements, with heritability estimates ranging
between 0.53 and 0.75 (6, 10, 13). In a sample of 748 young, Belgian, male siblings from 335
families, upper-​limit heritabilities (including genetic and shared environmental variances) were
estimated at 89% and 91% for anthropometrically estimated muscle plus bone cross-​sectional
area of the arm and thigh, respectively (21).

Heritability of regional and total muscle mass based

on imaging techniques
Significant sibling correlations for muscle diameter measures from radiographic analyses in pre-
school children (rmales=0.56 and rfemales=0.63) and monozygotic twin (MZ) intraclass correlations
(rmales=0.83 and rfemales=0.85) in adolescents indicate an important genetic contribution in calf
muscle mass (16, 19). In 41 young adult, male twins, the contribution of additive genetic factors
on muscle cross-​sectional area of the mid-​upper arm determined by CT scans was estimated at
92% (95% confidence interval (CI) 79–​97%) (56). In 227 pairs of MZ and 126 pairs of dizygotic
(DZ) postmenopausal twins, total body lean mass was measured by a Hologic QDR–​2000 DXA
scanner (Hologic Inc., Waltham, MA, USA). Heritability was estimated using the Falconer for-
mula and showed that 56% (95% CI 24–​88%) of variation in age, height, and weight-​corrected
whole-​body lean mass was attributable to genetic factors (1). Appendicular lean mass (ALM;
sum of lean body mass measured at the four limbs) was determined by DXA (QDR 4500W
system; Hologic, Inc) in a large sample of adult (51.8 ± 13.7 years) women within the TwinsUK
Adult Twin Registry (1196 individuals: 119 DZ and 428 MZ twin pairs and 102 singletons,
56 MZ and 46 DZ twins without sibling measurements). The heritability estimate obtained
using variance decomposition analysis was 0.81 ± 0.05 (30). Family members (men, N=580;
women, N=766, 327 pedigrees) in the Framingham Osteoporosis Study had leg lean mass
(LLM) measured by DXA. Age-​adjusted LLM (and estrogen status in women) was highly herit-
able (upper-​limit heritability), with a significant h2 estimate of 69%. After adjustment for height
and body mass index, h2 decreased to 42% (24). Similar family-​based heritability estimates were
reported within 244 families ascertained for two adult siblings having type 2 diabetes in the
Diabetes Heart Study (20). Estimations of the upper-​limit heritability were 0.63 and 0.67 for
whole-​body lean mass and ALM, respectively, with phenotypes controlled for covariates age,
sex, ethnicity, height, diabetes status, smoking, dietary intake, and physical activity.

254
255

Heritability of muscle size

Heritability of muscle fiber type distribution

Early reports presented high heritability estimates (h2=0.93) for the proportion of type I fibers in
skeletal muscle (26). Komi et al. found that MZ twins (N=15) were almost identical in the per-
centage of type I fibers in the vastus lateralis muscle, while DZ twins (N=16) were quite variable.
Also, high MZ twin intrapair correlations for type I fiber composition (r=0.94) (35) are indicative of
a high genetic contribution in young adult males. However, Lortie et al. (32) showed less evidence for
high genetic contributions in a study on 32 pairs of brothers and 35 pairs of MZ twins. Significant
F-​ratios in brothers and MZ twins indicated a significant biological resemblance for type I fiber per-
centage. The contribution of nongenetic factors was estimated at 30% of the variance in MZ fiber
type I distribution, while another 12% of the variance was accounted for by sampling variability and
technical error.There was no significant resemblance in brothers for any of the fiber type areas, while
MZ twins only exhibited a significant resemblance in type I fiber area.With the inclusion of an add-
itional set of 32 male and 20 female DZ twins to these data, estimates of heritability were based on
several estimation formulas and indicated no significant genetic effect for the distribution of fiber
type I, IIa, and IIx fibers or fiber areas. In a review on this topic, variation in type I fiber distribution
was decomposed in three sources: variance due to sampling variability and technical errors (15%),
environmental variance (40%), and genetic variance (40–​50%) (49). In 78 sedentary subjects from
19 families of the HERITAGE Family Study, results suggested a weak familial aggregation (F-​test
for comparison of within and between family variance) for type I fiber areas in the sedentary state
(F=2.4, P=0.007). There was some evidence for familial resemblance in the number of capillaries
around type I (F=1.9, P=0.039) and type IIa fibers (F=1.9, P=0.044), and in the fiber area per capil-
lary in type I (P=0.011) and type IIa fibers (P=0.042) in the sedentary state (45).
Baguet et al. (3) investigated carnosine as a stable muscle metabolite measured by proton
magnetic resonance spectroscopy in a twin cohort (N=25 MZ, N=22 DZ). The content of this
metabolite is closely related to the muscle fiber type distribution in human gastrocnemius and
soleus muscle (4). In soleus muscle, intratwin similarities were higher for MZ twins compared
to DZ twins, suggesting that muscle carnosine levels and fiber type distribution are highly herit-
able (AE model: h2=85%). However, these observations could not be observed in gastrocnemius
muscle. Although part of the large interindividual variability in proportions of different fiber
types in human mixed muscles is genetically determined, its specific contribution and additional
role of epigenetic factors (2) still need to be resolved.

Heritability of muscle strength

Muscle strength-​related phenotypes are used as an overall term to indicate the different subtypes
and evaluation methods to study aspects of contractile outcomes of activated muscle (see
Chapter 19 for further details). In general, muscle strength is subdivided into static or iso-
metric strength (force generated at a specific joint angle without lengthening or shortening
of the muscle fibers), dynamic concentric contractions (shortening of the muscle fibers at a
specific contraction velocity), dynamic eccentric contractions (fibers are lengthened against
an external force), muscle power (maximal dynamic strength delivered in the shortest possible
time), and muscular endurance (e.g., repeated muscle contractions). Evaluation of each of these
characteristics can be based on a wide variety of test protocols.
Recently, two meta-​ analyses have been performed on heritability estimates of muscle
strength-​related phenotypes. Schutte et al. (47) reported the weighted mean heritability of (iso-
metric) handgrip strength in 9–​25-​year-​olds (N=4516) at 63% (95% CI 47–​73%, five studies)

255
256

M. Thomis

and explosive strength (vertical jump; N=874) at 62% (95% CI 47–​77%, four studies) in adoles-
cent twin samples. Zempo et al. (60) included 24 studies (with 58 measurements) on heritability
of human muscle strength-​related phenotypes in their meta-​analysis based on a systematic lit-
erature search before August 2016.

Heritability of handgrip strength

Handgrip strength is a widely used measure of isometric strength applied over a large age range.
Ten studies based on twin or family data reported (upper-​limit) heritability estimates for grip
strength (N=320,588 in total, average ages between 9 and 70.5 years) between 0.36 and 0.88,
with a weighted mean heritability of 0.56 (95% CI 0.46–​0.67) (1, 14, 22, 25, 27, 34, 37, 44, 47,
48, 57). A significant degree of heterogeneity in these studies was further explored and showed
a significant negative correlation of grip-​strength heritability with age (r=−0.70) and sex (males
>females, P <0.05).

Heritability of other isometric strength tests

Other isometric or maximal static strength tests include elbow flexion/​extension, knee exten-
sion/​flexion, or trunk flexion/​extension tests using standardized dynamometers, ankle plantar
flexion, or arm pull tests. Zempo et al. (60) included 16 measurements from 11 studies for
isometric strength measures (21, 27, 33, 38, 42, 48, 56, 57, 59, 62). The weighted h2 was 0.49
(95% CI 0.47–​0.52). Although there was a considerable degree of heterogeneity, there was no
significant relationship between age and reported h2 or sex. Part of the variability in heritability
estimates might come from the force–​length relationship that determines the optimal strength
performance over a range of motion. Thomis et al. (56) found a similar, slightly “mirrored”
pattern of h2 estimates for elbow flexor strength at different elbow angles compared to the
observed mean torque values. Different contributions of unique environmental variation might
correspond to the degree of variability in muscle activation and performance discomfort at each
angle. Also, genetic factors related to the moment arm (tendon insertions and bone structures)
might contribute differently at different muscle lengths. However, muscle activation as measured
by the maximal value of the filtered differences of the transformed electromyograms during
maximal isometric flexion at 140° and 110° elbow flexion was highly determined by genetic
factors (h2=0.78–​0.83) (56).

Heritability of isokinetic torque

Isokinetic tests aim to measure the ability of contracting muscle to generate force at a preset
speed of contraction –​though only over a limited range of motion during shorting (concentric)
or lengthening (eccentric) of the muscle. A subset of three studies (12, 21, 46) measuring mainly
elbow isokinetic flexion were included in the overview by Zempo et al. (60) and showed a
weighted mean h2 of 0.49 (95% CI 0.37–​0.61). Dynamic strength depends on the speed of con-
traction (lower strengths/​torques at higher speeds) and type of contraction (eccentric torques
exceeding concentric torques). This force–​velocity and contraction type specificity was also
partially observed when studying elbow flexor strength at different velocities in both concentric
and eccentric contraction types in 41 male twins (56). A trend was observed that heritabilities
for eccentric contraction torques were higher than isometric and concentric contractions, and
that heritabilities for contraction torques at higher concentric speeds were lower compared to
contractions at slower speeds. Also for torques at the elbow, knee, and trunk it was found that

256
257

Heritability of muscle size

dynamic isometric torques showed lower upper-​heritability values (63–​87%) compared to iso-
metric strength test (82–​96%) at a same muscle length in 748 males siblings from 335 families
from the Leuven Genes for Muscular Strength project (21).

Heritability of power and jump test performance

Muscle contractions employing as much strength as possible in the shortest possible time, also
indicated as “power,” are often quantified through jumps or specific dynamometer tests. Six
studies (7, 9, 27, 33, 37, 47) including 30–​498 young subjects (average ages between 9 and
23.8 years) from (extended) twin designs estimated the contribution of genetic factors to indi-
vidual variation in jump performance between 0.49 and 0.86 (meta-​analysis average h2=0.55;
95% CI 0.45–​0.65). Heritability estimates for leg extension tests and back lifting work were
between 0.37 and 0.60 in middle-​aged twins (1, 46, 59).

Heritability of muscle functional tests

More functional strength tests that measure components of dynamic strength endurance (e.g.,
push-​ups, sit-​ups, bent arm hang, leg lifts, elbow flexion one-​repetition maximum (1RM) tests)
are mostly studied in younger twins and families. Since contraction type and velocity can be
mixed within such tests and different between tests, estimation of an average heritability by
means of meta-​analysis might be less informative. Zempo et al. (60) included nine measurements
from five studies and reported a meta-​analysis average h2 of 0.49 (95% CI 0.32–​0.67).
The weighted mean of h2 for muscle strength-​related phenotypes (effects of additive genetic
and genetic dominance effects summed) over all selected studies and measurements –​different
strength phenotypes or different estimates for males and females were included as separate
entries in the meta-​analysis –​was 0.52 (95% CI 0.48–​0.56) (60). This analysis showed signifi-
cant heterogeneity (I2=91.0%, P <0.001) and was further explored by meta-​regression analysis.
Only age was a significant contributor to the heterogeneity, with lower heritability estimates
in older samples (r=−0.43; specifically for handgrip r=−0.70). The mean h2 for strength-​related
phenotypes tended to be higher in men than in women, which only became significant for
the sex difference in heritability of handgrip strength. This meta-​analysis did not show signifi-
cant differences between h2 estimates for the different subphenotypes (grip, isometric, dynamic,
isotonic, jump tests), upper or lower body strength, type of SEM model reported, or cohort
type (60).

Multivariate heritability studies for muscle mass and strength

Differences in the broad set of strength-​related subphenotypes seem not to vary independently
from each other within and between individuals. Statistical evidence for a “strength gener-
ality” factor is suggested by moderate to high correlations between different strength tests (18).
Whether observed phenotypic correlations are based on shared genetic (additive genetic, A; or
genetic dominance, D) and/​or shared environmental (environmental factors common within
families, C; or specific to each individual, E) factors can be explored using multivariate twin
models.
Schutte et al. (47) estimated a genetic correlation between handgrip and vertical jump of
rG=0.46 (95% CI 0.27–​0.65) in 227 17.2 ± 1.2 year-​old twins. Elbow strength measured at
a flexion angle of 110° during maximum concentric and eccentric contractions at 120°/​s, as
well as isometrically, was related to the muscle cross-​sectional area (MCSA) at the mid-​upper

257
258

M. Thomis

arm in a study of 41 young adult male twins (22.4 ± 3.7 year) (12). A model with a “general”
(Ac) genetic factor and a general specific environmental factor (Ec) together with phenotype-​
specific genetic (AS1–​4) and environmental factors (ES1–​4) fitted the data best. The shared
genetic and environmental component, accounted for 43% and 6% in MCSA (h2=81%), 47%
and 20% in eccentric (h2=65%), 58% and 4% in isometric (h2=70%), and 32% and 1% in con-
centric strength (h2=32%), respectively. Contraction type-​specific and muscle cross-​sectional
area-​specific genetic and environmental effects, accounted for 38% and 14% in MCSA, 18%
and 15% in eccentric, 12% and 26% in isometric, and 0% and 67% in concentric strength,
respectively. These results support a shared pleiotropic gene action for MCSA, eccentric, iso-
metric, and concentric strength, with a moderate to high genetic contribution to the vari-
ability of these characteristics. In the same set of Flemish male twins, covariation in MCSA
and static elbow torques in at different muscle lengths (at 140°, 110°, and 80° elbow flexion)
could also be explained by a set of shared genetic factors explaining 82% (MSCA), 66% (at
140°), 61% (at 110°), and 50% (at 80°) of the variance (54). An additional torque-​specific
genetic factor was also included (explaining an additional 6–​24% of variance in the elbow
flexion torques).
In an older, female twin sample (101 MZ, 116 DZ, 63–​76 years) within the Finnish Twin
Study on Aging (FITSA), knee extension power and maximal isometric knee extension strength
were analyzed in a bivariate reduced ACE Cholesky decomposition model (58). A shared gen-
etic component accounted for 32% of the total variance in leg extensor power and 48% in
isometric knee extensor strength. A nonshared environmental effect shared by both phenotypes
accounted for 4% of the variance in power and 52% in isometric strength. For leg extensor
power, common environmental factors were also significant (28%). Environmental sources of
variation might include general levels of physical activity or nutritional choices that act upon
the development of muscle mass and all measures of strength (isometric, dynamic, power) and
therefore contribute to the shared sets of unique environmental factors. However, specificity
of training effects might contribute to the different contributions of environmental factors in
different types of muscle contraction. As for genes contributing to multiple forms of muscular
strength, candidates can be found in structural elements of muscle and contractile proteins or
fiber type distribution. Specific genetic contributions might also relate to the importance of
fiber type distribution (e.g., related more strongly to muscle power than muscle strength), and
passive elastic components that are contributing to torques in specific muscle contraction types
(e.g., eccentric torques).
Back strength was evaluated by an isometric trunk extensor endurance test, isokinetic lifting
force, and power and psychophysical lifting test in 122 MZ and 131 DZ male, middle-​aged
(49.9 ± 7.7 years) twins (46). A general additive genetic factor was included in the model
and explained 57–​92% of the covariance among the three back tests and 31% of the covari-
ance between isokinetic lifting force and work. Genetic dominance was significantly shared
between isokinetic lifting work and force and isokinetic lifting work had a specific dominance
factor. Common environmental factors contributed test-​specifically to variance in psychophys-
ical lifting force (18%) and isometric trunk extensor endurance (34%). Unique environmental
factors were shared between the four phenotypes (explaining between 1% and 49% of the vari-
ance) and had test-​specific contributions (between 20% and 60%). Covariation explained by
unique environmental factors showed much variation between the tests (re=0.03–​0.43). Broad
heritability (sum of A and D sources) was highest for isokinetic lifting force (65%) and work
(60%), and lowest for psychophysical lifting force (33%) and the trunk extensor endurance test
(5%). A set of models including measured environmental determinants were tested and showed
significance for participation in power sports and body weight for isokinetic force and power,

258
259

Heritability of muscle size

but limited contributions of other factors (health compared to others, type of job, doing aerobic
sports) in the other back strength measures.

Heritability of muscle mass and strength from longitudinal studies

Sex differences in the intercept and slope of growth in handgrip strength were the focus of
variance components analysis using a latent growth curve model analysis in a study of 2513
same-​sex twins from 11.8 years (baseline) to 17.8 years of age (second follow-​up) (22). High
heritabilities were reported for the variability in handgrip strength at 11 years of age (intercept)
(males=88%, females=79%). Additive genetic effects accounted for most (80%) of the variance
around the slope in males but were of less importance in females (h2=28%). The absolute gen-
etic variance around the slope (increased by age) was nearly ninefold higher in males.Variation
in timing of the adolescent growth spurt is a confounding factor when studying development
of muscle mass and strength and its genetic and environmental components during adolescence.
Longitudinal models were applied to study the tracking in isometric (arm pull) (41) and explo-
sive (vertical jump) (40) strength in 105 female and male twins of the Leuven Longitudinal
Twins Study, between 10 and 18 years of age. All measurements were aligned according to the
age at peak height velocity (APHV) for each individual. A model including genetic and envir-
onmental innovation and transmission paths could account for the tracking observed in these
measurements during adolescence. Arm pull strength 3 years after peak height velocity could
be explained by genetic (44.3% in males, 22.5% in females) and environmental factors (31.2%
in males, 44.5% in females) transmitted from previous time points (41). In males, new gen-
etic factors seem to increase the overall variance by age, specifically at the time around APHV
(innovation A=30.7%). Heritability estimates in vertical jump performance ranged between
60.8% (95% CI 37.7–​77.2%) and 87.3% (95% CI 74.2–​94.0%) for boys and between 76.5%
(95% CI 56.7–​89.0%) and 88.6% (95% CI 77.8–​94.1%) for girls. Up to 56.4% and 62.8% of
the total variation 3 years after APHV was explained by additive genetic factors that already
explained a significant amount of variation at previous measurement occasions in boys and girls
respectively (40). It thus can be concluded that the observed stability of strength during adoles-
cence is caused by a stable genetic influence, more so in boys than in girls.
Longitudinal genetic modeling in other age periods are limited. One study reported 10-​year
aging effects on handgrip strength in 77 MZ and 75 DZ twin pairs at mean ages of 63 and
73 years. A bivariate ACE Cholesky decomposition model was used to explore the observed
stability in handgrip strength (r=0.62). Shared genetic factors accounted for 35% of the pheno-
typic correlation, while shared familial environmental effects accounted for 48% of the stability
(8). In middle-​aged siblings (N=115 subjects from 48 families, age at baseline 44.8 ± 7 years)
upper-​limit heritability estimates for changes in knee strength over 2.4 years (−2.8 ± 8.6%/​
year) were between 54% and 74% (depending on the correction for covariates related to bone
and cartilage characteristics related to osteoarthritis) (61).

Role of genetic factors in muscle mass and strength responses to exercise

Few studies have used a family or twin design to investigate the role of genetic factors in
the interindividual responses in muscle mass and strength after resistance/​strength training.
The MZ twin design to explore genotype–​training interaction effects was applied in five MZ
pairs (17–​26 years) in response to 10 weeks of maximal isokinetic knee flexion/​extension
training (52). No significant genotype–​training interaction could be found for the response in
peak torque (mean ± SD: 24% ± 12%), although the response in oxoglutarate dehydrogenase

259
260

M. Thomis

activity showed significant intrapair resemblance (r=0.76). Male MZ (N=25) and DZ (N=16)
twins (22.4 ± 3.7 years) participated in a 10-​week resistance training program for the elbow
flexors (55). The evidence for genotype–​training interaction, or association of interindividual
differences in training effects with the genotype, was tested by a two-​way ANOVA in the MZ
twins and using a bivariate model-​fitting approach on pre-​and post-​training phenotypes in MZ
and DZ twins. Evidence for genotype–​training interaction was found for 1RM and isometric
strength, with MZ intrapair correlations of 0.46 and 0.30, respectively. Bivariate model-​fitting
indicated that about 20% of the variation in post-​training 1RM, isometric strength, and con-
centric moment at 120°/​s was explained by training-​specific genetic factors that were inde-
pendent from genetic factors that explained variation in the pretraining phenotype (30–​77%).
A hypertrophic response of arm cross-​sectional muscle area (CT) was significant and varied
between subjects (2.2 cm2 ± 2.7); however, no significant genotype–​training interaction effect
was found. Compared to studies in which anaerobic/​intermittent or endurance training were
used as exercise regimens (15, 50), strength responses seem to show lower/​smaller evidence for
genotype–​training interaction.

The studies summarized in this chapter indicate that individual differences in muscle mass,
muscle strength and power, and responses to resistance training are partially determined by gen-
etic factors. Although heritability estimates vary according to sample characteristics (age, sex),
heritability estimation technique, specific strength test, and characteristic (isometric, isotonic,
dynamic, functional, eccentric, concentric, strength endurance), it has pushed the field into the
search for specific gene variants that underlie the heritability of these traits (53). Environmental
factors are mainly identified as unique to the individual and less often “shared” by family
members, however, these specific factors (e.g., nutrition, occupation, type of exercise activities,
etc.) can be shared between different types of muscle strength. Covariation and uniqueness of
multiple types of muscle contraction and strength tests have been explored in multivariate gen-
etic and environmental models. Although shared genetic variation indicated pleiotropic gene
actions in most of these analyses, genetic variation specific to each subtrait of strength also
indicate specific gene sets for specific strength characteristics. Stability of strength characteristics
over time in adolescence and in older age also show a genetic component. In summary, twin
and family studies have indicated that genetic factors significantly (>0.50–​0.70) contribute to
the observed differences in muscle mass and strength characteristics in young and old, males and
females and that efforts to investigate DNA sequence variation in candidate gene or genome-​
wide approaches was merited.

References
1. Arden NK and Spector TD. Genetic influences on muscle strength, lean body mass, and bone min-
eral density: a twin study. J Bone Miner Res 12: 2076–​2081, 1997.
2. Baar K. Epigenetic control of skeletal muscle fibre type. Acta Physiol (Oxf) 199: 477–​487, 2010.
3. Baguet A, Everaert I, Achten E, Thomis M and Derave W. The influence of sex, age and herit-
ability on human skeletal muscle carnosine content. Amino Acids 43: 13–​20, 2012.
4. Baguet A, Everaert I, Hespel P, Petrovic M, Achten E and Derave W. A new method for non-​
invasive estimation of human muscle fiber type composition. Plos One 6: e21956, 2011.
5. Bouchard C, Malina R and Bar-​Or O. Growth, Maturation, and Physical Activity. Champaign,
IL: Human Kinetics Publishers, 2004.
6. Byard PJ, Sharma K, Russell JM and Rao DC. A family study of anthropometric traits in a Punjabi
community: II. An investigation of familial transmission. Am J Phys Anthropol 64: 97–​104, 1984.

260
261

Heritability of muscle size

7. Calvo M, Rodas G, Vallejo M, Estruch A, Arcas A, Javierre C, Viscor G and Ventura JL.
Heritability of explosive power and anaerobic capacity in humans. Eur J Appl Physiol 86: 218–​225, 2002.
8. Carmelli D and Reed T. Stability and change in genetic and environmental influences on hand-​
grip strength in older male twins. J Appl Physiol (1985) 89: 1879–​1883, 2000.
9. Chatterjee S and Das N. Physical and motor fitness in twins. Jpn J Physiol 45: 519–​534, 1995.
10. Clark PJ. The heritability of certain anthropometric characters as ascertained from measurements of
twins. Am J Hum Genet 8: 49–​54, 1956.
11. Culjkovic-​Kraljacic B, Baguet A, Volpon L, Amri A and Borden KL. The oncogene eIF4E
reprograms the nuclear pore complex to promote mRNA export and oncogenic transformation. Cell
Rep 2: 207–​215, 2012.
12. De Mars G, Thomis MAI, Windelinckx A, Van Leemputte M, Maes HH, Blimkie CJ, Claessens
AL, Vlietinck R and Beunen G. Covariance of isometric and dynamic arm contractions: multi-
variate genetic analysis. Twin Res Hum Genet 10: 180–​190, 2007.
13. Dupae E, Defrise-​Gussenhoven E and Susanne C. Genetic and environmental influences on
body measurements of Belgian twins. Acta Genet Med Gemellol (Roma) 31: 139–​144, 1982.
14. Frederiksen H, Gaist D, Petersen HC, Hjelmborg J, McGue M, Vaupel JW and Christensen
K. Hand grip strength: a phenotype suitable for identifying genetic variants affecting mid-​and late-​
life physical functioning. Genet Epidemiol 23: 110–​122, 2002.
15. Hamel P, Simoneau JA, Lortie G, Boulay MR and Bouchard C. Heredity and muscle adapta-
tion to endurance training. Med Sci Sports Exerc 18: 690–​696, 1986.
16. Hewitt D. Sib resemblance in bone, muscle and fat measurements of the human calf. Ann Hum Genet
22: 213–​221, 1958.
17. Heymsfield SB, Adamek M, Gonzalez MC, Jia G and Thomas DM. Assessing skeletal muscle
mass: historical overview and state of the art. J Cachexia Sarcopenia Muscle 5: 9–​18, 2014.
18. Hortobagyi T, Katch FI and LaChance PF. Interrelationships among various measures of upper
body strength assessed by different contraction modes. Evidence for a general strength component.
Eur J Appl Physiol Occup Physiol 58: 749–​755, 1989.
19. Hoshi H, Ashizawa K, Kouchi M and Koyama C. On the intra-​pair similarity of Japanese mono-
zygotic twins in some somatological traits. Okajimas Folia Anat Jpn 58: 675–​686, 1982.
20. Hsu FC, Lenchik L, Nicklas BJ, Lohman K, Register TC, Mychaleckyj J, Langefeld CD,
Freedman BI, Bowden DW and Carr JJ. Heritability of body composition measured by DXA in
the diabetes heart study. Obes Res 13: 312–​319, 2005.
21. Huygens W, Thomis MA, Peeters MW, Vlietinck RE and Beunen GP. Determinants and
upper-​limit heritabilities of skeletal muscle mass and strength. Can J Appl Physiol 29: 186–​200, 2004.
22. Isen J, McGue M and Iacono W. Genetic influences on the development of grip strength in ado-
lescence. Am J Phys Anthropol 154: 189–​200, 2014.
23. Janssen I, Shepard DS, Katzmarzyk PT and Roubenoff R. The healthcare costs of sarcopenia in
the United States. J Am Geriatr Soc 52: 80–​85, 2004.
24. Karasik D, Zhou Y, Cupples LA, Hannan MT, Kiel DP and Demissie S. Bivariate genome-​
wide linkage analysis of femoral bone traits and leg lean mass: Framingham study. J Bone Miner Res
24: 710–​718, 2009.
25. Katzmarzyk PT, Gledhill N, Perusse L and Bouchard C. Familial aggregation of 7-​year changes
in musculoskeletal fitness. J Gerontol A Biol Sci Med Sci 56: B497–​502, 2001.
26. Komi PV, Viitasalo JH, Havu M, Thorstensson A, Sjodin B and Karlsson J. Skeletal muscle
fibres and muscle enzyme activities in monozygous and dizygous twins of both sexes. Acta Physiol
Scand 100: 385–​392, 1977.
27. Kovar R. Genetic analysis of motor performance. J Sports Med Phys Fitness 16: 205–​208, 1976.
28. Landers KA, Hunter GR, Wetzstein CJ, Bamman MM and Weinsier RL. The interrelationship
among muscle mass, strength, and the ability to perform physical tasks of daily living in younger and
older women. J Gerontol A Biol Sci Med Sci 56: B443–​448, 2001.
29. Leong DP, Teo KK, Rangarajan S, Lopez-​Jaramillo P, Avezum A, Jr., Orlandini A, Seron P,
Ahmed SH, Rosengren A, Kelishadi R, Rahman O, Swaminathan S, Iqbal R, Gupta R, Lear
SA, Oguz A, Yusoff K, Zatonska K, Chifamba J, Igumbor E, Mohan V, Anjana RM, Gu H,
Li W, Yusuf S and Prospective Urban Rural Epidemiology Study investigators. Prognostic
value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study.
Lancet 386: 266–​273, 2015.

261
262

M. Thomis

30. Livshits G, Gao F, Malkin I, Needhamsen M, Xia Y, Yuan W, Bell CG, Ward K, Liu Y, Wang
J, Bell JT and Spector TD. Contribution of heritability and epigenetic factors to skeletal muscle
mass variation in United Kingdom twins. J Clin Endocrinol Metab 101: 2450–​2459, 2016.
31. Loos R, Thomis M, Maes HH, Beunen G, Claessens L, Derom C, Legius E, Derom R and
Vlietinck R. Gender-​specific regional changes in genetic structure of muscularity in early adoles-
cence. J Appl Physiol (1985) 82: 1802–​1810, 1997.
32. Lortie G. Muscle fiber type composition and enzyme activities in brothers and monozygotic twins.
In: Sports and Human Genetics, edited by Malina R and Bouchard C. Champaign, IL: Human Kinetics,
1986, p. 147–​153.
33. Maes HHM, Beunen GP, Vlietinck RF, Neale MC, Thomis M, VandenEynde B, Lysens R,
Simons J, Derom C and Derom R. Inheritance of physical fitness in 10-​yr-​old twins and their
parents. Med Sci Sports Exerc 28: 1479–​1491, 1996.
34. Malina RM and Mueller WH. Genetic and environmental influences on the strength and motor
performance of Philadelphia school children. Hum Biol 53: 163–​179, 1981.
35. Masschelein E, Van Thienen R, Thomis M and Hespel P. High twin resemblance for sensitivity
to hypoxia. Med Sci Sports Exerc 47: 74–​81, 2015.
36. Mitchell WK, Williams J, Atherton P, Larvin M, Lund J and Narici M. Sarcopenia, dynapenia,
and the impact of advancing age on human skeletal muscle size and strength; a quantitative review.
Front Physiol 3: 260, 2012.
37. Okuda E, Horii D and Kano T. Genetic and environmental effects on physical fitness and motor
performance. Int J Sport Health Sci 3: 1–​9, 2005.
38. Pajala S, Era P, Koskenvuo M, Kaprio J, Tolvanen A, Heikkinen E, Tiainen K and Rantanen
T. Contribution of genetic and environmental effects to postural balance in older female twins. J Appl
Physiol (1985) 96: 308–​315, 2004.
39. Payne N, Gledhill N, Katzmarzyk PT, Jamnik V and Ferguson S. Health implications of mus-
culoskeletal fitness. Can J Appl Physiol 25: 114–​126, 2000.
40. Peeters MW, Thomis MA, Maes HH, Loos RJ, Claessens AL, Vlietinck R and Beunen GP.
Genetic and environmental causes of tracking in explosive strength during adolescence. Behav Genet
35: 551–​563, 2005.
41. Peeters MW, Thomis MA, Maes HHM, Beunen GP, Loos RJF, Claessens AL and Vlietinck
R. Genetic and environmental determination of tracking in static strength during adolescence. J Appl
Physiol (1985) 99: 1317–​1326, 2005.
42. Perusse L, Lortie G, Leblanc C, Tremblay A, Theriault G and Bouchard C. Genetic and envir-
onmental sources of variation in physical fitness. Ann Hum Biol 14: 425–​434, 1987.
43. Pitsiladis Y, Wang G, Wolfarth B, Scott R, Fuku N, Mikami E, He Z, Fiuza-​Luces C, Eynon
N and Lucia A. Genomics of elite sporting performance: what little we know and necessary
advances. Br J Sports Med 47: 550–​555, 2013.
44. Reed T, Fabsitz RR, Selby JV and Carmelli D. Genetic influences and grip strength norms in the
NHLBI twin study males aged 59–​69. Ann Hum Biol 18: 425–​432, 1991.
45. Rico-​Sanz J, Rankinen T, Joanisse DR, Leon AS, Skinner JS, Wilmore JH, Rao DC, Bouchard
C and Study HF. Familial resemblance for muscle phenotypes in the HERITAGE Family Study. Med
Sci Sports Exerc 35: 1360–​1366, 2003.
46. Ropponen A, Levalahti E, Videman T, Kaprio J and Battie MC. The role of genetics and envir-
onment in lifting force and isometric trunk extensor endurance. Phys Ther 84: 608–​621, 2004.
47. Schutte NM, Nederend I, Hudziak JJ, de Geus EJ and Bartels M. Differences in adolescent
physical fitness: a multivariate approach and meta-​analysis. Behav Genet 46: 217–​227, 2016.
48. Silventoinen K, Magnusson PK, Tynelius P, Kaprio J and Rasmussen F. Heritability of body
size and muscle strength in young adulthood: a study of one million Swedish men. Genet Epidemiol
32: 341–​349, 2008.
49. Simoneau JA and Bouchard C. Genetic determinism of fiber type proportion in human skeletal
muscle. FASEB J 9: 1091–​1095, 1995.
50. Simoneau JA, Lortie G, Boulay MR, Marcotte M, Thibault MC and Bouchard C. Inheritance
of human skeletal muscle and anaerobic capacity adaptation to high-​intensity intermittent training.
Int J Sports Med 7: 167–​171, 1986.
51. Strollo SE, Caserotti P, Ward RE, Glynn NW, Goodpaster BH and Strotmeyer ES. A review
of the relationship between leg power and selected chronic disease in older adults. J Nutr Health Aging
19: 240–​248, 2015.

262
263

Heritability of muscle size

52. Thibault MC, Simoneau JA, Cote C, Boulay MR, Lagasse P, Marcotte M and Bouchard
C. Inheritance of human muscle enzyme adaptation to isokinetic strength training. Hum Hered
36: 341–​347, 1986.
53. Thomis M. Genes and strength and power phenotypes. In: Genetic and Molecular Aspects of Sport
Performance, edited by Bouchard C and Hoffman EP. Chichester: Wiley-​Blackwell, 2011, p. 159–​176.
54. Thomis MA, VanLeemputte M, Maes HH, Blimkie CJR, Claessens AL, Marchal G, Willems
E, Vlietinck RF and Beunen GP. Multivariate genetic analysis of maximal isometric muscle force
at different elbow angles. J Appl Physiol (1985) 82: 959–​967, 1997.
55. Thomis MAI, Beunen GP, Maes HH, Blimkie CJ, Van Leemputte M, Claessens AL, Marchal
G, Willems E and Vlietinck RF. Strength training: importance of genetic factors. Med Sci Sports
Exerc 30: 724–​731, 1998.
56. Thomis MAI, Beunen GP, Van Leemputte M, Maes HH, Blimkie CJ, Claessens AL, Marchal
G, WIllems E and Vlietinck RF. Inheritance of static and dynamic arm strength and some of its
determinants. Acta Physiol Scand 163: 59–​71, 1998.
57. Tiainen K, Sipila S, Alen M, Heikkinen E, Kaprio J, Koskenvuo M, Tolvanen A, Pajala S and
Rantanen T. Heritability of maximal isometric muscle strength in older female twins. J Appl Physiol
(1985) 96: 173–​180, 2004.
58. Tiainen K, Sipila S, Alen M, Heikkinen E, Kaprio J, Koskenvuo M, Tolvanen A, Pajala S and
Rantanen T. Shared genetic and environmental effects on strength and power in older female twins.
Med Sci Sports Exerc 37: 72–​78, 2005.
59. Tiainen KM, Perola M, Kovanen VM, Sipila S, Tuononen KA, Rikalainen K, Kauppinen MA,
Widen EI, Kaprio J, Rantanen T and Kujala UM. Genetics of maximal walking speed and skeletal
muscle characteristics in older women. Twin Res Hum Genet 11: 321–​334, 2008.
60. Zempo H, Miyamoto-​Mikami E, Kikuchi N, Fuku N, Miyachi M and Murakami H.
Heritability estimates of muscle strength-​related phenotypes: A systematic review and meta-​analysis.
Scand J Med Sci Sports 27: 1537–​1546, 2017.
61. Zhai G, Ding C, Stankovich J, Cicuttini F and Jones G. The genetic contribution to longitudinal
changes in knee structure and muscle strength: a sibpair study. Arthritis Rheum 52: 2830–​2834, 2005.
62. Zhai G, Stankovich J, Ding C, Scott F, Cicuttini F and Jones G. The genetic contribution to
muscle strength, knee pain, cartilage volume, bone size, and radiographic osteoarthritis: a sibpair study.
Arthritis Rheum 50: 805–​810, 2004.
63. Zierath JR and Hawley JA. Skeletal muscle fiber type: influence on contractile and metabolic
properties. PLoS Biol 2: e348, 2004.

263
264

19
GENETIC CONTRIBUTIONS TO
MUSCLE STRENGTH
Matthew D. Barberio, Emidio E. Pistilli, and Monica J. Hubal

Muscular strength is defined as the ability to generate force against a resistance and can be
thought of in terms of force output by an individual skeletal muscle or a group of skeletal
muscles exerting force in a coordinated pattern. Strength is influenced by both intrinsic (i.e.,
genetic) and extrinsic (i.e., environmental) factors. This book’s previous chapter (Chapter 18)
explored the heritability of muscle size and strength traits at baseline and changes in these traits
following exercise training.This chapter will further explore muscular strength as a multifaceted
trait that is affected by underlying components such as muscle mass, neuromuscular activity
patterns, and executable skill in performing a variety of tasks. Muscle size and strength are gen-
erally positively related to one another (especially at baseline), but this relationship is complex
with high intersubject variability, especially during training or detraining.This chapter will serve
as a primer to identify important components of muscle strength, highlight biological pathways
that modify strength at baseline and strength changes with training, and identify molecular and
genetic modifiers of strength. We refer the reader to later chapters for expanded discussions of
modifiers of complementary traits such as muscle size (Chapter 20) and neuroendocrine status
(Chapter 21).

Muscular strength primer

Skeletal muscle strength is a key determinant of overall functional capacity, and progressive resist-
ance training can stimulate significant gains in muscular strength in both men and women at
any stage of life (1–​3). In exploring the muscle strength trait, defining a set of commonly used
terms is useful. In terms of muscle actions performed during resistance exercise training, three
types of action are possible (4, 5). A concentric contraction occurs when activated muscle shortens
and produces greater force than the external resistance. An eccentric or lengthening action occurs
when activated muscle lengthens rather than shortens because the external resistance is greater
than the force that muscle can generate. An isometric contraction occurs when activated muscle does
not change length as the external force matches the internal force generation. Due to muscle’s
length–​tension relationship, the magnitude of force production during these types of muscle
actions are eccentric >isometric >concentric (6). Also, force is inversely related to contraction
speed (7). Absolute strength is a measure of the total amount of weight that can be lifted by a

264
265

Genetic contributions to muscle strength

muscle or group of muscles during a given task. In contrast, relative strength is a measure of the
amount of weight lifted by a muscle or group of muscles normalized to some measure of body
size. One common method to determine absolute strength is to quantify a repetition maximum
(RM). For example, a one-​repetition maximum would be the maximal weight that a person
can lift for one repetition when performing a resistance exercise task. Adding a time aspect to
strength, power output is the rate of doing work, defined as force output applied over a given
distance and time (4, 5). Although often used interchangeably, muscular strength and muscular
power refer to different types of adaptations that can result from resistance training. Muscular
endurance is the ability to repeatedly contract a muscle against a resistance while maintaining force
output over time. Strength training programs can be designed to target these different aspects of
muscle strength. For example, increases in absolute strength could be achieved by lifting maximal
weights for very few repetitions (i.e., 1–​3 repetitions), while muscular endurance can be trained
by lifting submaximal weights for a higher number of repetitions (i.e., 20 repetitions).
Unlike other types of fitness (e.g., cardiovascular, flexibility), strength should not be thought
of as a single factor (6). Resistance exercise training can be used to enhance many different
expressions of muscular strength, such as maximal muscle force, muscular power, contraction
velocity, or muscular endurance, and no single strength training program could enhance all
of these factors simultaneously. In other words, muscle performance adaptations in response
to strength training are specific to the strength training program utilized. In a similar way, the
underlying molecular adaptations within skeletal muscles that occur in response to different
resistance training programs are also generally protocol-specific (8, 9). For example, training
for muscular endurance would evoke more changes in mitochondrial volume and activity
compared to maximal weight/​low-​repetition training.
Strength training programs can be designed to specifically focus on strength adaptations in
individual muscles or on the performance of a coordinated movement task. Resistance training
programs that are designed to promote strength and size of individual skeletal muscles are
referred to as “structural” resistance training programs, while programs aimed at initiating a per-
formance increase are referred to as “functional” resistance training programs (6). With this in
mind, it is important to note that maximal strength production requires not only contraction
of the skeletal muscles involved in the movement, but also the maximal activation of all motor
units available in those contracting skeletal muscles by the nervous system. Siff states that a “fun-
damental principle of strength training is that all strength increases are initiated by neuromus-
cular stimulation” (6). Therefore, the interaction of the nervous system and the skeletal muscles
must be considered in the overall training program if the goal is the development of maximal
strength. For readers interested in an extended view of neuroendocrine responses to training,
Chapter 21 of this book will discuss this topic at length.

Composition of skeletal muscle

Skeletal muscle is a complex tissue that is composed of multiple layers that each contribute
to muscle contraction and force production (reviewed in 10, 11). At the molecular level, indi-
vidual myosin heads interact with binding sites on actin filaments to form crossbridges, and
these crossbridges perform the myosin power stroke using the energy of ATP hydrolysis. The
coordinated, and asynchronous, movement of individual crossbridges results in shortening of
sarcomeres, which are the basic units of muscle contraction. As sarcomeres shorten, the actin
and myosin filaments slide past each other (i.e., sliding filament theory). Sarcomeres are lined up
in series within myofibrils, and multiple myofibrils are contained within skeletal muscle fibers.

265
266

M.D. Barberio et al.

Skeletal muscle fibers are the cells of skeletal muscle. Numerous muscle fibers are packaged into
whole skeletal muscles and, architecturally, can either run from tendon to tendon (i.e., fusiform
muscle) or at angles to the line of pull (i.e., pennate muscle). During a concentric contraction,
the myosin power stroke and resultant force output of each individual crossbridge is summated
within all the sarcomeres in a myofibril, for all myofibrils within each muscle fiber, and for all
muscle fibers in a skeletal muscle, which gives rise to muscle contraction and skeletal muscle
force output.
Baseline muscle strength in humans and the muscle strength responses to training both
display high intersubject variability, due to various factors such as muscle composition,
training status, age, and sex. In vastus lateralis muscle biopsies obtained from untrained,
sedentary humans, the percentage of type I and type II fiber expression is approximately
50% (reviewed in 12). Although muscle fiber type expression is a dynamic process and is
influenced by training stimuli, there is a limit to the extent that fiber type changes can be
manipulated by training alone. Genetic predisposition for a specific muscle fiber type dis-
tribution is heritable and can influence muscle strength and athletic performance. Genetic
predisposition for a greater expression of type II muscle fibers is associated with greater
baseline muscle strength and enhanced responses to strength training programs, while a
greater expression of type I muscle fibers is associated with enhanced adaptations to endur-
ance training programs (13–​16). Heritability and the genetic basis for muscle strength and
size traits at baseline and in response to exercise training were explored to a greater extent
in Chapter 18 of this book.

Development of muscular strength

In the seminal paper by Moritani and DeVries (17), the time course for development of mus-
cular strength upon initiation of resistance training was established (also reviewed in 18). Early
increases in strength in response to resistance exercise training are ascribed to adaptations in
the nervous system, as a neuromuscular program is imprinted to the new movement patterns
being performed. During this initial period of time, as muscle strength increases in response to
the resistance exercise training, minimal muscle hypertrophy is observed. With continuation of
resistance training, however, strength continues to increase via muscle hypertrophy. Adaptations
in the nervous system are thought to be responsible for the gains in strength that occur within
the first 3–​5 weeks of starting resistance exercise training (17). After this initial period of time,
muscle hypertrophy becomes more evident, which coincides with further strength increases.
However, muscle hypertrophy is not infinite, and there are plateaus in both strength and size
development, even in those who continue to train for a number of years.
Adaptations in the nervous system that contribute to increases in muscular strength are well
established in the literature (reviewed in 19, 20). For example, DeLorme and Watkins (21), first
noted that “this initial increase in strength on progressive resistance exercise occurs at a rate far
greater than can be accounted for by changes in the muscle.” Studies have observed that increases
in strength from resistance training within the first few weeks of training are quantifiable des-
pite minimal measurable increases in muscle size (22, 23). In further support of the ability of
the nervous system to induce strength gains, unilateral resistance training is associated with
gains in strength in the untrained limb (24, 25). This was documented in the Functional Single
Nucleotide Polymorphisms Associated with Human Muscle Size and Strength (FAMuSS) study
(26, 27), in which 12 weeks of resistance training targeting the upper arm of the nondominant
arm was associated with a 54% increase in strength in the trained arm and an 11% increase in
strength in the nontrained arm. These post-​training increases in strength were associated with a

266
267

Genetic contributions to muscle strength

19% increase in size of the trained arm, compared to a 1.4% increase in size of the nontrained
arm (28), supporting a complex relationship between strength and size with training.
Continued increases in muscle strength, following early adaptations in the nervous system,
are ascribed to exercise-​induced muscle hypertrophy. Muscle hypertrophy can be induced
through the activation of numerous signaling pathways (reviewed in 29). Resistance exercise
initiates increases in protein accumulation, which is mainly due to greater protein synthesis and
slowed protein degradation (30). Greater protein accumulation within skeletal muscle gives
rise to larger and more abundant myofibrils within skeletal muscle fibers. In addition, resist-
ance exercise training stimulates the proliferation of satellite cells, which also contribute to
increases in muscle size. Following resistance training, the number of myonuclei is increased
relative to pretraining numbers (31). Therefore, resistance training stimulates protein synthesis
that leads directly to larger myofibrils, a greater number of myofibrils (9), as well as an increase
in myonuclear number (31–​33). Multiple molecular pathways are upregulated in response to
resistance training and help to facilitate increases in muscle size and strength. However, genetic
variability in the molecules at play in these pathways can lead to variability in the responses to
resistance training (i.e., muscle strength, muscle hypertrophy). More information about factors
that underlie muscle mass and its changes with training can be found in Chapter 20.

Genetic variations associated with muscular strength –​

methodological approaches
Many factors mediate muscle strength (size, neuromuscular efficiency, training history, etc.),
making it difficult to isolate individual genetic variations or molecular pathways as being
important drivers of muscle strength, yet strength as a phenotype is moderately to highly her-
itable (34). Approaches to identifying genetic contributions to strength include both global
(genome-​wide) and targeted (selected gene/​variant) association studies. In the early 2000s, with
the publication of the human genome, there was a plethora of studies targeting genes related to
subcomponents of strength such as muscle size (examples include myostatin, insulin-​like growth
factor 1 (IGF1), alpha-​actinin-​3 (ACTN3), etc.). We briefly review some of the stronger asso-
ciations later in this chapter, and other literature reviews of single nucleotide polymorphisms
and strength traits are available (26, 35, 36). However, many of these studies suffered from lack
of statistical power, lack of adequate controls, lack of validation/​replication, and other problems
related to population genetics studies (for review, see 37). In addition, there are few studies that
adopt a systems biology approach to strength genetics –​following up association studies with
detailed molecular studies to demonstrate functional consequences of genetic variation (though
there are some exceptions to this, such as ACTN3, further discussed in this book in Chapter 23).
Attempting these types of studies in trained populations adds in additional problems, given
that training introduces a large new amount of variability to the strength phenotype. One of
the largest training studies to date was the FAMuSS study, which tested strength and size gains
following 12 weeks of unilateral resistance training in approximately 750 subjects. We refer the
reader to reviews of findings from FAMuSS related to size and strength (26, 27).
Because many of the associations reported in these early studies either have small effect sizes
or have not been consistently replicated (especially in trained athletes), single-​gene variant asso-
ciation studies have fallen out of favor in the literature, replaced by more complex multigene
variant models or genome-​wide association studies (GWAS). The multigene variant model
approach has yet to be applied well to the phenotype of muscle strength, so it will not be
addressed further here. Future work from large consortia such as the Molecular Transducers of
Physical Activity (MoTrPAC) will likely shed light on genetic contributions to strength using

267
268

M.D. Barberio et al.

multivariant modeling. Likewise, future studies using the GWAS approach should be inform-
ative. Unfortunately, the GWAS approach requires extremely large population sizes, as well as
large validation and replication sets, to identify loci associated with a particular trait. GWAS
have not lived up to the expectation that once scientists could look across the entire genome at
once, that disease-​or trait-​associated loci would be easily identified. In the largest (N=195,180)
GWAS study to date to specifically examine strength (38), 16 loci were deemed significantly
associated with grip strength, all in genes that were largely untargeted by previous studies but
that represent new areas of exploration, including neuronal maintenance and signaling and psy-
chomotor skills.
As we await more results from large, well-​controlled studies of genetic associations with
strength, we will conclude this chapter with a summary of cellular and molecular pathways
that play direct or indirect roles in muscle strength and how genetic variation in pathway
subcomponents contribute to strength variability. By necessity, these results overlap somewhat
with pathways that control muscle growth, due to the overall relationship between strength
and size.

Cellular pathways related to muscle strength

The functional relationship between strength and muscle size at the molecular level is best
explained by the relationship between protein synthesis and protein degradation. Depending on
the health status of the population, a single bout of resistance exercise can result in significantly
increased rates of protein synthesis as early as 1 hour after exercise which persist for up to 48
hours. Protein degradation is also upregulated in response to resistance exercise, though propor-
tionally less so than protein synthesis resulting in a net protein accretion. The molecular drivers
and contributors to regulating protein synthesis and protein degradation pathways are complex
and not fully understood. Here we will explore major growth, inflammatory, and neuromus-
cular pathways that are responsive to tension/​load stimuli and their role in the development of
muscular strength.

Regulation of muscle size and strength by growth factors

While there are numerous pathways associated with regulation of protein synthesis, the central,
and most well understood, pathway is the phosphatidylinsositol-​3-​kinase (PI3K)/​protein kinase
B (Akt)/​mammalian target of rapamycin (mTOR) axis. It should be noted that these pathways,
independent of other signaling molecules, are activated by mechanotransduction, but are fur-
ther augmented by naturally occurring proteins called growth factors. Figure 19.1 shows an
integrated, but not comprehensive, schematic of the major growth factor signaling cascades in
regulating protein synthesis and degradation. We discuss a few key elements of the PI3K/​Akt/​
mTOR pathway in the following subsections.

Insulin-​like growth factor 1

Both skeletal muscle and liver cells produce IGF1, though paracrine/​autocrine actions of skel-
etal muscle-​derived IGF1 appear more critical for skeletal muscle growth than circulating
levels (39). IGF1, upon stimulation of the IGF1 receptor, stimulates the activation of PI3K/​
Akt that promotes skeletal muscle growth through both the activation of protein synthesis
and inhibition of protein degradation pathways (40). IGF1-​stimulated activation of Akt indir-
ectly promotes protein synthesis through activation of mTOR, resulting in further activation

268
269

Genetic contributions to muscle strength

Figure 19.1 Molecular pathways involved in the regulation and balance of protein synthesis and protein
degradation. Skeletal muscle contraction (tension/​load) during resistance exercises stimulates both
protein synthesis and protein degradation. These processes are primarily modulated through the PI3K/​
Akt/​mTOR signaling axis. Growth factors, such as IGF1, increase Akt/​mTOR signaling in response to
contractile activity, which enhances protein production through regulation of translational regulators such
as eukaryotic translation initiation factor 4E-​binding protein (4E-​BP1). Protein degradation pathways
are controlled by forkhead box proteins (FoxO). Other growth factors, such as circulating or skeletal
muscle-​produced MSTN, inhibit this pathway, resulting in protein degradation. Inflammatory mediators,
also increased as a result of contractile activity, such a TNFα, may indirectly regulate PI3K/​Akt/​mTOR
signaling through c-​Jun N-​terminal kinases (JNK) and also stimulate protein degradation pathways
through p38 MAPK and NFkB signaling.

of the translational protein S6 kinase (S6K1) and by inactivating inhibitory eIF4E-​binding

proteins (4EBPs), allowing for activation of eukaryotic translation initiation factor 4E (eIF4E)
(40). Simultaneously, activation of Akt also results in inhibition of protein degradation pathways
through phosphorylation and repression of forkhead box O (FoxO) transcriptional factors (40).
The effect of resistance exercise on IGF1 expression is less clear. A number of studies have
resulted in conflicting results for IGF1 mRNA expression in skeletal muscle. Resistance exercise
studies have shown upregulation (41–​43), downregulation (44), and no changes (45) in IGF1
mRNA expression. However, transgenic animal models have provided direct evidence of its
role in promoting increased muscle mass and preservation of motoneurons during age-​related

269
270

M.D. Barberio et al.

loss, both of which would likely result in increased strength (46, 47). Furthermore, human
studies using recombinant human IGF1 administration demonstrated improved strength in
older females (48).Thus, IGF1 is certainly involved in the regulation of skeletal muscle mass and
strength, though the role of resistance exercise in increasing IGF1 expression and its subsequent
role in resistance exercise-​induced strength improvements is less clear.
One reason there may be conflicting responses between resistance exercise studies and IGF1
is the role of other interacting proteins and signaling pathways in its mechanism of action. First,
a binding protein, IGF-​binding protein 3 (IGFBP3) is responsible for extending the half-​life of
circulating IGF1 and IGF2. Increasing the half-​life would serve to augment their interactions
with surface receptors on skeletal muscle and cardiomyocytes (49, 50). However, other IGFBPs
are responsible for inhibiting its interactions with the surface receptor. Furthermore, IGF1-​
induced hypertrophy in both skeletal and cardiac tissue is largely influenced by signaling
pathways independent of the PI3K/​Akt/​mTOR axis previously described. Specifically, the pro-
tein phosphatase calcineurin is involved in the Ca2+-​dependent calcineurin signaling pathway,
which influences IGF1-​induced growth (51). Lastly, the timing and developmental phase of the
tissue may be critical as well. During fetal growth, IGF2 acts as the primary skeletal muscle tissue
growth factor and is later involved in regulation of skeletal muscle satellite cells (52).
As previously mentioned, expression levels of IGF1 following resistance exercise have provided
conflicting results, making interpretation of its role in resistance exercise-​induced strength changes
complicated. However, the nature of the IGF1 gene itself may be partially responsible for these
conflicting results. Splice variants of the IGF1 gene give rise to multiple mature protein products,
one of which is IGF-​1Ec, also known as mechanogrowth factor (MGF). MGF has been shown to
increase the proliferative capacity of skeletal muscle progenitor cells (53) as well as myotubes in vitro
(54), thus, providing a clear role in the development of muscle mass. However, animal and human
models of skeletal muscle overload using synergistic ablation (55) and resistance exercise (56) have
shown an age-​dependent upregulation of MGF mRNA.

Myostatin
Myostatin (MSTN), previously known as growth differentiation factor-​8 (GDF-​8), is a highly
conserved negative regulator of muscle development (57). This negative regulation has been
shown to occur through multiple signaling pathways: circulating MSTN activates activin
type IIa and IIb receptors (ACVR2A and ACVR2B) (58, 59) while skeletal muscle-​produced
MSTN has been shown to inhibit IGF1-​induced PI3K/​Akt/​mTOR signaling (57). Naturally
occurring loss of MSTN in humans (60), as well as in multiple animal models, results in signifi-
cantly more muscle mass in comparison to genetic wild types (61–​64). A number of studies have
explored circulating and tissues levels of MSTN in response to single or repeated bouts of resist-
ance training. Kim et al. (65) demonstrated reductions of MSTN mRNA in the vastus lateralis
following a single bout of resistance training in both old and young males and females. The
resistance exercise-​induced reduction was blunted in older females however. A study by Dalbo
et al. (66) showed reduced MSTN mRNA expression in old, but not young, men following
three bouts of resistance exercise. Furthermore, older participants also showed upregulation of
follistatin (FST), which is known to inhibit circulating MSTN activity by preventing binding
with the ACVR2A and ACVR2B receptors. Roth et al. (67) showed 9 weeks of heavy exer-
cise reduced vastus lateralis MSTN mRNA expression by 37% in men and women. Expression
levels in this study were not correlated to any measure of skeletal muscle mass or strength at
any time point. An extended review of MSTN’s role in muscle size is presented in Chapter 22
of this handbook.

270
271

Genetic contributions to muscle strength

Regulation of muscle strength and size by inflammatory factors

The inflammatory response to various exercise modalities plays a role in systemic and cel-
lular adaptation. The cellular and systemic mediators of this response are a class of proteins
called cytokines. While numerous, a key subset of cytokines has been of particular interest to
those trying to understand their role in protein accretion and degradation. Tumor necrosis
factor-​alpha (TNFα), interleukin (IL)-​6, and IL-​15 are some of the most well studied to date.
Independently these cytokines result in distinct cellular signaling cascades, but also work syner-
gistically and in opposition to each other as well as growth factors.

Tumor necrosis factor-​alpha

Originally identified as “cachectin” (due to elevated systemic levels in muscle wasting during
cancer cachexia) (68), TNFα is a proinflammatory mediator of skeletal muscle protein degrad-
ation (69).TNFα exerts its proteolytic actions through activation of multiple signaling pathways.
TNFα activates the classical proinflammatory protein complex nuclear factor kappa-​light-​chain-​
enhancer of activated B cells (NF-​κB) (70). Activation of TNFα membrane-​bound receptors,
TNFR1 and TNFR2, results in phosphorylation and degradation of the enzyme complex IκB
kinase (IKK). Degradation of this complex results in activation and nuclear translocation of NF-​
κB which signals for further production of proinflammatory cytokines, including TNFα, and
the proteolytic E3 ubiquitin-​ligase MuRF1 (71). Studies have also demonstrated a non-​NF-​κB
dependent proteolytic pathway activated by TNFα. P38 mitogen-​activated protein kinase (p38)
is activated through TNFα receptor signaling which results in activation of the proteolytic E3
ubiquitin ligase muscle atrophy F box (MAFbx/​Atrogin-​1) protein (72). Lastly, TNFα has been
demonstrated to inhibit both the expression and protein activity of growth factors such as IGF1
(73, 74).
The proteolytic actions of TNFα can be demonstrated or correlated at the cellular and
whole-​organ level. Blockade of TNFα signaling mitigates loss of muscle mass in models of
muscle wasting such as cancer or sepsis (75, 76). Li and Reid (77) also demonstrated the role
of NF-​κB-​mediated TNFα-​induced protein loss in C2C12 myoblasts. Although there is a high
degree of variability in the TNFα response to resistance training, circulating levels of TNFα have
been shown to decrease following strength training and have an inverse relationship to strength
gains (78). However, skeletal muscle expression of TNFα following resistance exercise has been
demonstrated to increase for up to 12 hours (79, 80). Louis et al. (80) also demonstrated that
elevations in TNFα resulted in upregulation of MuRF1 and MAFbx expression for up to 4
hours after resistance exercise. These studies, as well as observational studies (78, 81), clearly
demonstrate a relationship between TNFα and muscle mass and muscle strength in disease
states and during strength training-​induced adaptations. The early upregulation in expression
of TNFα and downstream effectors following a resistance exercise load indicates it is likely
involved in the exercise-​induced increase in protein degradation, which is outpaced by the pro-
tein synthesis resulting in a net protein accretion.

Interleukin-​6 and ciliary neurotrophic factor

Similar to TNFα, an inverse relationship between circulating levels of IL-​6 and measures of
muscle mass and muscle strength in older men and women has been demonstrated (78, 81, 82).
However, the role of IL-​6 in regulating the balance between protein synthesis and degradation

271
272

M.D. Barberio et al.

is less clear. Due to skeletal muscle being a primary site of synthesis and release in response
to exercise, IL-​6 has been termed a myokine (83). Traditionally a regulator of the acute phase
response to various cellular and systemic stressors, IL-​6 synthesis and release from skeletal muscle
is regulated by the contractile process (84). Once into circulation, IL-​6 can interact with either
the interleukin 6 receptor (IL-​6R) or the glycoprotein 130 (GP130) cytokine receptor, which
results in the activation of JAK/​STAT3 signaling and transcription. Multiple studies of resistance
exercise demonstrate an acute upregulation of IL-​6 expression, with one showing a biphasic
response over 24 hours (79, 80). Circulating levels of IL-​6 are highly variable between individ-
uals and are highly influenced by age and health status. Furthermore, circulating levels following
exercise are highly dependent on exercise modality, intensity, and duration (83). While there is
evidence that IL-​6 may interfere with the anabolic actions of IGF1 (85), it is possible that the
IL-​6 relationship to protein balance is indirect, likely through a negative regulation of TNFα.
Ciliary neurotrophic factor (CNTF) shares a similar molecular structure with IL-​6 and is
able to bind to the GP130 membrane receptor, resulting in activation of the STAT3 and AMPK
signaling pathways (86). Interest in CNTF as a regulator of muscle mass and strength is based
on its ability to promote increased cellular growth and survival of neurons (87). Multiple resist-
ance training studies (88, 89) have identified a positive correlation between CNTF expression
in skeletal muscle and resistance training outcomes. Furthermore, these studies also identified an
age effect on CNTF expression, with older individuals expressing less. However, use of recom-
binant human CNTF for treatment of amyotrophic lateral sclerosis (ALS), a neurodegenerative
disease, did not attenuate loss of isometric grip strength (90).

Interleukin-​15
IL-​15 was originally identified as a growth factor after it was demonstrated to increase myoblast
proliferation and muscle-​specific myosin heavy chain composition (91). Further in vitro (92)
and overexpression (93) studies supported the role of IL-​15 in promoting increased protein
accretion as well as preventing protein degradation. More recent studies in mice have altered
this view of IL-​15, instead focusing on its role in mitochondrial biogenesis and fiber type com-
position. Overexpression of IL-​15 results in a more oxidative skeletal muscle phenotype (94),
while loss of the IL-​15 receptor α subunit of the IL-​15 plasma membrane receptor increased
endurance capacity and resistance to fatigue (95).

Limitations, practical applications, and future directions

In this chapter, we have provided a primer for understanding the muscular strength trait at
baseline and following training and have explored molecular pathways that can contribute to
variations in strength within a population. We noted that while the heritability of strength is
moderate to high, genetic contributions can be quite varied due to the complex nature of the
strength trait. We discussed current limitations to identifying genetic associations with strength,
including limited sample sizes and the introduction of variability during training. With a few
exceptions (such as ACTN3 and MSTN, discussed further in their own chapters), both targeted
and global association studies have yet to identify strong single gene–​strength interactions, and
it is likely that multivariant modeling will be needed to explain currently missing strength her-
itability. This will be enabled by efforts such as MoTrPAC and other large consortia projects,
though results from these studies are likely years away. In the meantime, there is strong evidence
that pathway modification of underlying strength components such as muscle size accounts for
significant variability in baseline strength and in functional gains with training. Understanding

272
273

Genetic contributions to muscle strength

these to a greater extent should help develop training programs that optimize gains among a
variety of different people.

References
1. Campos, G.E., et al., Muscular adaptations in response to three different resistance-​
training regimens: specificity of repetition maximum training zones. Eur J Appl Physiol, 2002.
88(1–​2): p. 50–​60.
2. Chilibeck, P.D., et al.,Twenty weeks of weight training increases lean tissue mass but not bone min-
eral mass or density in healthy, active young women. Can J Physiol Pharmacol, 1996. 74(10): p. 1180–​5.
3. Peterson, M.D., A. Sen, and P.M. Gordon, Influence of resistance exercise on lean body mass in
aging adults: a meta-​analysis. Med Sci Sports Exerc, 2011. 43(2): p. 249–​58.
4. Harman, E., Biomechanics of resistance training, in Essentials of Strength Training and Conditioning,
T.R. Baechle and R.W. Earle, Editors. 2008, Human Kinetics.
5. Knuttgen, H.G. and P.V. Komi, Basic definitions for exercise, in Strength and Power in Sport, P.V.
Komi, Editor. 1992, Blackwell Science.
6. Siff, M.C., Supertraining. 5th ed. 2000, Supertraining Institute.
7. Wilkie, D.R., The relation between force and velocity in human muscle. J Physiol, 1949. 110(3–​4): p.
249–​80.
8. Goldspink, G., Cellular and molecular aspects of adaptation in skeletal muscle, in Strength and Power
in Sport, P.V. Komi, Editor. 1992, Blackwell Science.
9. Bickel, C.S., et al., Time course of molecular responses of human skeletal muscle to acute bouts of
resistance exercise. J Appl Physiol (1985), 2005. 98(2): p. 482–​8.
10. Caiozzo, V.J., The muscular system: structural and functional plasticity, in ACSM’s Advanced Exercise
Physiology, P.A. Farrell, M.J. Joyner, and V.J. Caiozzo, Editors. 2012, Lippincott Williams & Wilkens.
11. Billeter, R. and H. Hoppeler, Muscular basis of strength, in Strength and Power in Sport, P.V. Komi,
Editor. 1992, Blackwell Science.
12. Saltin, B. and P.D. Gollnick, Skeletal muscle adaptability: significance for metabolism and perform-
ance, in Handbook of Physiology: Skeletal Muscle, L.D. Peachey, R.H. Adrian, and R.S. Geiger, Editors.
1983, Williams & Wilkins.
13. Simoneau, J.A. and C. Bouchard, Genetic determinism of fiber type proportion in human skeletal
muscle. FASEB J, 1995. 9(11): p. 1091–​5.
14. Simoneau, J.A. and C. Bouchard, Human variation in skeletal muscle fiber-​type proportion and
enzyme activities. Am J Physiol, 1989. 257(4 Pt 1): p. E567–​72.
15. Simoneau, J.A., et al., Inheritance of human skeletal muscle and anaerobic capacity adaptation to
high-​intensity intermittent training. Int J Sports Med, 1986. 7(3): p. 167–​71.
16. Prud’homme, D., et al., Sensitivity of maximal aerobic power to training is genotype-​dependent.
Med Sci Sports Exerc, 1984. 16(5): p. 489–​93.
17. Moritani, T. and H.A. deVries, Neural factors versus hypertrophy in the time course of muscle
strength gain. Am J Phys Med, 1979. 58(3): p. 115–​30.
18. Moritani, T., Time course of adaptations during strength and power training, in Strength and Power
in Sport, P.V. Komi, Editor. 1992, Blackwell Science.
19. Sale, D.G., Neural adaptations to strength training, in Strength and Power in Sport, P.V. Komi, Editor.
1992, Blackwell Science.
20. Aagaard, P., Training-​induced changes in neural function. Exerc Sport Sci Rev, 2003. 31(2): p. 61–​7.
21. DeLorme, T.L. and A.L. Watkins, Progressive Resistance Exercise. 1951, Appleton Century Inc.
22. DeVries, H.A., “Efficiency of electrical activity” as a physiological measure of the functional state of
muscle tissue. Am J Phys Med, 1968. 47(1): p. 10–​22.
23. Bowers, L., Effects of autosuggested muscle contraction on muscular strength and size. Res Q, 1966.
37(3): p. 302–​12.
24. Hellebrandt, F.A., A.M. Parrish, and S.J. Houtz, The influence of unilateral exercise on the
contralateral limb. Arch Phys Med Rehabil, 1947. 28(2): p. 76–​85.
25. Coleman, E.A., Effect of unilateral isometric and isotonic contractions on the strength of the
contralateral limb. Res Q, 1969. 40: p. 490–​495.
26. Pescatello, L.S., et al., Highlights from the functional single nucleotide polymorphisms associated
with human muscle size and strength or FAMuSS study. Biomed Res Int, 2013. 2013: p. 643575.

273
274

M.D. Barberio et al.

27. Thompson, P.D., et al., Functional polymorphisms associated with human muscle size and strength.
Med Sci Sports Exerc, 2004. 36(7): p. 1132–​9.
28. Hubal, M.J., et al.,Variability in muscle size and strength gain after unilateral resistance training. Med
Sci Sports Exerc, 2005. 37(6): p. 964–​72.
29. Tsika, R., The muscular system: the control of muscle mass, in ACSM’s Advanced Exercise Physiology,
P.A. Farrell, M.J. Joyner, and V.J. Caiozzo, Editors. 2012, Lippincot Williams and Wilkens.
30. Biolo, G., et al., Increased rates of muscle protein turnover and amino acid transport after resistance
exercise in humans. Am J Physiol, 1995. 268(3 Pt 1): p. E514–​20.
31. Allen, D.L., R.R. Roy, and V.R. Edgerton, Myonuclear domains in muscle adaptation and disease.
Muscle Nerve, 1999. 22(10): p. 1350–​60.
32. Petrella, J.K., et al., Potent myofiber hypertrophy during resistance training in humans is associated
with satellite cell-​mediated myonuclear addition: a cluster analysis. J Appl Physiol (1985), 2008.
104(6): p. 1736–​42.
33. Verdijk, L.B., et al., Skeletal muscle hypertrophy following resistance training is accompanied by a
fiber type-​specific increase in satellite cell content in elderly men. J Gerontol A Biol Sci Med Sci, 2009.
64(3): p. 332–​9.
34. Zempo, H., et al., Heritability estimates of muscle strength-​related phenotypes: A systematic review
and meta-​analysis. Scand J Med Sci Sports, 2017. 27(12): p. 1537–​1546.
35. Rankinen, T., et al., Advances in exercise, fitness, and performance genomics. Med Sci Sports Exerc,
2010. 42(5): p. 835–​46.
36. Venezia, A.C. and S.M. Roth, Recent research in the genetics of exercise training adaptation. Med
Sport Sci, 2016. 61: p. 29–​40.
37. Bouchard, C., Exercise genomics –​a paradigm shift is needed: a commentary. Br J Sports Med, 2015.
49(23): p. 1492–​6.
38. Willems, S.M., et al., Large-​scale GWAS identifies multiple loci for hand grip strength providing
biological insights into muscular fitness. Nat Commun, 2017. 8: p. 16015.
39. Adams, G.R., F. Haddad, and K.M. Baldwin, Time course of changes in markers of myogenesis
in overloaded rat skeletal muscles. J Appl Physiol (1985), 1999. 87(5): p. 1705–​12.
40. Schiaffino, S. and C. Mammucari, Regulation of skeletal muscle growth by the IGF1-​Akt/​PKB
pathway: insights from genetic models. Skelet Muscle, 2011. 1(1): p. 4.
41. Bamman, M.M., et al., Cluster analysis tests the importance of myogenic gene expression during
myofiber hypertrophy in humans. J Appl Physiol (1985), 2007. 102(6): p. 2232–​9.
42. Deldicque, L., et al., Increased IGF mRNA in human skeletal muscle after creatine supplementation.
Med Sci Sports Exerc, 2005. 37(5): p. 731–​6.
43. Greig, C.A., et al., Skeletal muscle IGF-​I isoform expression in healthy women after isometric exer-
cise. Growth Horm IGF Res, 2006. 16(5–​6): p. 373–​6.
44. Psilander, N., R. Damsgaard, and H. Pilegaard, Resistance exercise alters MRF and IGF-​I
mRNA content in human skeletal muscle. J Appl Physiol (1985), 2003. 95(3): p. 1038–​44.
45. Coffey, V.G. and J.A. Hawley,The molecular bases of training adaptation. Sports Med, 2007. 37(9): p.
737–​63.
46. Messi, M.L. and O. Delbono, Target-​derived trophic effect on skeletal muscle innervation in sen-
escent mice. J Neurosci, 2003. 23(4): p. 1351–​9.
47. Gonzalez, E., et al., Insulin-​like growth factor-​1 prevents age-​related decrease in specific force and
intracellular Ca2+ in single intact muscle fibres from transgenic mice. J Physiol, 2003. 552(Pt 3): p.
833–​44.
48. Boonen, S., et al., Musculoskeletal effects of the recombinant human IGF-​I/​IGF binding protein-​3
complex in osteoporotic patients with proximal femoral fracture: a double-​blind, placebo-​controlled
pilot study. J Clin Endocrinol Metab, 2002. 87(4): p. 1593–​9.
49. Ferry, R.J., Jr., et al., Cellular actions of insulin-​like growth factor binding proteins. Horm Metab Res,
1999. 31(2–​3): p. 192–​202.
50. Jones, J.I. and D.R. Clemmons, Insulin-​like growth factors and their binding proteins: biological
actions. Endocr Rev, 1995. 16(1): p. 3–​34.
51. Musaro, A., et al., IGF-​1 induces skeletal myocyte hypertrophy through calcineurin in association
with GATA-​2 and NF-​ATc1. Nature, 1999. 400(6744): p. 581–​5.
52. Goldberg, A.L., et al., Mechanism of work-​induced hypertrophy of skeletal muscle. Med Sci Sports,
1975. 7(3): p. 185–​98.

274
275

Genetic contributions to muscle strength

53. Kandalla, P.K., et al., Mechano growth factor E peptide (MGF-​E), derived from an isoform of IGF-​
1, activates human muscle progenitor cells and induces an increase in their fusion potential at different
ages. Mech Ageing Dev, 2011. 132(4): p. 154–​62.
54. Philippou, A., et al., Expression of IGF-​ 1 isoforms after exercise-​induced muscle damage in
humans: characterization of the MGF E peptide actions in vitro. In Vivo, 2009. 23(4): p. 567–​75.
55. Owino, V., S.Y. Yang, and G. Goldspink, Age-​related loss of skeletal muscle function and the
inability to express the autocrine form of insulin-​like growth factor-​1 (MGF) in response to mech-
anical overload. FEBS Lett, 2001. 505(2): p. 259–​63.
56. Hameed, M., et al., Expression of IGF-​I splice variants in young and old human skeletal muscle after
high resistance exercise. J Physiol, 2003. 547(Pt 1): p. 247–​54.
57. Morissette, M.R., et al., Myostatin inhibits IGF-​I-​induced myotube hypertrophy through Akt. Am
J Physiol Cell Physiol, 2009. 297(5): p. C1124–​32.
58. Lee, S.J., et al., Regulation of muscle growth by multiple ligands signaling through activin type II
receptors. Proc Natl Acad Sci U S A, 2005. 102(50): p. 18117–​22.
59. Lee, S.J. and A.C. McPherron, Regulation of myostatin activity and muscle growth. Proc Natl Acad
Sci U S A, 2001. 98(16): p. 9306–​11.
60. Schuelke, M., et al., Myostatin mutation associated with gross muscle hypertrophy in a child. N Engl
J Med, 2004. 350(26): p. 2682–​8.
61. Clop, A., et al., A mutation creating a potential illegitimate microRNA target site in the myostatin
gene affects muscularity in sheep. Nat Genet, 2006. 38(7): p. 813–​8.
62. McPherron, A.C. and S.J. Lee, Double muscling in cattle due to mutations in the myostatin gene.
Proc Natl Acad Sci U S A, 1997. 94(23): p. 12457–​61.
63. Mosher, D.S., et al., A mutation in the myostatin gene increases muscle mass and enhances racing
performance in heterozygote dogs. PLoS Genet, 2007. 3(5): p. e79.
64. Szabo, G., et al., A deletion in the myostatin gene causes the compact (Cmpt) hypermuscular muta-
tion in mice. Mamm Genome, 1998. 9(8): p. 671–​2.
65. Kim, J.S., J.M. Cross, and M.M. Bamman, Impact of resistance loading on myostatin expression
and cell cycle regulation in young and older men and women. Am J Physiol Endocrinol Metab, 2005.
288(6): p. E1110–​9.
66. Dalbo, V.J., et al., Acute loading and aging effects on myostatin pathway biomarkers in human
skeletal muscle after three sequential bouts of resistance exercise. J Gerontol A Biol Sci Med Sci, 2011.
66(8): p. 855–​65.
67. Roth, S.M., et al., Myostatin gene expression is reduced in humans with heavy-​resistance strength
training: a brief communication. Exp Biol Med (Maywood), 2003. 228(6): p. 706–​9.
68. Beutler, B., et al., Purification of cachectin, a lipoprotein lipase-​suppressing hormone secreted by
endotoxin-​induced RAW 264.7 cells. J Exp Med, 1985. 161(5): p. 984–​95.
69. Argiles, J.M. and F.J. Lopez-​Soriano,The role of cytokines in cancer cachexia. Med Res Rev, 1999.
19(3): p. 223–​48.
70. von Haehling, S., et al., Cachexia: a therapeutic approach beyond cytokine antagonism. Int J Cardiol,
2002. 85(1): p. 173–​83.
71. Bodine, S.C., et al., Identification of ubiquitin ligases required for skeletal muscle atrophy. Science,
2001. 294(5547): p. 1704–​8.
72. Li, Y.P., et al.,TNF-​alpha acts via p38 MAPK to stimulate expression of the ubiquitin ligase atrogin1/​
MAFbx in skeletal muscle. FASEB J, 2005. 19(3): p. 362–​70.
73. Fernandez-​Celemin, L., et al., Inhibition of muscle insulin-​like growth factor I expression by
tumor necrosis factor-​alpha. Am J Physiol Endocrinol Metab, 2002. 283(6): p. E1279–​90.
74. Langen, R.C., et al., Inflammatory cytokines inhibit myogenic differentiation through activation of
nuclear factor-​kappaB. FASEB J, 2001. 15(7): p. 1169–​80.
75. Breuille, D., et al., Pentoxifylline decreases body weight loss and muscle protein wasting characteristics
of sepsis. Am J Physiol, 1993. 265(4 Pt 1): p. E660–​6.
76. Costelli, P., et al., Tumor necrosis factor-​alpha mediates changes in tissue protein turnover in a rat
cancer cachexia model. J Clin Invest , 1993. 92(6): p. 2783–​9.
77. Li, Y.P. and M.B. Reid, NF-​kappaB mediates the protein loss induced by TNF-​alpha in differentiated
skeletal muscle myotubes. Am J Physiol Regul Integr Comp Physiol, 2000. 279(4): p. R1165–​70.
78. Schaap, L.A., et al., Higher inflammatory marker levels in older persons: associations with 5-​year
change in muscle mass and muscle strength. J Gerontol A Biol Sci Med Sci, 2009. 64(11): p. 1183–​9.

275
276

M.D. Barberio et al.

79. Buford, T.W., M.B. Cooke, and D.S. Willoughby, Resistance exercise-​induced changes of inflam-
matory gene expression within human skeletal muscle. Eur J Appl Physiol, 2009. 107(4): p. 463–​71.
80. Louis, E., et al.,Time course of proteolytic, cytokine, and myostatin gene expression after acute exer-
cise in human skeletal muscle. J Appl Physiol (1985), 2007. 103(5): p. 1744–​51.
81. Visser, M., et al., Relationship of interleukin-​6 and tumor necrosis factor-​alpha with muscle mass
and muscle strength in elderly men and women: the Health ABC Study. J Gerontol A Biol Sci Med Sci,
2002. 57(5): p. M326–​32.
82. Schaap, L.A., et al., Inflammatory markers and loss of muscle mass (sarcopenia) and strength. Am J
Med, 2006. 119(6): p. 526 e9–​17.
83. Pedersen, B.K., et al., Role of myokines in exercise and metabolism. J Appl Physiol (1985), 2007.
103(3): p. 1093–​8.
84. Pedersen, B.K. and M.A. Febbraio, Muscle as an endocrine organ: focus on muscle-​derived
interleukin-​6. Physiol Rev, 2008. 88(4): p. 1379–​406.
85. Lazarus, D.D., L.L. Moldawer, and S.F. Lowry, Insulin-​like growth factor-​1 activity is inhibited
by interleukin-​1 alpha, tumor necrosis factor-​alpha, and interleukin-​6. Lymphokine Cytokine Res,
1993. 12(4): p. 219–​23.
86. Watt, M.J., et al., CNTF reverses obesity-​induced insulin resistance by activating skeletal muscle
AMPK. Nat Med, 2006. 12(5): p. 541–​8.
87. Ip, N.Y. and G.D. Yancopoulos, Ciliary neurotrophic factor and its receptor complex. Prog Growth
Factor Res, 1992. 4(2): p. 139–​55.
88. Dennis, R.A., et al., Aging alters gene expression of growth and remodeling factors in human skel-
etal muscle both at rest and in response to acute resistance exercise. Physiol Genomics, 2008. 32(3): p.
393–​400.
89. Dennis, R.A., et al., Muscle expression of genes associated with inflammation, growth, and remod-
eling is strongly correlated in older adults with resistance training outcomes. Physiol Genomics, 2009.
38(2): p. 169–​75.
90. Group, A.C.T.S., A double-​blind placebo-​controlled clinical trial of subcutaneous recombinant
human ciliary neurotrophic factor (rHCNTF) in amyotrophic lateral sclerosis. Neurology, 1996.
46(5): p. 1244–​9.
91. Quinn, L.S., K.L. Haugk, and K.H. Grabstein, Interleukin-​15: a novel anabolic cytokine for skel-
etal muscle. Endocrinology, 1995. 136(8): p. 3669–​72.
92. Furmanczyk, P.S. and L.S. Quinn, Interleukin-​15 increases myosin accretion in human skeletal
myogenic cultures. Cell Biol Int, 2003. 27(10): p. 845–​51.
93. Quinn, L.S., et al., Overexpression of interleukin-​ 15 induces skeletal muscle hypertrophy in
vitro: implications for treatment of muscle wasting disorders. Exp Cell Res, 2002. 280(1): p. 55–​63.
94. Quinn, L.S., et al., IL-​15 overexpression promotes endurance, oxidative energy metabolism, and
muscle PPARdelta, SIRT1, PGC-​1alpha, and PGC-​1beta expression in male mice. Endocrinology,
2013. 154(1): p. 232–​45.
95. Pistilli, E.E., et al., Loss of IL-​15 receptor alpha alters the endurance, fatigability, and metabolic
characteristics of mouse fast skeletal muscles. J Clin Invest, 2011. 121(8): p. 3120–​32.

276
277

20
GENETIC CONTRIBUTIONS TO
SKELETAL MUSCLE SIZE
Philip J. Atherton, Jessica Cegielski, and Daniel J. Wilkinson

Resistance exercise training and muscle hypertrophy

The main physiological adaptive feature of resistance exercise training (RET) is increased
muscle mass, which manifests as an increase in the dimensions of whole muscle groups as a
result of growth of individual muscles.These gross physiological changes in skeletal muscle mass
are most commonly determined using imaging techniques such as dual-​energy X-​ray absorpti-
ometry (DXA), computed tomography (CT), magnetic resistance imaging (MRI), or ultrason-
ography (43). Muscle hypertrophy can also be measured on a single-​fiber level by estimation of
individual myofiber cross-​sectional area in biopsy samples using histological techniques coupled
to imaging software analysis to determine hypertrophy on a cell-​by-​cell level (20). It is believed
that postnatal growth of skeletal muscle occurs via the process of hypertrophy, which represents
an increase in cross-​sectional area of existing post-​mitotic muscle cells, rather than through fiber
splitting or hyperplasia (61).
RET-​induced hypertrophy is reliant on a number of different components and is regulated
through complex and integrated gene–​environment interactions. Moreover, using standard
muscle loading regimens, approximately 20% of adults do not demonstrate measurable increases
in muscle mass (6, 23, 74) even after months of resistance training; this intersubject variability
is important to keep in mind when considering the various potential mediators of skeletal
muscle cell hypertrophy (ensuing sections). Muscle mass in adults is maintained through the
diurnal pattern of building muscle proteins via muscle protein synthesis (MPS) following
periods of feeding, with dietary amino acids (AAs) replenishing muscle proteins that are broken
down via muscle protein breakdown (MPB) during periods of fasting. Under normal healthy
circumstances, diurnal MPS matches MPB and muscle mass is maintained. In contrast, in
chronic periods of growth (e.g., adolescence) or RET-​induced hypertrophy, MPS >MPB, while
in situations of skeletal muscle wasting (e.g., catabolic disease or ageing), MPS <MPB (15).

Resistance exercise training-​induced effects on muscle protein

synthesis and muscle protein breakdown
In the immediate period following a single bout of resistance exercise (RE), MPS is markedly
increased approximately two-​to fourfold above baseline rates (at rest, human muscle proteins

277
278

P.J. Atherton et al.

turnover at approximately 0.05%/​hour or 1.2%/​day) (55). The magnitude of increase in MPS

in response to RE displays a dose-​dependent increase in MPS, reaching near maximal rates
at approximately 60% of one-​repetition maximum (1RM) (55), beyond which no additional
stimulation is observed. Despite this observation, this does not mean exercising at lower inten-
sities is of no benefit (when it comes to muscle size and hypertrophy that is), since increases in
acute MPS at 30% 1RM are equivalent to those at 90% 1RM, but only when RE is performed
to failure (17). This is likely due to increased type II muscle fiber recruitment through fatiguing
contractions (Hennemen’s size principle) (42), resulting in maximal fiber recruitment and hence
maximal recruitment and thereby anabolic stimulation of that muscle.
It is the accumulation of postexercise increases in MPS that, over a period of RET, ultim-
ately creates a greater sustained positive net protein balance (i.e., MPS >MPB) culminating in
muscle growth or hypertrophy (13, 14). Few studies have addressed the temporal nature of MPS
response throughout a RET program due to traditional measurement approaches requiring
constant infusions of AA tracers, limiting studies to hours rather than days or weeks. However,
recent advances using the D2O stable isotope tracer have allowed further understanding of
the cumulative, temporal nature of MPS responses over progressive RET programs. What can
be observed from these studies is that the major hypertrophic responses to (unaccustomed)
RET occurs during the early phase of RET, with significant increases in D2O-​derived MPS
only observed over 0–​3 weeks of a 6-​week resistance exercise program. No significant increase
was observed between 3 and 6 weeks (13). These conclusions are supported by a number of
RET studies showing pronounced increases in muscle mass using CT, DXA, and ultrason-
ography, during the early phases of RET, which curtail within the later weeks (9, 25, 85).
Despite the idea of “early hypertrophy” still being considered a somewhat controversial idea
by some, based on findings from acute MPS tracer studies, conceptually it makes perfect sense.
Indeed, early acute stable isotope tracer studies highlighted that following a single bout of RET,
MPS responses were more pronounced in individuals unaccustomed to RET as compared with
those accustomed to regular RET (77). Moreover, following unilateral RET, MPS responses are
reduced within the trained versus untrained leg (75), with this training-​induced attenuation of
MPS beginning following only a few RET bouts (22, 89, 94). These findings support the idea
of rapid remodeling and growth in the early phases due to heightened MPS responses to RE,
followed by a period of sustained maintenance and neuromuscular adaptations.

Nutrient interactions with resistance exercise training for

regulation of muscle size
The impact of nutrients on muscle protein turnover is now very well established (see 5, 15, 24,
65, 71). In terms of muscle protein turnover, it is the essential amino acid (EAA) components
of a mixed meal that are the primary driver behind increases in MPS in the postprandial phase,
reverting the muscle to a net positive balance and replenishing the protein lost during the pre-
vious fasted/​postabsorptive period (5, 67, 86, 87). In particular, the branched-​chain AA leucine
has been found to have potent anabolic properties, acting as both a substrate and a signal for
muscle protein synthesis (12, 40, 95, 96). In terms of MPB, the release of insulin –​either through
the carbohydrate content of a mixed meal, or via AA-​mediated insulin release (36) –​helps to
suppress MPB, augmenting the shift to positive protein balance further. However, the acute
response to nutrition is finite; MPS peaks at about 1.5 hours following feeding and then rapidly
returns to baseline values, once the fasted loss of protein has been replenished in the fed state,
a term known as “muscle full” (4, 66). Once this is complete, these energetically demanding
metabolic processes can be shut down until required again. While muscle size is dependent on

278
279

Genetic contributions to skeletal muscles

adequate nutrition, nutrition alone does not regulate muscle size, growth, or hypertrophy –​at
least in adult muscle. It is the important interaction between nutrition and exercise that is key
to regulating growth.
It should also be noted, that RE in the absence of nutrition is actually a significant catabolic
stressor. Following an acute bout of RE, MPB increases by as much as 50%, which, despite
concomitant increases in MPS in the postabsorptive state (i.e., absence of nutrition), leads to
an overall negative net balance (54, 55, 76). Crucially, nutrition plays a significant facilitating
role in RE-​induced anabolism, helping to suppress the exercise-​induced increase in MPB
(via insulin-​mediated mechanism discussed above) while also potentiating the RE-​induced
increase in MPS (8). Thus, in terms of protein turnover, while postexercise increases in MPS
are the major determinants of adaptive hypertrophy, the potentiation of this MPS response
through interactions with dietary protein intake rich in EAAs, alongside the associated
attenuation of MPB, is necessary to maximize positive net protein balance and muscle gains.
Following RE, MPS can remain elevated for up to, and potentially beyond, 24 hours, essen-
tially “sensitizing” the muscle to future anabolic stimuli, such as food intake (16, 64, 76). This
interaction between nutrient intake and RE has led to the investigation of whether there
may be an optimal dosage or timing of nutrient intake allied to RE-​induced anabolism (71).
It appears that, as with protein intake alone, there is a “muscle-​full” effect, whereby MPS
response is near maximal following the provision of between 0.25 and 0.4 g/​kg (~20–​30g
of protein for the average size male) (69), with higher doses providing no additional benefit
and only leading to increased AA oxidation (69, 98). Nonetheless, due to the extended ana-
bolic window provided following RE, many have investigated whether the timing of pro-
tein ingestion is an important consideration; however, little evidence suggests that anything
other than providing protein in the post-​RE period close to the cessation of exercise is of
any additional benefit (68).

Intramuscular sensing and signaling regulating resistance exercise

training-​induced muscle hypertrophy
Of all the signaling pathways associated with RET response, the mechanistic target of rapamycin,
complex 1 (mTORC1: composed of the genes mTOR, raptor, PRAS40, Deptor, and LST8) is
considered a “master” determinant of the regulation of RET-​induced hypertrophy (33). For
instance, experiments where rapamycin, an inhibitor of mTORC1, was administered to humans
revealed a complete blockade of RE-​mediated increases in MPS (28).The notion that activation
of mTORC1 might represent a central node communicating RE to MPS can be reconciled by
the number of substrates associated with mRNA translation. Indeed, activation of mTORC1
is associated with the phosphorylation of multiple translational initiation factors substrates
including 4E-​binding protein (4EBP1), ribosomal protein S6 kinase (p70S6K1), and eukary-
otic initiation factors 4 G/​A/​B (eIF4G/​A/​B), that facilitate formation of an eIF3F scaffold and
assembly of a “competent” 48S pre-​initiation complex. In a parallel pathway, activation of the
guanine exchange factor, eukaryotic initiation factor 2B (eIF2B), shuttles the initiator tRNA
(Met-​tRNAi) to the ribosome during formation of the 48S pre-​initiation complex, thereby
promoting “global” MPS (mRNA translation) and thus coordinately enhancing translational
efficiency (see 45 for control of translation). Practically all of these factors have been shown to
be stimulated in response to RE, and many are used as a proxy for mTORC1 signaling (e.g.,
p70S6K1) since kinase activity of mTORC1 is a complex combination of cellular localization,
affinity to its multiple binding partners, and phosphorylation events.

279
280

P.J. Atherton et al.

Upstream signaling to mTORC1 in response to resistance exercise

While the mTORC1 signaling pathway is thought to be an affirmed central regulator of
MPS and hypertrophy, the proximal mechanisms involved in the activation of mTORC1 by
RE remain incompletely defined. In the most basic sense, “upstream” signals triggered by RE
cause the activation of intracellular signaling pathways that are responsible for modulating
post-​translational control of MPS and also that of muscle mRNA expression. Like any form
of physical activity, RE triggers complex physiochemical events (i.e., mechanotransduction
(33), endocrine (41), and auto/​paracrine (7) responses) which combine to determine the
net adaptive responses. The genetic bases of neurohormonal responses to RET are fur-
ther discussed in Chapter 21. Much of the early preclinical (10) and cell culture work
(78) implicated insulin-​like growth factor (IGF) canonical signaling in RE-​mediated alter-
ations in MPS and hypertrophy. Increases in the expression of IGF-​1 and its splice variants
(mechanogrowth factor, liver IGF-​1 isoforms) bind the insulin-​like growth factor 1 receptor
(IGF-​1R), which in turn yields activation of a classical insulin-​mediated PI3K-​protein
kinase B (AKT)–​mTORC1 signaling pathway. However, it is now thought that canonical
IGF-​1R–​AKT–​mTOR signaling is an oversimplification of the complex regulation of RE-​
induced muscle hypertrophy. An example of this can be found in one elegant study, where
adult human volunteers performed RE either in arm muscles alone or concurrently in arm
and leg muscles; the rationale being to create either a high (i.e., arm and leg exercise) or
a low (i.e., arm exercise alone) systemic hormone milieu based on differences in recruit-
ment of absolute muscle mass. Despite markedly different systemic concentrations of growth
hormone, testosterone and IGF-​1 between groups, there was no difference in mTORC1-​
signaling or in MPS –​or critically –​chronic adaptations to training in terms of muscle mass/​
strength gains (92, 93). These data indicate that modulating the systemic concentrations of
IGF-​1 or testosterone and growth hormone (GH, the latter of which acts via inducing IGF-​
1 expression rather than having any direct effect on MPS in adult humans) (27) within the
physiological range might not play a role in adaptive hypertrophy and challenges the classic
view of canonical insulin-​signaling pathway. This does not however preclude that IGF-​1
regulates AKT-​mTOR signaling via “local” autoparacrine signaling mechanisms undetect-
able by measuring serum abundance of IGF-​1. That being said, IGF-​1R knockout did not
attenuate functional loading-​induced muscle hypertrophy in animal models (88) (synergist
ablation model), once again suggesting that neither IGF-​1 canonical signaling, nor perhaps,
systemic hormone milieu can explain RE-​induced hypertrophy (i.e., there must be other
mechanisms at play).
The absence of accord over the role of canonical signaling from exogenous (to the muscle)
IGF-​1 (or other circulating hormones) has led to new hypotheses surrounding the mechan-
istic basis of RET-​induced hypertrophy with greater focus on it being a process both arising
from and being regulated within the muscle.The intrinsic mechanisms by which this is thought
to occur is via mechanotransduction; the process of converting mechanical signals sensed in
response to cellular movement into molecular signals. One mechanoresponsive target was
recently highlighted, where activation of phospholipase D (PLD) enzymes increases produc-
tion of the lipid second messenger phosphatidic acid (PA) in a mechanosensitive manner.
Mechanistically, it has been demonstrated that PA signaling is upstream of contraction-​induced
activation of mTOR, since pharmacological inhibition of PLD effectively ablated activation of
mTOR in response to contractions (73). While pharmacological inhibitors are rarely entirely
selective, this perhaps represents at least one line of evidence for a mechanism by which muscle
can adapt independently of systemic hormones and/​or a canonical IGF-​1 pathway.

280
281

Genetic contributions to skeletal muscles

Another possible mechanosensitive pathway is that of the muscle attachment, or focal adhe-
sion complexes: macromolecular structures situated in the sarcolemma of muscle fibers, that
link the extracellular matrix to the cytoplasmic cytoskeleton, consisting of a variety of extra-
cellular matrix receptors/​integrins and intracellular cytoskeletal and signaling molecules (52).
Indeed, interactions of extracellular matrix proteins with integrin receptors stimulate intracel-
lular signaling pathways important in cell growth and migration in adult skeletal muscle (84).
Moreover, activation of integrin receptors appears to be a common feature of muscle remod-
eling in response to a variety of conditions including endurance training and muscular dys-
trophy (90). Focal adhesion kinase (FAK) is a nonreceptor tyrosine kinase that localizes to focal
adhesion complexes and is thought to represents a key component of integrin-​mediated sig-
naling (19). Engagement of integrin receptors induces phosphorylation of FAK at Tyr397, which
correlates with its activation (18) and a growing body of evidence has associated FAK activation
with the hypertrophic response to mechanical stress in skeletal muscle (29). Indeed, expression
patterns of FAK have been reported to be load-​dependent, i.e., phosphorylation of FAK was
lowered following hindlimb suspension in rodents (34), immobilization in humans (11), and
increased in models of overload (31, 35) and following RE training in humans (97). Finally, local
overexpression of FAK (FAK pLKO.1 plasmid electroporation) in vivo in rodents was shown to
stimulate muscle hypertrophy (52), suggesting FAK is a genuine mechanosensitive component
of muscle hypertrophy. This notion is further supported by recent work demonstrating that
FAK is associated with skeletal muscle remodeling in response to RET in humans, with its acti-
vation being associated with areas of most significant RET-​induced hypertrophy (32).

PGC1 and myostatin as regulators of RET-​induced muscle hypertrophy

The identification of splice variants of peroxisome proliferator-​ activated receptor gamma
coactivator 1-​alpha (PGC1α: α1, α2, α3, and α4,) in muscle has led to the suggestion that
PGC1α4 is critical for the regulation of muscle hypertrophy through upregulation of IGF-​1
signaling and inhibition of myostatin (80) in mice. However, translation of these findings were
questioned in a study which reported no differential upregulation of PGC1α splice variants
in response to acute exercise of contrasting type and effects on muscle hypertrophy (resistance
vs. endurance) (99). Moreover, exercise-​induced expression of PGC1α variants following 5
weeks of unilateral RET showed no correlation with increases in muscle size and strength (57),
casting doubt upon a role of such splice variants regulating human muscle growth in responses
to RET. In terms of other key purported regulators, myostatin remains the most potent (nega-
tive) regulator of muscle mass since its discovery about 20 years ago (63).Yet, despite the com-
pelling studies in animals revealing the gross musculature of animals in which myostatin itself
or its endogenous serum binding partners (follistatin, G protein-​coupled receptor-​associated
sorting protein (GASP-​1), myostatin propeptide) were genetically manipulated, the role of
myostatin in human muscle hypertrophy remains poorly understood with studies in humans
aimed at defining the role of myostatin in hypertrophy yielding contentious data. Indeed, while
downregulation of myostatin would be expected in relation to hypertrophy-​inducing exercise,
some studies have shown this (3, 44, 51, 83), while others have noted no differences, or a poor
association with resulting muscle hypertrophy (50, 56). Much of the problem associated with
studying the role of myostatin may be due to its complex post-​translational processing (62),
which makes accurate protein identification problematic. In addition to its purported effects on
anabolic signaling, inhibition of myostatin has also been suggested as a mediator of satellite cell
activity (satellite cells represent a pool of muscle “stem cells” able to contribute to muscle repair
or growth). However, while the use of a soluble activin receptor 2B antagonist was associated

281
282

P.J. Atherton et al.

with muscle growth, this was associated with minimal labelling of satellite cell-​derived nuclei,
suggesting that myostatin acts on myofibers, rather than satellite cell to influence muscle hyper-
trophy (91). Therefore, while myostatin remains a therapeutic target for muscle growth, its
physiological role in human muscle hypertrophy remains unclear. It could be postulated that
while myostatin is important in development, its homeostatic role in the regulation and main-
tenance of adult skeletal muscle is limited, thereby limiting scope for any downregulation to
have effects. Myostatin is described in greater detail in Chapter 22.

Genetic contributors to muscle mass/​strength

As alluded to earlier, there is marked intersubject variation in an individual’s response to RET
in terms of mass and strength gains. It is therefore intuitive that this variation may be associated
with genetic variability in individuals, and the presence of common inherited mutations (or
polymorphisms).There have been a number of studies investigating this, and a number of poten-
tial candidate genes and gene variants have been proposed; for example, angiotensin-​converting
enzyme (ACE; more detail in Chapter 17) insertion/​deletion polymorphism, ACTN3 (detailed
in Chapter 23), ciliary neurotrophic factor (CNTF) and receptor, myostatin-​related genes
(MSTN; detailed in Chapter 22), and the vitamin D receptor (VDR), as highlighted by Kostek
et al. (53) and Roth et al. (79). Nevertheless, findings have been somewhat inconclusive and
lacking in reproducibility, such that definitive conclusions are difficult to draw as to whether
these gene variants are important for heritability in mass and strength or not. As such, there
remains a great deal of work to be done to investigate relationships between heritability and
muscle mass/​strength. There has been some progress however, with one of the first large-​scale
genome-​wide association meta-​analyses studies being performed and identifying five single
nucleotide polymorphisms in/​near hydroxysteroid 17-​beta dehydrogenase 11 (HSD17B11),
veriscan (VCAN), a disintegrin-​like and metalloprotease domain with thrombospondin type-​
1 motifs-​like-​3 (ADAMTSL3), insulin receptor substrate-​1 (IRS1), and fat-​mass and obesity
associated gene (FTO), which were associated with lean body mass, and with successful replica-
tion across multiple different cohorts (100).This perhaps provides an important breakthrough in
our attempts to understand the genetic variation in skeletal muscle size and strength. Due to the
complexity of analysis in “omics” and the study sizes needed to explore this complexity, it still
remains to be seen as to whether these gene variants can explain the heterogeneous responses
to RET in humans.

Relationships between muscle mass and strength gains in response to

resistance exercise training
As described in Chapter 19, the positive relationship between muscle strength and muscle
size (cross-​sectional area (CSA) and muscle volume), is well established (2, 46, 47). However,
numerous studies utilizing RET have shown early increases in strength where muscle hyper-
trophy is not yet evident (21, 81), suggesting that other variables contribute to strength gain,
such as changes to neural drive or a change in fiber composition. Moreover, classical studies
comparing strength and muscle CSA in sprinters and endurance runners (59) have shown
that sprinters produced higher maximal knee extensor voluntary isometric force compared to
endurance runners, despite there being no significant difference in muscle CSA between the
two. In addition, numerous studies have compared the effects of RET in relation to training
status (i.e., between trained and untrained individuals) (1, 60).The results of such studies suggest
that despite comparable muscle CSA, greater gains in strength (from baseline) are found in

282
283

Genetic contributions to skeletal muscles

untrained individuals (38). This is likely due to trained subjects already possessing forms of
favorable neuromuscular or morphological adaptations (e.g., altered fiber type composition
and/​or muscle architecture), in comparison with the untrained such that the untrained individ-
uals have a greater potential for adaptation and thus a more pronounced gain in strength (58).
This demonstrates that strength is not solely dependent on muscle mass; instead, there are many
adaptations contributing to strength outcomes (i.e., neuromuscular/​architectural/​fiber type).
As discussed above, neuromuscular adaptations are an important contributor to strength
changes. These adaptations can occur at the neuromuscular junction or central nervous
system (30, 49). Electromyography (EMG) allows for the investigation of neural adaptations
in response to exercise (26, 70, 72). Using surface electrodes, motor unit (MU) activation
can be recorded as an integral of the electrical activity measured (iEMG), before, during
and after contraction (39, 82). (For a more detailed review on EMG methods, see 82.) As
previously described, studies have identified that initial increases in strength correlate with
increases in iEMG (i.e., MU activation) during early periods of training where muscle
hypertrophy was not evident. Therefore, these increases in strength are due to increased
neural adaptations, such as increased MU recruitment. In addition to changes in muscle
strength following RET in a trained limb, a phenomenon termed “cross-​education” or
“cross-​training effect” was identified whereby increases in strength were also detected in the
contralateral untrained limb (47). Consequently, it was hypothesized that neural adaptations
had to be occurring more centrally to elicit such effects on the contralateral untrained limb.
It is since known that with long-​term RET, neural adaptations occur in the primary motor
cortex, M1 (49). Similarly, neuromuscular plasticity can also be observed during detraining.
Narici et al. (72) observed similar rates of change in CSA, iEMG, and maximum voluntary
contraction during the training and detraining period. Conversely, Ishida et al. (48) observed
a significant increase in maximum voluntary contraction during training but an insignifi-
cant decrease during the detraining period. This suggested that there were potential strength
training adaptations that remained. It was proposed that this was due to enzyme activity in
the muscle, as described by Guy and Snow (37), or that it was as a result of changes in the
size of sarcoplasmic reticulum or release of calcium ions. The latter is a crucial component
of the excitation–​contraction coupling cascade.

Conclusion
Muscle hypertrophy is the most recognized feature of adaptation to RET, governed by both
genetic and environmental factors.The general time-​course of RET-​induced hypertrophy is one
of rapid onset in the first weeks, before gradual slowing, despite pursuing progressive regimens
(adjusting loads for strength gains). The upstream mechanisms triggering sustained increases in
MPS (the driver of protein accretion) has been related to induction of systemic hormones and
contraction-​induced mechanotransduction within muscle; with the latter perhaps being more
critical. Either way, the activation of proximal signaling to mTORC1 is a critical step since its
substrates are linked to global mRNA translation and enhanced MPS. Nonetheless, responses are
highly variable among individuals, albeit for poorly defined reasons.This variation may relate to
impaired activation of MPS, or satellite cell recruitment, and be driven by genetic predisposition
to trainability; nonetheless, any gene variants regulating hypertrophic responses to RET and its
variation remain to be robustly identified. Strength gains accompany muscle hypertrophy and
are related to the degree of muscle hypertrophy, but also neuromuscular adaptations, perhaps at
the level of the brain and spinal cord –​but also –​at the level of individual motor units and the
excitation–​contraction coupling apparatus.

283
284

P.J. Atherton et al.

References
1. Ahtiainen JP, Pakarinen A, Alen M, Kraemer WJ, Häkkinen K. Muscle hypertrophy, hormonal
adaptations and strength development during strength training in strength-​trained and untrained
men. Eur J Appl Physiol 89: 555–​563, 2003.
2. Akagi R, Takai Y, Ohta M, Kanehisa H, Kawakami Y, Fukunaga T. Muscle volume compared
to cross-​sectional area is more appropriate for evaluating muscle strength in young and elderly indi-
viduals. Age Ageing 38: 564–​569, 2009.
3. Aoki MS, Soares AG, Miyabara EH, Baptista IL, Moriscot AS. Expression of genes related to
myostatin signaling during rat skeletal muscle longitudinal growth. Muscle Nerve 40: 992–​999, 2009.
4. Atherton PJ, Etheridge T, Watt PW, Wilkinson D, Selby A, Rankin D, Smith K, Rennie MJ.
Muscle full effect after oral protein: time-​dependent concordance and discordance between human
muscle protein synthesis and mTORC1 signaling. Am J Clin Nutr 92: 1080–​1088, 2010.
5. Atherton PJ, Wilkinson DJ, Smith K. Feeding modulation of amino acid utilization: role of insulin
and amino acids in skeletal muscle. In: The Molecular Nutrition of Amino Acids and Proteins, edited by
Dardevet D. London: Academic Press, 2016, p. 109–​124.
6. Bamman MM, Petrella JK, Kim JS, Mayhew DL, Cross JM. Cluster analysis tests the import-
ance of myogenic gene expression during myofiber hypertrophy in humans. J Appl Physiol
102: 2232–​2239, 2007.
7. Bamman MM, Shipp JR, Jiang J, Gower BA, Hunter GR, Goodman A, McLafferty CL,
Urban RJ. Mechanical load increases muscle IGF-​I and androgen receptor mRNA concentrations
in humans. Am J Physiol Endocrinol Metab 280: E383–​E390, 2001.
8. Biolo G, Tipton KD, Klein S, Wolfe RR. An abundant supply of amino acids enhances the meta-
bolic effect of exercise on muscle protein. Am J Physiol 273: E122–​E129, 1997.
9. Blazevich AJ, Cannavan D, Coleman DR, Horne S. Influence of concentric and eccen-
tric resistance training on architectural adaptation in human quadriceps muscles. J Appl Physiol
103: 1565–​1575, 2007.
10. Bodine SC, Stitt TN, Gonzalez M, Kline WO, Stover GL, Bauerlein R, Zlotchenko E,
Scrimgeour A, Lawrence JC, Glass DJ, Yancopoulos GD. Akt/​mTOR pathway is a crucial
regulator of skeletal muscle hypertrophy and can prevent muscle atrophy in vivo. Nat Cell Biol
3: 1014–​1019, 2001.
11. de Boer MD, Selby A, Atherton P, Smith K, Seynnes OR, Maganaris CN, Maffulli N, Movin
T, Narici M V, Rennie MJ. The temporal responses of protein synthesis, gene expression and cell
signalling in human quadriceps muscle and patellar tendon to disuse. J Physiol 585: 241–​251, 2007.
12. Bonfils G, Jaquenoud M, Bontron S, Ostrowicz C, Ungermann C, De Virgilio C. Leucyl-​
tRNA synthetase controls TORC1 via the EGO complex. Mol Cell 46: 105–​110, 2012.
13. Brook M, Wilkinson D, Mitchell W, Lund J, Szewczyk NJ, Greenhaff P, Smith K, Atherton
P. Skeletal muscle hypertrophy is most active during early resistance exercise training responses,
matching long term deuterium oxide (D2O)-​derived measures of muscle protein synthesis and
mTORc1-​signaling. FASEB J 29: 4485–​4496, 2015.
14. Brook MS, Wilkinson DJ, Mitchell WK, Lund JN, Phillips BE, Szewczyk NJ, Greenhaff
PL, Smith K, Atherton PJ. Synchronous deficits in cumulative muscle protein synthesis and
ribosomal biogenesis underlie age-​ related anabolic resistance to exercise in humans. J Physiol
594: 7399–​7417, 2016.
15. Brook MS, Wilkinson DJ, Phillips BE, Perez-​Schindler J, Philp A, Smith K, Atherton PJ.
Skeletal muscle homeostasis and plasticity in youth and ageing: impact of nutrition and exercise. Acta
Physiol 216: 15–​41, 2016.
16. Burd NA, West DWD, Moore DR, Atherton PJ, Staples AW, Prior T, Tang JE, Rennie MJ,
Baker SK, Phillips SM. Enhanced amino acid sensitivity of myofibrillar protein synthesis persists for
up to 24 h after resistance exercise in young men. J Nutr 141: 568–​573, 2011.
17. Burd NA, West DWD, Staples AW, Atherton PJ, Baker JM, Moore DR, Holwerda AM, Parise
G, Rennie MJ, Baker SK, Phillips SM. Low-​load high volume resistance exercise stimulates muscle
protein synthesis more than high-​load low volume resistance exercise in young men. PLoS One
5: e12033, 2010.
18. Calalb MB, Polte TR, Hanks SK, Calalb MB Polte TR HSK. Tyrosine phosphorylation of focal
adhesion kinase at sites in the catalytic domain regulates kinase activity: a role for Src family kinases.
Mol Cell Biol 15: 954–​963, 1995.

284
285

Genetic contributions to skeletal muscles

19. Cary LA, Guan JL. Focal adhesion kinase in integrin-​ mediated signaling. Front Biosci
4: D102–​D113, 1999.
20. Ceglia L, Niramitmahapanya S, Price LL, Harris SS, Fielding RA, Dawson-​Hughes B. An
evaluation of the reliability of muscle fiber cross-​sectional area and fiber number measurements in rat
skeletal muscle. Biol Proced Online 15: 6, 2013.
21. Chilibeck PD, Caldera W, Sale DG, Webber CE. A comparison of strength and muscle
mass increases during resistance training in young women. Eur J Appl Physiol Occup Physiol
77: 170–​175, 1998.
22. Damas F, Phillips S, Vechin FC, Ugrinowitsch C. A review of resistance training-​induced
changes in skeletal muscle protein synthesis and their contribution to hypertrophy. Sports Med
45: 801–​807, 2015.
23. Davidsen PK, Gallagher IJ, Hartman JW, Tarnopolsky MA, Dela F, Helge JW, Timmons JA,
Phillips SM. High responders to resistance exercise training demonstrate differential regulation of
skeletal muscle microRNA expression. J Appl Physiol 110: 309–​317, 2011.
24. Deane CS, Wilkinson DJ, Phillips BE, Smith K, Etheridge T, Atherton PJ. “Nutraceuticals”
in relation to human skeletal muscle and exercise. Am J Physiol –​Endocrinol Metab 312,
E282–​E299, 2017.
25. DeFreitas JM, Beck TW, Stock MS, Dillon MA, Kasishke PR. An examination of the time
course of training-​induced skeletal muscle hypertrophy. Eur J Appl Physiol 111: 2785–​2790, 2011.
26. DeVries HA. “Efficiency of electrical activty” as a physiological measure of the functional state of
muscle tissue. Am J Phys Med 47: 10–​22, 1968.
27. Doessing S, Heinemeier KM, Holm L, Mackey AL, Schjerling P, Rennie M, Smith K,
Reitelseder S, Kappelgaard A-​M, Rasmussen MH, Flyvbjerg A, Kjaer M. Growth hormone
stimulates the collagen synthesis in human tendon and skeletal muscle without affecting myofibrillar
protein synthesis. J Physiol 588: 341–​351, 2010.
28. Drummond MJ, Fry CS, Glynn EL, Dreyer HC, Dhanani S, Timmerman KL, Volpi E,
Rasmussen BB. Rapamycin administration in humans blocks the contraction-​induced increase in
skeletal muscle protein synthesis. J Physiol 587: 1535–​1546, 2009.
29. Durieux AC, Desplanches D, Freyssenet D, Fluck M. Mechanotransduction in striated muscle
via focal adhesion kinase. Biochem Soc Trans. 35: 1312–​1313, 2007.
30. Fang Y, Siemionow V, Sahgal V, Xiong F, Yue GH. Greater movement-​related cortical potential
during human eccentric versus concentric muscle contractions. J Neurophysiol 86: 1764–​1772, 2001.
31. Flück M, Carson JA, Gordon SE, Ziemiecki A, Booth FW. Focal adhesion proteins FAK and
paxillin increase in hypertrophied skeletal muscle. Am J Physiol 277: C152–​C162, 1999.
32. Franchi MV, Ruoss S, Valdivieso P, Mitchell KW, Smith K, Atherton PJ, Narici MV, Flück
M. Regional regulation of focal adhesion kinase after concentric and eccentric loading is related to
remodelling of human skeletal muscle. Acta Physiol 223: e13056, 2018.
33. Goodman CA, Frey JW, Mabrey DM, Jacobs BL, Lincoln HC, You J-​S, Hornberger TA.
The role of skeletal muscle mTOR in the regulation of mechanical load-​induced growth. J Physiol
589: 5485–​5501, 2011.
34. Goodman CA, Mabrey DM, Frey JW, Miu MH, Schmidt EK, Pierre P, Hornberger TA. Novel
insights into the regulation of skeletal muscle protein synthesis as revealed by a new nonradioactive
in vivo technique. FASEB J 25: 1028–​1039, 2011.
35. Gordon SE, Flück M, Booth FW. Selected Contribution: skeletal muscle focal adhesion kinase,
paxillin, and serum response factor are loading dependent. J Appl Physiol 90: 1174–​1183, 2001.
36. Greenhaff PL, Karagounis LG, Peirce N, Simpson EJ, Hazell M, Layfield R, Wackerhage H,
Smith K, Atherton P, Selby a, Rennie MJ. Disassociation between the effects of amino acids and
insulin on signaling, ubiquitin ligases, and protein turnover in human muscle. Am J Physiol Endocrinol
Metab 295: E595–​E604, 2008.
37. Guy PS, Snow DH. The effect of training and detraining on muscle composition in the horse. J
Physiol 269: 33–​51, 1977.
38. Häkkinen K, Komi P., Tesch P. Effect of combined concentric and eccentric strength training and
detraining on force-​time, muscle fiber and metabolic characteristics of leg extensor muscles. Scand J
Med Sci Sport 3: 50–​58, 1981.
39. Häkkinen K, Kraemer WJ, Newton RU, Alen M. Changes in electromyographic activity, muscle
fibre and force production characteristics during heavy resistance/​power strength training in middle-​
aged and older men and women. Acta Physiol Scan 171: 51–​62, 2001.

285
286

P.J. Atherton et al.

40. Han JM, Jeong SJ, Park MC, Kim G, Kwon NH, Kim HK, Ha SH, Ryu SH, Kim S. Leucyl-​
tRNA synthetase is an intracellular leucine sensor for the mTORC1-​ signaling pathway. Cell
149: 410–​424, 2012.
41. Hansen S, Kvorning T, Kjaer M, Sjøgaard G. The effect of short-​term strength training on
human skeletal muscle: the importance of physiologically elevated hormone levels. Scand J Med Sci
Sports 11: 347–​354, 2001.
42. Henneman E. Relation between size of neurons and their susceptibility to discharge. Science
126: 1345–​1347, 1957.
43. Heymsfield SB, Gallagher D, Visser M, Nuñez C, Wang ZM. Measurement of skeletal
muscle: laboratory and epidemiological methods. J Gerontol A Biol Sci Med Sci 50: 23–​29, 1995.
44. Hulmi JJ, Tannerstedt J, Selänne H, Kainulainen H, Kovanen V, Mero AA. Resistance exercise
with whey protein ingestion affects mTOR signaling pathway and myostatin in men. J Appl Physiol
106: 1720–​1729, 2009.
45. Iadevaia V, Huo Y, Zhang Z, Foster LJ, Proud CG. Roles of the mammalian target of rapamycin,
mTOR, in controlling ribosome biogenesis and protein synthesis. Biochem Soc Trans 40: 168–​172, 2012.
46. Ikai M, Fukunaga T. Calculation of muscle strength per unit cross-​sectional area of human muscle
by means of ultrasonic measurement. Int Zeitschrift für Angew Physiol Einschließlich Arbeitsphysiologie
26: 26–​32, 1968.
47. Ikai M, Fukunaga T. A study on training effect on strength per unit cross-​sectional area of muscle
by means of ultrasonic measurement. Int Zeitschrift für Angew Physiol Einschließlich Arbeitsphysiologie
28: 173–​180, 1970.
48. Ishida K, Moritani T, Itoh K. Changes in voluntary and electrically induced contractions during
strength training and detraining. Eur J Appl Physiol Occup Physiol 60: 244–​248, 1990.
49. Kami A, Meyer G, Jezzard P, Adams MM, Turner R, Ungerleider LG. Functional MRI evi-
dence for adult motor cortex plasticity during motor skill learning. Nature 377: 155, 1995.
50. Kim J-​S, Petrella JK, Cross JM, Bamman MM. Load-​mediated downregulation of myostatin
mRNA is not sufficient to promote myofiber hypertrophy in humans: a cluster analysis. J Appl Physiol
103: 1488–​1495, 2007.
51. Kim J, Cross JM, Bamman MM. Impact of resistance loading on myostatin expression and cell cycle
regulation in young and older men and women. Am J Physiol Endocrinol Metab 288: E1110–​E1119, 2005.
52. Klossner S, Durieux A-​C, Freyssenet D, Flueck M. Mechano-​transduction to muscle protein
synthesis is modulated by FAK. Eur J Appl Physiol 106: 389–​398, 2009.
53. Kostek M, Hubal MJ, Pescatello LS.The role of genetic variation in muscle strength. Am J Lifestyle
Med 5: 156–​170, 2011.
54. Kumar V, Atherton P, Smith K, Rennie MJ. Human muscle protein synthesis and breakdown
during and after exercise. J Appl Physiol 106: 2026–​2039, 2009.
55. Kumar V, Selby A, Rankin D, Patel R, Atherton P, Hildebrandt W, Williams J, Smith K,
Seynnes O, Hiscock N, Rennie MJ. Age-​related differences in the dose-​response relationship of
muscle protein synthesis to resistance exercise in young and old men. J Physiol 587: 211–​217, 2009.
56. Kvorning T, Andersen M, Brixen K, Schjerling P, Suetta C, Madsen K. Suppression of testos-
terone does not blunt mRNA expression of myoD, myogenin, IGF, myostatin or androgen receptor
post strength training in humans. J Physiol 578: 579–​593, 2007.
57. Lundberg TR, Fernandez-​Gonzalo R, Norrbom J, Fischer H, Tesch PA, Gustafsson T.
Truncated splice variant PGC-​1α4 is not associated with exercise-​induced human muscle hyper-
trophy. Acta Physiol (Oxf) 1: e00140, 2014.
58. Mangine GT, Gonzalez AM, Townsend JR, Adam J, Beyer KS, Miramonti AA, Ratamess
NA. Influence of baseline muscle strength and size measures on training adaptations in resistance-​
trained men. Int J Exerc Sci 11: 198–​213, 2018.
59. Maughan RJ, Watson JS, Weir J. Relationships between muscle strength and muscle cross-​sectional
area in male sprinters and endurance runners. Eur J Appl Physiol Occup Physiol 50: 309–​318, 1983.
60. Maughan RJ, Watson JS, Weir J. Muscle strength and cross-​sectional area in man: a comparison of
strength-​trained and untrained subjects. Br J Sports Med 18: 149–​157, 1984.
61. McCall GE, Byrnes WC, Dickinson A, Pattany PM, Fleck SJ. Muscle fiber hypertrophy, hyper-
plasia, and capillary density in college men after resistance training. J Appl Physiol 81: 2004–​2012, 1996.
62. McFarlane C, Langley B, Thomas M, Hennebry A, Plummer E, Nicholas G, McMahon C,
Sharma M, Kambadur R. Proteolytic processing of myostatin is auto-​regulated during myogenesis.
Dev Biol 283: 58–​69, 2005.

286
287

Genetic contributions to skeletal muscles

63. McPherron AC, Lawler AM, Lee SJ. Regulation of skeletal muscle mass in mice by a new TGF-​
beta superfamily member. Nature 387: 83–​90, 1997.
64. Miller BF, Olesen JL, Hansen M, Døssing S, Crameri RM, Welling RJ, Langberg H, Flyvbjerg
A, Kjaer M, Babraj JA, Smith K, Rennie MJ. Coordinated collagen and muscle protein synthesis
in human patella tendon and quadriceps muscle after exercise. J Physiol 567: 1021–​1033, 2005.
65. Mitchell W, Wilkinson D, Phillips B, Lund J, Smith K, Atherton P. Human skeletal muscle
protein metabolism responses to amino acid nutrition. Adv Nutr 7: 828S–​838S, 2016.
66. Mitchell WK, Phillips BE, Williams JP, Rankin D, Lund JN, Smith K, Atherton PJ. A dose-​
rather than delivery profile-​dependent mechanism regulates the “muscle-​full” effect in response to
oral essential amino acid intake in young men. J Nutr 145: 207–​214, 2015.
67. Mitchell WK, Wilkinson DJ, Phillips BE, Lund JN, Smith K, Atherton PJ. Human skeletal
muscle protein metabolism responses to amino acid nutrition. Adv Nutr 7: 828S–​838S, 2016.
68. Moore DR, Churchward-​Venne TA, Witard O, Breen L, Burd NA, Tipton KD, Phillips SM.
Protein ingestion to stimulate myofibrillar protein synthesis requires greater relative protein intakes in
healthy older versus younger men. J Gerontol Ser A Biol Sci Med Sci 70: 57–​62, 2015.
69. Moore DR, Robinson MJ, Fry JL, Tang JE, Glover EI, Wilkinson SB, Prior T, Tarnopolsky
MA, Phillips SM. Ingested protein dose response of muscle and albumin protein synthesis after
resistance exercise in young men. Am J Clin Nutr 89: 161–​168, 2009.
70. Moritani T, de Vries HA. Neural factors versus hypertrophy in the time course of muscle strength
gain. Am J Phys Med 58: 115–​130, 1979.
71. Morton RW, McGlory C, Phillips SM. Nutritional interventions to augment resistance training-​
induced skeletal muscle hypertrophy. Front Physiol 6: 245, 2015.
72. Narici M V., Roi GS, Landoni L, Minetti AE, Cerretelli P. Changes in force, cross-​sectional area
and neural activation during strength training and detraining of the human quadriceps. Eur J Appl
Physiol Occup Physiol 59: 310–​319, 1989.
73. O’Neil TK, Duffy LR, Frey JW, Hornberger TA. The role of phosphoinositide 3-​ kinase
and phosphatidic acid in the regulation of mammalian target of rapamycin following eccentric
contractions. J Physiol 587: 3691–​3701, 2009.
74. Phillips BE, Williams JP, Gustafsson T, Bouchard C, Rankinen T, Knudsen S, Smith K,
Timmons JA, Atherton PJ. Molecular networks of human muscle adaptation to exercise and age.
PLoS Genet 9: e1003389, 2013.
75. Phillips SM, Parise G, Roy BD, Tipton KD, Wolfe RR, Tamopolsky MA. Resistance-​training-​
induced adaptations in skeletal muscle protein turnover in the fed state. Can J Physiol Pharmacol
80: 1045–​1053, 2002.
76. Phillips SM, Tipton KD, Aarsland a, Wolf SE, Wolfe RR. Mixed muscle protein synthesis and
breakdown after resistance exercise in humans. Am J Physiol 273: E99–​E107, 1997.
77. Phillips SM, Tipton KD, Ferrando AA, Wolfe RR. Resistance training reduces the acute exercise-​
induced increase in muscle protein turnover. Am J Physiol 276: E118–​E124, 1999.
78. Rommel C, Bodine SC, Clarke BA, Rossman R, Nunez L, Stitt TN, Yancopoulos GD, Glass
DJ. Mediation of IGF-​1-​induced skeletal myotube hypertrophy by PI(3)K/​Akt/​mTOR and PI(3)K/​
Akt/​GSK3 pathways. Nat Cell Biol 3: 1009–​13, 2001.
79. Roth SM, Rankinen T, Hagberg JM, Loos RJF, Pérusse L, Sarzynski MA, Wolfarth B,
Bouchard C. Advances in exercise, fitness, and performance genomics in 2011. Med Sci Sports Exerc
44: 809–​817, 2012.
80. Ruas JL, White JP, Rao RR, Kleiner S, Brannan KT, Harrison BC, Greene NP, Wu J, Estall
JL, Irving BA, Lanza IR, Rasbach KA, Okutsu M, Nair KS, Yan Z, Leinwand LA, Spiegelman
BM. A PGC-​1α isoform induced by resistance training regulates skeletal muscle hypertrophy. Cell
151: 1319–​1331, 2012.
81. Rutherford OM, Jones DA. The role of learning and coordination in strength training. Eur J Appl
Physiol Occup Physiol 55: 100–​105, 1986.
82. Sale DG. Neural adaptation to resistance training. Med Sci Sport Exerc 20: 135–​145, 1988.
83. Saremi A, Gharakhanloo R, Sharghi S, Gharaati MR, Larijani B, Omidfar K. Effects of oral cre-
atine and resistance training on serum myostatin and GASP-​1. Mol Cell Endocrinol 317: 25–​30, 2010.
84. Schlaepfer DD, Hauck CR, Sieg DJ. Signaling through focal adhesion kinase. Prog Biophys Mol Biol
71: 435–​478, 1999.
85. Seynnes OR, de Boer M, Narici M V. Early skeletal muscle hypertrophy and architectural changes
in response to high-​intensity resistance training. J Appl Physiol 102: 368–​373, 2007.

287
288

P.J. Atherton et al.

86. Smith K, Barua JM, Watt PW, Scrimgeour CM, Rennie MJ. Flooding with L-​[1-​13C]leu-
cine stimulates human muscle protein incorporation of continuously infused L-​[1–​13C]valine. Am J
Physiol 262: E372–​E376, 1992.
87. Smith K, Reynolds N, Downie S, Patel a, Rennie MJ. Effects of flooding amino acids on incorp-
oration of labeled amino acids into human muscle protein. Am J Physiol 275: E73–​E78, 1998.
88. Spangenburg EE, Le Roith D, Ward CW, Bodine SC. A functional insulin-​like growth factor
receptor is not necessary for load-​induced skeletal muscle hypertrophy. J Physiol 586: 283–​291, 2008.
89. Tang JE, Perco JG, Moore DR, Wilkinson SB, Phillips SM. Resistance training alters the
response of fed state mixed muscle protein synthesis in young men. Am J Physiol Regul Integr Comp
Physiol 294: R172–​R178, 2008.
90. Timmons JA, Larsson O, Jansson E, Fischer H, Gustafsson T, Greenhaff PL, Ridden J,
Rachman J, Peyrard-​Janvid M, Wahlestedt C, Sundberg CJ. Human muscle gene expression
responses to endurance training provide a novel perspective on Duchenne muscular dystrophy.
FASEB J 19: 750–​760, 2005.
91. Wang Q, McPherron AC. Myostatin inhibition induces muscle fibre hypertrophy prior to satellite
cell activation. J Physiol 590: 2151–​2165, 2012.
92. West DWD, Burd NA, Tang JE, Moore DR, Staples AW, Holwerda AM, Baker SK, Phillips
SM. Elevations in ostensibly anabolic hormones with resistance exercise enhance neither training-​
induced muscle hypertrophy nor strength of the elbow flexors. J Appl Physiol 108: 60–​67, 2010.
93. West DWD, Kujbida GW, Moore DR, Atherton P, Burd NA, Padzik JP, De Lisio M, Tang
JE, Parise G, Rennie MJ, Baker SK, Phillips SM. Resistance exercise-​induced increases in putative
anabolic hormones do not enhance muscle protein synthesis or intracellular signalling in young men.
J Physiol 587: 5239–​5247, 2009.
94. Wilkinson DJ, Franchi M V, Brook MS, Narici M V, Williams JP, Mitchell WK, Szewczyk
NJ, Greenhaff PL, Atherton PJ, Smith K. A validation of the application of D2O stable iso-
tope tracer techniques for monitoring day-​to-​day changes in muscle protein subfraction synthesis in
humans. Am J Physiol Endocrinol Metab 306: E571–​E579, 2014.
95. Wilkinson DJ, Hossain T, Hill DS, Phillips BE, Crossland H, Williams J, Loughna P,
Churchward-​Venne TA, Breen L, Phillips SM, Etheridge T, Rathmacher J a, Smith K,
Szewczyk NJ, Atherton PJ. Effects of leucine and its metabolite β-​hydroxy-​β-​methylbutyrate on
human skeletal muscle protein metabolism. J Physiol 591: 2911–​2923, 2013.
96. Wilkinson DJ, Hossain T, Limb MC, Phillips BE, Lund J, Williams JP, Brook MS, Cegielski
J, Philp A, Ashcroft S, Rathmacher JA, Szewczyk NJ, Smith K, Atherton PJ. Impact of the
calcium form of β-​hydroxy-​β-​methylbutyrate upon human skeletal muscle protein metabolism. Clin
Nutr 37: 2068–​2075, 2017.
97. Wilkinson SB, Phillips SM, Atherton PJ, Patel R, Yarasheski KE, Tarnopolsky MA, Rennie
MJ. Differential effects of resistance and endurance exercise in the fed state on signalling molecule
phosphorylation and protein synthesis in human muscle. J Physiol 15: 3701–​3717, 2008.
98. Witard OC, Jackman SR, Breen L, Smith K, Selby A, Tipton KD. Myofibrillar muscle protein
synthesis rates subsequent to a meal in response to increasing doses of whey protein at rest and after
resistance exercise. Am J Clin Nutr 99: 86–​95, 2014.
99. Ydfors M, Fischer H, Mascher H, Blomstrand E, Norrbom J, Gustafsson T. The truncated
splice variants, NT-​PGC-​1α and PGC-​1α4, increase with both endurance and resistance exercise in
human skeletal muscle. Physiol Rep 1: e00140, 2013.
100. Zillikens MC, Demissie S, Hsu Y-​H, Yerges-​Armstrong LM, Chou W-​C, Stolk L, Livshits G,
Broer L, Johnson T, Koller DL, Kutalik Z, Luan J, Malkin I, Ried JS, Smith A V., Thorleifsson
G, Vandenput L, Hua Zhao J, Zhang W, Aghdassi A, Åkesson K, Amin N, Baier LJ, Barroso
I, Bennett DA, Bertram L, Biffar R, Bochud M, Boehnke M, Borecki IB, Buchman AS,
Byberg L, Campbell H, Campos Obanda N, Cauley JA, Cawthon PM, Cederberg H, Chen
Z, Cho NH, Jin Choi H, Claussnitzer M, Collins F, Cummings SR, De Jager PL, Demuth I,
Dhonukshe-​Rutten RAM, Diatchenko L, Eiriksdottir G, Enneman AW, Erdos M, Eriksson
JG, Eriksson J, Estrada K, Evans DS, Feitosa MF, Fu M, Garcia M, Gieger C, Girke T, Glazer
NL, Grallert H, Grewal J, Han B-​G, Hanson RL, Hayward C, Hofman A, Hoffman EP,
Homuth G, Hsueh W-​C, Hubal MJ, Hubbard A, Huffman KM, Husted LB, Illig T, Ingelsson
E, Ittermann T, Jansson J-​O, Jordan JM, Jula A, Karlsson M, Khaw K-​T, Kilpeläinen TO,
Klopp N, Kloth JSL, Koistinen HA, Kraus WE, Kritchevsky S, Kuulasmaa T, Kuusisto J,
Laakso M, Lahti J, Lang T, Langdahl BL, Launer LJ, Lee J-​Y, Lerch MM, Lewis JR, Lind

288
289

Genetic contributions to skeletal muscles

L, Lindgren C, Liu Y, Liu T, Liu Y, Ljunggren Ö, Lorentzon M, Luben RN, Maixner

W, McGuigan FE, Medina-​Gomez C, Meitinger T, Melhus H, Mellström D, Melov S,
Michaëlsson K, Mitchell BD, Morris AP, Mosekilde L, Newman A, Nielson CM, O’Connell
JR, Oostra BA, Orwoll ES, Palotie A, Parker SCJ, Peacock M, Perola M, Peters A, Polasek
O, Prince RL, Räikkönen K, Ralston SH, Ripatti S, Robbins JA, Rotter JI, Rudan I, Salomaa
V, Satterfield S, Schadt EE, Schipf S, Scott L, Sehmi J, Shen J, Soo Shin C, Sigurdsson G,
Smith S, Soranzo N, Stančáková A, Steinhagen-​Thiessen E, Streeten EA, Styrkarsdottir U,
Swart KMA, Tan S-​T, Tarnopolsky MA, Thompson P, Thomson CA, Thorsteinsdottir U,
Tikkanen E, Tranah GJ, Tuomilehto J, van Schoor NM, Verma A, Vollenweider P, Völzke H,
Wactawski-​Wende J, Walker M, Weedon MN, Welch R, Wichmann H-​E, Widen E, Williams
FMK, Wilson JF, Wright NC, Xie W, Yu L, Zhou Y, Chambers JC, Döring A, van Duijn
CM, Econs MJ, Gudnason V, Kooner JS, Psaty BM, Spector TD, Stefansson K, Rivadeneira
F, Uitterlinden AG, Wareham NJ, Ossowski V, Waterworth D, Loos RJF, Karasik D, Harris
TB, Ohlsson C, Kiel DP. Large meta-​analysis of genome-​wide association studies identifies five loci
for lean body mass. Nat Commun 8: 80, 2017.

289
290

21
GENETIC CONTRIBUTIONS
TO NEUROENDOCRINE
RESPONSES TO RESISTANCE
TRAINING
William J. Kraemer, Nicholas A. Ratamess, and Jakob L. Vingren

Introduction
The endocrine system plays a substantial role in mediating the acute responses to resistance
exercise (RE) as well as influencing subsequent baseline resistance training (RT) adaptations.
Hormonal signals play a variety of roles in anabolism (tissue growth, substrate restoration,
recovery) and catabolism (tissue breakdown, metabolic regulation). The interactions of hor-
monal signals with the cells’ genetic machinery are complex and the understanding of this
area is still in its infancy. Recent discoveries show that multiple pathways relay hormonal
signals to affect nuclear targets (48) and thus induce acute and chronic anabolic or cata-
bolic effects on target cells (3, 71). Our evolving understanding of RE-​induced hormonal
influences must be put into context of an individual’s training state, RE protocols used, and
inherent genetic potential for adaptive changes in specific phenotypic characteristics (17, 40,
45, 46) (Figure 21.1). However, hormones do not function within an isolated setting. Rather,
a specific hormonal response must be viewed within the context of the entire endocrine
system and its relationship with other physiological systems (some of which are depicted in
Figure 21.1).
It is clear that genetics substantially contribute to an individual’s baseline adaptability in muscle
strength, power, endurance, and hypertrophy from RT. Several candidate genes (i.e., ACTN3,
PPARA, CNTF, ACE, and VDR) have been identified that contribute to a strength pheno-
type where either a specific polymorphism(s) has been associated with baseline muscle strength
and mass or with the magnitude of muscle strength increases and hypertrophy from training
(8). Heritability studies have shown that genetics account for 31–​83% of variability in muscle
strength and lean tissue mass (11, 60) (see Chapter 18 for further details). From a neuroendocrine
perspective, genetics contribute to the amount of each hormone produced through regula-
tion of enzymes in biosynthetic pathways, the quantity/​activity/​affinity of transport proteins,
coregulatory proteins, hormone receptors, and molecular pathway intermediates, all of which
affect the magnitude of adaptations. Epigenetics and transcriptional modifications increase the
complexity of genetic control over neuroendocrine function. A genetic polymorphism of a
receptor can strengthen or weaken its affinity for hormone binding and thus be associated with
responder or nonresponder status for a given trait. RT influences genetic expression of many

290
291

Neuroendocrine responses to training

Amount of hormones and RE stress – interaction of

isoforms synthesized and intensity, volume, rest interval
released (pulsatility)– length, muscle mass
including biosynthetic involvement, muscle actions, Plasma volume shifts and
enzyme activity in tissues and motor units recruited (size changes in tissue blood flow
and secretory gland principle)
function

Age and sex Training

status and body
fat

Release hormone
responses and Acute blood testosterone, GH, Nutritional
autocrine/paracrine IGF-1, and cortisol responses intake and
isoform responses during and following RE hydration

Time of day Genetics

Hormone metabolism,
Acidosis and hypoxia
clearance, receptor
interaction/upregulation,
Quantity, partitioning, and
cross-reactivity and
conveyance of transport proteins
intracellular signaling
– free and bound hormone fractions

Figure 21.1 Factors influencing the acute hormonal responses to resistance exercise. RE=resistance
exercise.

genes thereby affecting mRNA and increasing synthesis of several critical proteins involved in
adaptation. In addition, gene doping has the potential to target anabolic hormone DNA (i.e.,
insulin-​like growth factor-​1 (IGF-​1)) to elicit further gains in muscle strength and hypertrophy.
Specific pathways that play roles in muscle strength (Chapter 19) and hypertrophy (Chapter 20)
are covered previously in this book. Here, we explore genetic contributions of key anabolic
hormones (androgens, growth hormone (GH) superfamily, IGF-​1), glucocorticoids (cortisol),
and their receptors to acute RE responses and adaptations to RT.

Androgens
Testosterone (T) is the primary anabolic androgen that interacts with androgen receptors
(ARs) within skeletal muscle; whereas, the more potent dihydrotestosterone (DHT) primarily
acts within sex-​linked tissues with a secondary role in skeletal muscle. T is released from the
Leydig cells of the testes (in men) under the control of gonadotropin releasing hormone
(GnRH)-​stimulated release of luteinizing hormone (LH) via neural stimulation and kisspeptin.
Kisspeptins, produced from the KISS1 gene, regulate GnRH. Kisspeptin neurons are found in
the arcuate nucleus and anteroventral periventricular nucleus of the brain (26). Mutations and
polymorphisms of the KISS1 gene and kisspeptin receptor (KISS1R) genes alter sexual mat-
uration and can lead to early onset puberty (72). How these gene polymorphisms affect mus-
cular performance has yet to be elucidated. Other sources of T synthesis include the adrenal
cortex, ovaries, and skeletal muscle (77). Skeletal muscle T may act in an intracrine rather

291
292

W.J. Kraemer et al.

CYP11A CYP17 CYP17

Cholesterol Pregnenolone 17α -hydroxpregnenolone DHEA Adrenal cortex,
testis, ovaries,
3βHSD 3βHSD 3βHSD muscle

CYP17 17α -hydroxpregnenolone CYP17

Progesterone Androstenedione

Prostate, ovary, Testis, muscle

mammary gland 17HSD5 17HSD3

Testosterone

Testis, ovaries,
CYP19 liver

Estradiol

Figure 21.2 Testosterone biosynthesis. Key enzymes catalyzing androgen biosynthesis from cholesterol
include cytochrome P450 cholesterol side chain cleavage enzyme (CYP11A); cytochrome P450 17α-​
hydroxylase/​17,20 lyase (CYP17), 3ß-​hydroxysteroid dehydrogenase (3ßHSD, types 1 and 2), 17ß-​
hydroxysteroid dehydrogenases (17ßHSD, types 3 and 5 (AKR1C3)). Aromatase cytochrome P450
(CYP19) catalyzes the conversion of T to estradiol; whereas, 5α-​reductase catalyzes the conversion of T
to the more potent DHT (not shown).

than endocrine manner. Testosterone is synthesized from cholesterol in a series of steps that
involve the intermediate formation of weaker steroids including DHEA, androstenedione, and
androstenediol (Figure 21.2).T is released into circulation upon synthesis where it is transported
bound to binding proteins, primarily sex hormone-​binding globulin (SHBG). Free (unbound,
~2% of total T in circulation) testosterone (FT) is taken up at the tissue level for potential
binding to ARs. A change in SHBG concentrations may influence the magnitude of FT avail-
able for diffusion across the cell membrane.

Acute testosterone responses and chronic adaptations to resistance training

RE can elicit elevations in T and FT concentrations for up to 30 minutes after exercise in men
with either small or no increases in women (41, 79).Acute elevations in androstenedione, DHEA
sulfate, and SHBG, as well as delayed increases in LH have also been reported. The magnitude
of the acute androgen response is affected by the exercise selection and muscle mass involve-
ment, the interaction of intensity, volume, and rest interval lengths, nutritional intake, metabolic
demand, age, and training experience (41, 45, 46). A few studies have shown augmented acute
T responses following a period of RT. The ramifications of acute elevations in androgens with
RE are unclear as some studies show relationships between T elevations and AR upregulation
and strength and hypertrophy enhancement whereas others do not (45).Thus, acute T responses
must be viewed within the context of multiple skeletal muscle signaling pathway adaptations as
well as how it coincides with other androgen, GH, IGF-​1, insulin, and cortisol responses.
Findings for effects of RT on resting T are inconsistent and may be confounded by other
variables (i.e., diurnal variations, nutritional interventions, and the time blood sample was
obtained relative to last workout), thereby making interpretation difficult. Most studies have
shown no changes in resting T or FT concentrations following RT while some have found
elevations or reductions (41, 45). Other resting androgen (DHT, DHEA or DHEA sulfate,
androstenedione) or SHBG concentrations may decrease or not change (41, 45). Because

292
293

Neuroendocrine responses to training

diurnal T concentrations are tightly regulated, it is unlikely for a permanent change to occur as
regulatory mechanisms are quickly re-​engaged after acute RE (41). Rather, an elevation/​reduc-
tion may reflect a transient change that will return to a normal circadian pattern soon thereafter.

Androgen biosynthesis and skeletal muscle steroidogenesis

Androgen biosynthesis (Figure 21.2) is affected by other hormones including insulin and IGF-​1
(31). Several genes encode for these biosynthetic enzymes, resulting in varied concentrations
of androgens and other steroids in multiple tissues. For example, mutations in the HSD17B3
gene encoding the enzyme 17β-​HSD3 cause male pseudo-​hermaphroditism during fetal devel-
opment due to low T and DHT (51). Polymorphisms of genes encoding for 17β-​HSD3 and
17β-​HSD5 have been related to increased risk for diseases such as polycystic ovary syndrome
and cancer. Although some polymorphisms are associated with increased androgen levels, their
potential effects on muscle performance are poorly understood.
Of recent interest is the investigation of steroidogenic enzymes in skeletal muscle capable
of synthesizing T from DHEA, androstenedione, and androstenediol. Steroidogenic enzyme
content and T concentrations in skeletal muscle appear to be similar between men and women
(77). RE has been shown to increase skeletal muscle DHEA, FT, DHT, 3β-​HSD, 17β-​HSD, and
5α-​reductase type I content in older men (65) but not in resistance-​trained young men and
women within 70 minutes of exercise (77). Statistical trends for increased muscle concentrations
of T and DHT have been observed 3 and 24 hours after RE. Sato et al. (65) showed signifi-
cant correlations between muscle DHEA, FT, and 5α-​reductase concentrations and isokinetic
strength and muscle cross-​sectional area (CSA) in older men. These findings indicate that skel-
etal muscle is a source of T, albeit the effects of RE remain unclear but could be a mechanism
to help counteract circulating T reductions in older men.

Androgen signaling
Androgens affect acute performance and increase CSA of type I and II muscle fibers, strength,
power, and endurance via several molecular pathways (Figure 21.3). Androgens increase protein
synthesis, alter fiber types, increase lactate transport via upregulation of monocarboxylate trans-
porter 1 (MCT1) and 4 (MCT4) proteins, increase satellite cell activation, proliferation, differ-
entiation, and incorporation into skeletal muscle as myonuclei, differentiation of mesenchymal
pluripotent cells, follistatin expression, myotube formation, increase polyamine synthesis, GH,
muscle mechanogrowth factor (MGF), and IGF-​1Ea mRNA, downregulate and upregulate
myostatin gene expression, downregulate Acvr2b receptor mRNA and Ankrd1, inhibit forkhead
box O (FoxO) family of transcription factors, and downregulate glucocorticoid receptor expres-
sion (46). Nongenomic actions are rapid and result in a rise in intracellular calcium and may
lead to PI3K/​Akt/​mTORC1/​S6K1 pathway-​induced muscle hypertrophy.

Androgen receptor genetic variations

The androgen receptor gene (AR) is located on the q arm of chromosome X at position 11–​12
(Figure 21.4) and is widely expressed in most tissues. The first exon contains several regions
of repetitive DNA sequences, one of which is the CAG triplet repeat that begins at codon 58,
with the length varying between 8 and 35 repeats. Polymorphisms yielding a variety of repeats
are associated with a variety of conditions including male infertility, prostate and testicular
cancer, bone disease, and cardiovascular disease. Long CAG repeats may interfere with androgen

293
294

5
Cytoplasm Nucleus
GR
AR
FT/DHT Target Gene
Hsp 90
1
AR Coregulator Proteins
Hsp 70
• Dimerization
•Binding to ARE
Hsp 40
Transcription
2
Wnt F D GSK-3

TCF/LEF
-catenin
mRNA

Protein Synthesis
↑ β-catenin
3 6
Satellite Cell
AR

Activation AR 4
PL-C IP3 Sarcoplasmic
Reticulum
G-protein

PKA

SHBG-R
G-protein
7

SHBG
Ca++ Transcription
Myogenesis Factors
AR cAMP
MPC Muscle Contractility

Figure 21.3 A simplified model of intracellular androgen signaling in skeletal muscle. A truncated
model of anabolic and anticatabolic actions of androgens is shown. Pathway 1 depicts classical genomic
activation. FT or DHT diffuses through the cell membrane and binds to an AR in the cytoplasm –​
binding dissociates heat shock proteins (HSP) stabilizing the AR; the FT-​AR complex translocates to
the nucleus and binds to specific DNA elements (androgen response elements or AREs). Coactivators,
general transcription factors, and RNA polymerase II are recruited, transcription is increased, and protein
synthesis increases. Pathway 2 depicts a secondary genomic pathway, canonical Wnt-​β-​catenin signaling.
Wnt binds to frizzled (F) receptors and activates Dishevelled (D) and inhibits glycogen synthase kinase-​3
(GSK-​3) reducing β-​catenin degradation. The FT-​AR complex inhibits GSK-​3 and increases β-​catenin
where it translocates to the nucleus, binds to DNA response elements (T-​cell factor/​lymphoid enhancer
factor 1 (TCF/​LEF)), increases transcription, and activation of muscle satellite cells. Pathway 3 depicts
a nongenomic pathway where FT binds to a membrane-​bound AR that is coupled with a pertussis
toxin (PTX)-​sensitive G protein that activates phospholipase C (PL-​C) and increases IP3. IP3 binds to
specific receptors on the sarcoplasmic reticulum (SR) and increases intracellular calcium. Calcium may
mediate/​augment muscle contraction or activate several pathways including Src/​Ras/​ERK/​MAPK/​
mTOR to increase transcription. Pathway 4 depicts the SHBG receptor (SHBG-​R) activation upon
binding of SHBG-​T activating a G-​protein to form cyclic AMP (cAMP), activation of protein kinase
A (PKA), and increased transcriptional activation of nuclear AR. Pathway 5 depicts T/​DHT ability to
decrease glucocorticoid receptor (GR) expression and increase competitive binding between androgen-​
AR complex and common response elements between AR and GR shared cistromes on target
genes. Pathway 6 depicts androgen-​AR binding on satellite cells increases proliferation, mobilization,
differentiation, and incorporation into skeletal muscle. T/​AR binding to myogenic intermediates may
also occur, favoring increased differentiation. Pathway 7 depicts androgen-​AR binding on mesenchymal
pluripotent cells (MPC) increases propensity towards myogenesis in skeletal muscle. Crosstalk between
pathways, muscle steroidogenesis, and other noncortisol hormone-​receptor interactions are not shown
thereby showing the vast nature of androgen signaling.
295

Neuroendocrine responses to training

actions (reduce interaction between the N and C termini), whereas short repeats are associated
with increased AR protein expression and enhance androgen action through stronger binding
yielding greater transcriptional activity.
The potential relationship between CAG repeat number and muscle strength and lean body
mass (LBM) might be critical. AR gene polymorphisms affecting CAG repeat length could help
explain the concept of responders and nonresponders to RT in addition to the physiological
responses to exogenous T administration. However, contradictory results have been reported
where CAG repeat number was positively related, inversely related, or not related to LBM, T or
FT concentrations, and muscle strength in young and older men (12, 69, 81). In a study examining
a large number of young men aged 20–​29 years, Nielsen reported significant inverse relationships
between CAG repeat number and thigh and trunk muscle size (52). Reductions in repeats of ten
equaled increases of muscle size of 6–​11 cm2.Thus, performance phenotypes based on AR candi-
date gene polymorphisms remain unclear but is an area of great RT.

Androgen receptor structure

The nuclear AR is a 110 kD receptor consisting of 919 amino acids, 12 α-​helices, and 2 ß-​
sheets that belongs to a large family of nuclear transcription factors (Figure 21.4). A truncated

q11–12

X Chromosome

Intron size (kb)

>20 >15 26 5.6 4.8 0.8 0.7
1 2 3 4 5 6 7 8

Exon size (kb) 1613 152 117 288 145 131 158 155

DNA - Binding

Transcription Regulation Hinge Ligand - Binding

NH2 CAG

Amino Acid 1 537 Zn++ 626 919

COOH
Serine 81, 94, and 650: phosphorylation sites

Figure 21.4 The androgen receptor gene and protein. The AR gene (upper panel) is oriented with
the 5ʹ end toward the centromere and spans approximately 90 kb of DNA containing eight exons that
code for approximately 2757-​base pair open reading frames within a 10.6 kb mRNA (14). The first
exon codes for the N-​terminus domain; exons 2–​3 code for the central DNA binding domain; exons
4–​8 code for the C terminus ligand-​binding domain (14). The AR (lower panel) forms ligand-​induced
homodimers that bind to inverted repeat DNA response elements and consist of four functional
domains: a C-​terminal ligand-​binding domain (LBD; comprises amino acids 662–​919), a DNA-​binding
domain that is built with two zinc finger motifs (amino acids 538–​611), a hinge region (amino acids
612–​661), and a N-​terminus transcription activation domain (amino acids 1–​537) (10).

295
296

W.J. Kraemer et al.

AR protein (87 kD) with similar function has been identified. Androgen binding (or binding
by selective androgen receptor modulators (SARMs)) to AR is part of a complex signaling
system (Figure 21.3). Androgen binding stabilizes the AR in a dose-​dependent manner and
this stabilization is critical to modifying AR content. The half-​life of AR without androgen
binding is 1 hour whereas the formation of the AR-​androgen complex extends the half-​life to
6 hours (32).T dissociates from the AR three times faster than DHT or synthetic androgens and
is less effective for stabilizing the AR (86), which helps explains the greater potency of these
compounds. The AR is capable of undergoing multiple bouts of recycling (at least four times)
between the nucleus and cytoplasm after ligand binding and dissociation (62).

Androgen receptor adaptations to training

Androgens and RE each upregulate AR protein content. AR gene expression in skeletal muscle
has correlated with FT concentrations and a few studies have shown relationships between
exercise-​induced T and FT elevations and AR content from 1 to 48 hours post RE (45,46). AR
quantity is related to muscle strength and CSA (61). The importance of androgens in mediating
training-​induced hypertrophy was shown in multiple studies (41). In male AR gene knockout
mice (but not female), significant reductions in muscle mass (up to 25%) and strength were
reported (49). Ninety-​three genes were up-​(N=47) or downregulated (N=46; including several
regulating muscle cytoskeletal proteins) in male AR knockout mice (49) showing the com-
plexity of androgen signaling across several pathways.
Single and sequential bouts of RE may increase AR protein or mRNA expression with the
magnitude and time course mediated by the volume of RE, muscle mass involvement, nutri-
tional consumption, and the quantity of circulating T in the blood (45). Initially, downregulation
or no change may occur in men and women but upregulation appear to follow within 3 hours
post RE and may persist for up to 48 hours post RE (45,61). Baseline AR protein content or
mRNA may not change over long-​term RT, suggesting that the acute cyclical upregulation in
response to sequential RE bouts may play a critical role in mediating androgen-​induced mus-
cular adaptations. AR up-​and downregulation may be cross-​regulated via hormonal or growth
factor (e.g., IGF-​1, cortisol, and the GH superfamily) interactions with other receptors. Such
cross-​regulation demonstrates the importance of several molecular pathways in AR regulation
and provides a combinatorial mechanism for hormone-​specific positive and/​or negative modu-
lation of different genes.

Growth hormone(s)
Our understanding of GH has evolved significantly over the years (33, 43, 50). The single 191-​
amino acid 22 kD monomer is not the only form found in anterior pituitary somatotrophs or
in circulation (4, 5, 55); hence, the concept of the GH “superfamily” of isoforms with different
molecular heterogeneity. Considering the diversity of GH isoforms, a host of questions remain
regarding the effects of RE on GH-​family molecules. Different immunoassays were used to
assay GH during RE in the 1990s as research focused on GH responses to a variety of exercise
protocols, the concept of pulsatility, and sexual dimorphism. Bioassays of GH revealed concen-
tration differences from radioimmunoassays. This created the concept that there is a difference
between the “bioactive” versus “immunoreactive” assay signals. Concerns about GH drug use
dramatically increased and stimulated various controversies regarding drug testing and if GH
was, in fact, an effective anabolic hormone in skeletal muscle (25).

296
297

Neuroendocrine responses to training

Basic growth hormone cybernetics

Multiple pathways are involved with regulating the release of GH. The control of GH release
starts with neural stimulation of a pulsatile release of the hypothalamic GH releasing hor-
mone (GHRH) contained in arcuate neurons of the neurosecretory nerve terminals into
hypothalamo-​hypophyseal portal circulation to the anterior pituitary where it interacts with
GHRH-​receptors, a class II G protein-​coupled receptor, and via an extensive signaling system
mediate the release of GH (13).The 22 kD monomer assayed by immunoassay (IGH) is released
in a pulsatile manner similar to its trophic hormone (76). The larger bioactive GH aggregates
(BGH) assay signals have not shown pulsatility. Somatostatin opposes the stimulation of GHRH
and is released from periventricular somatostatin neurons, carried by the same portal circulation
to the anterior pituitary where it inhibits GH secretion. This dynamic is thought to contribute
to the pulsatile release of IGH.The hypothalamus acts as the primary pulse generator of various
hormones. Interestingly, with high-​volume RE, IGH reduces pulsatility in the first half of sleep
followed by higher pulsatility in the second half of sleep, indicating pulse generation and the
types of GH released may vary from the 22 kD monomer (54). Such an effect may be due to an
increased somatostatin influence or more provocatively a greater synthesis of aggregated forms
of BGH, limiting the production and release of IGH from band I somatotrophs.
A series of steps initiate the synthesis of the 22 kD monomer in the somatotrophs of the
anterior pituitary. GHRH binds to its receptor and creates a conformational change in the alpha
Gs subunit of the G-​protein complex in the intracellular side (80). From here, a series of cAMP
pathways are stimulated ending with the transcription of GH by the interactions of cAMP
response elements with the promoter region of the GH gene. This produces the 191-​amino
acid, 22 kD monomer secreted from the somatotrophs. Increased transcription of the GHRHR
gene creates the well-​known positive feedback loop.

Somatotroph types and growth hormone isoforms

Two distinctly different somatotrophs of GH have been identified (20): band I contains lower-​
molecular-​weight forms of GH (predominantly 22 kD) whereas band II contains higher-​
molecular-​weight isoforms of GH (i.e., aggregates or BGH) (Figure 21.5). Aggregates may
be formed via zinc, disulfide, and hydrostatic bindings as partial degradation of the aggregates’
disulfide bonds increases the 22 kD monomer concentration reflecting one of these aggre-
gate binding mechanisms (30, 63). Additionally, tibial peptide, found in the granules and
somatotrophs, has biological functions with the tibial growth plate and possibly other unknown
cellular interfaces (29).

Growth hormone in circulation

GH has the potential for post-​translational processing within the periventricular space (e.g.,
GH to prolactin). Hosts of molecular GH isoforms, splice variants, and fragments have been
assayed in plasma (4, 5).There is a unique circulatory pathway directly into the brain potentially
allowing for the GH superfamily to influence neuroplasticity and function (7). Both high-​and
low-​affinity binding proteins have been identified with the high-​affinity GH binding pro-
tein (GHBP), considered to be the principal binding protein in circulation. GHBP arises from
proteolytic cleavage of the extracellular domain of the GH receptor (85). This process occurs
mostly in the liver, which is the most GH receptor-​r ich tissue, but could occur wherever GH
receptors exists. It has been suggested that GHBP may play a role in GH supply and receptor

297
 newgenrtpdf
298
Figure 21.5 Growth hormone somatotrophs in the anterior pituitary. The organization and biosynthesis of human growth hormone (GH) in the
anterior pituitary is related to the production of different hGH isoforms, from the 22 kD to splice variants (e.g., 20 kD) to larger molecular weight
aggregates (e.g., dimers GH-​GH; trimers GH-​GH-​GH etc.). Band 1 and 2 somatotrophs have been identified (20) and GH isoforms are organized
and stored for release into the circulation through a microtubular transport system. Release into the periventricular area also allows for last minute,
post-​translational processing of GH prior to being released into circulation (e.g., GH to prolactin (PL)). High-​and low-​affinity binding proteins
derived from the liver also may bind with GH, yet their functions remain unclear. GH can have direct effects on the target tissues and need not be
mediated via IGF-​1. Multiple pathways exist and there is a great deal of signal complexity (33). One common pathway is shown. GHR=growth
hormone receptor; JAK-​STAT=Janus kinase signal transducer and activator of transcription proteins; MAPK=mitogen-​activated protein kinase;
P=phosphate.
299

Neuroendocrine responses to training

function (67). Interestingly, with heavy RE, significant increases in GHBP were observed in
both sedentary and trained men indicating training does not augment the GHBP response to
RE. The BGH response indicates that some molecular weight fractions are more responsive
than others to RE. Nevertheless, signal strength varies with the type of assay used, e.g., IGH or
BGH (28). GH receptors are present in many different tissues including skeletal muscle and are
coded by the GH receptor gene. Ultimately, GH binds to the GH receptor resulting in receptor
dimerization, activating intra-​and intercellular transduction signal pathways influencing metab-
olism and growth (33).

Resistance exercise responses

RE alters plasma IGH concentrations more consistently than BGH (see Table 21.1). Our
understanding of RE effects has become complicated given the varieties of different assays
used to measure GH (28, 43). The RE protocol influences the IGH response with higher
volume of work, 10RM resistances, and short rest resulting in the greatest increases in men and
women (37, 38) likely because reduced pH mediates such responses (18). McCall et al. (50) first
observed that BGH was responsive to RE and had an inherent neurological afferent feedback
mechanism with unloading. Later, whole-​body RE protocols in women were found to elicit
increases in IGH, whereas BGH was unchanged (30). Only with long-​term (6 months) RT in
women did BGH concentrations significantly increase (42) but still with little or no response to
acute exercise. Interestingly, stronger women have been found to have higher concentrations of
BGH (38) and older women to have lower BGH than younger women post RE (19).
Health issues may also be reflected in the BGH signal. Obesity results in a lower BGH assay
signal in younger men at rest and during RE, implicating reduced function of the band II
somatotrophs (73). Recent findings of a lack of BGH signal in resting blood samples of older
men and women (60–​95 years) with no age relationships suggest that somatotrophs’ function-
ality is not guaranteed (44).

Insulin-​like growth factors

IGFs are structurally related to insulin. IGFs are small polypeptide hormones (70 and 67 amino
acids for IGF-​1 and IGF-​2, respectively) synthesized from a larger precursor peptide that is
post-​translationally processed into its active form. Of the two, IGF-​1 has been most exten-
sively studied and is secreted by the liver in response to GH stimulation. Only 2% of IGF-​1
circulates in its free form; most circulates as a binary (20–​25%) or ternary complex (~75%)
(53–​56). In its binary form, IGF-​1 circulates with one of seven binding proteins whereas in
its ternary form, IGF-​1 circulates with IGFBP-​3 and its acid labile subunit (ALS). IGF-​1 is a
systemic anabolic and metabolic hormone and also produced locally (i.e., autocrine and para-
crine mechanisms) in tissues and cells. IGF-​1 acts as both a cell cycle initiation and progression
factor and facilitates satellite cell activation, proliferation, survival, and differentiation; increases
myotube size and number of nuclei per myotube; stimulates amino acid uptake and protein
synthesis and muscle hypertrophy, neuronal myelinization, axonal sprouting, and damage repair;
reduces chronic inflammatory response; increases free fatty acid utilization; and enhances insulin
sensitivity upon receptor binding and subsequent intracellular signaling and glucose metabolism
(22, 46). Much of the hypertrophy-​induced effects of IGF-​1 occur via effects on satellite cells.
Because of these critical anabolic functions, IGF genes have been considered a potential target
for gene therapies and for gene doping in athletes (23). While liver-​derived IGF-​1 is under
direct regulation of GH, local mechanical-​stretch mechanisms can activate IGF-​1 synthesis in

299
300

W.J. Kraemer et al.

Table 21.1 Comparison of assay signal concentrations from immunoassays and bioassays during resistance
exercise. Postexercise (PE) concentrations were determined between 0 and 15 minutes in young men
(M) and women (W)

Study (reference) Sex Program Immunoassay Bioassay total (BGH)

(IGH) µg/​L µg/​L

Pre-​exercise PE Pre-​exercise PE

McCall et al. (50) M Isometric plantar 12.0 14.0 1300.0 3300.0*

flexion 5 min at
30% MVC
McCall et al. (50) M Isometric plantar 12.0 13.0 1100.0 2300.0*
flexion
1 min at 80% MVC
(B)
Kraemer et al. (37) M 8 EX, 10RM 1 min 2.2 23.9* − −
RI
Kraemer et al. (37) M 8 EX, 5 RM 3 min 2.1 5.2* − −
RI
Kraemer et al. (38) W 8 EX, 10RM 1 min 7.6 16.5* − −
RI
Kraemer et al. (38) W 8 EX, 5 RM 3 min 7.1 5.6 − −
RI
Hymer et al. (30) W Squat 6 × 10 80% 7.7 26.3* 3927.0 3787.0
1RM, 2 min RI
Kraemer et al. (42) W Squat 6 × 10 80% 2.6 11.2* 2450 3300
1RM, 2 min RI
Thomas et al. (73) M 3 × 10 reps: 85–​ 1.2 9.9* 6000.0 11,000.0
95% of 10 RM
for squat, bench
press, leg curls,
dumbbell
rows, dumbbell
shoulder press,
and dumbbell
step-​up

*=P ≤0.05 from corresponding pre-​exercise value; EX=exercises; MVC=maximal voluntary

contraction; RM=repetition maximum; RI=rest interval.

tissues. The potency of circulating IGF-​1 remains unclear and needs to be viewed in context
with its binding proteins that regulate bioavailability as several studies have shown systemic
elevations in IGF-​1 produced no elevations in protein synthesis or hypertrophy during RT
whereas upregulation in the muscle isoform was linked to significant muscle hypertrophy (22).

IGF and IGFBP genes

The IGF-​1 gene is shown in Figure 21.6. Some IGF-​1 gene variants (i.e., cytosine adenine
dinucleotide repeats in the promoter region) have been reported and associated with greater

300
301

Neuroendocrine responses to training

3 4 5 6

MGF (IGF-1Ec)

Leader sq.
Leader sq. 1 1 2 3 4 5 6

IGF-1 gene

3 4 5 6

IGF-1Ea

3 4 5

IGF-1Eb

Figure 21.6 IGF-​1 isoforms. The IGF-​1 gene is mapped to the long arm of chromosome 12 (12q22–​
q24.1) and the short arm of chromosome 11 (11p15). The gene contains two promoters and six exons
which could be spliced to produce at least six mRNAs. Exons 1 and 2 determine the protein class,
exons 3 and 4 encode the mature peptide and ligand-​binding site, and exons 5 and 6 determine the
E domain peptide and polyadenylating signals. The gene produces IGF-​1Ea, IGF-​1Eb, and IGF-​1Ec
(mechanogrowth factor (MGF)) (23).

strength changes during RT (34). This same polymorphism has been associated with baseline
muscle power, i.e., lower power outputs for those homozygous for the 192-​allele compared to
other genotypes, but not with power increases accompanying RT (70). The C–​1245T (rs35767)
is a single nucleotide polymorphism (SNP) found in the promoter region and has been associated
(the minor T allele) with systemic IGF-​1 concentrations, muscle strength, response to eccentric
exercise, bone mineral density, LBM, and left ventricular hypertrophy, in addition to being found in
a small percent (5%) of endurance and power athletes compared to other genotypes (6,35). Carriers
of the GA genotype with the rs6220 SNP have shown greater strength than AA homozygotes and
CC homozygotes of the SNP rs7136446 showed greater strength (by 9%) compared to TT and TC
genotypes (27). No such relationships have been found for the rs2854744 SNP of the IGFBP-​3
gene (27). Polymorphisms of the IGF-​2 gene (e.g.,ApaI) have been associated with muscle strength.
In men, the IGF-​2 SNPs G13790C, C16646T, and ApaI were significantly associated with strength
loss, creatine kinase activity, and soreness following eccentric exercise (9).

Acute responses and chronic adaptations of IGFs to resistance training

The systemic liver-​induced IGF-​1 response to RE is unclear as elevations (during and up to
30 minutes after exercise), decreases, and no changes have been found (41). A lack of change
may be attributed to, at least in part, the delay in secretion of IGF-​1 (i.e., 3–​9 hours) following
GH stimulation (41). Despite the anabolic actions of IGF-​1, a direct link between RE-​induced

301
302

W.J. Kraemer et al.

elevations and muscle strength and hypertrophy has been difficult to determine. Some evidence
indicates that the role of systemic IGF-​1 may lie more as a “regulator or amplifier” of muscle
remodeling (46). Of importance is the response of IGFBPs, which have generated more con-
sistent responses with RE acutely elevating IGFBP-​3 (41). Nindl et al. (54) monitored over-
night IGF-​1 following heavy RE and showed IGF-​1 concentrations remained unchanged but
IGFBP-​2 increased and ALS decreased, suggesting that binding protein partitioning, rather than
changes in systemic IGF-​1, was critical. Other studies have shown IGFBP-​1 was sensitive to
exercise duration, but not exercise mode. Using microdialysis to measure IGF-​1 in interstitial
fluid, Nindl et al. (56) showed total and free IGF-​1 and IGFBP-​3 were elevated; whereas, IGF-​1
in interstitial fluid was unaltered following high-​power exercise. Furthermore, IGF-​1 receptor
phosphorylation was not increased, and IGF mRNA content and Akt phosphorylation were
increased (56), reinforcing the concept that skeletal muscle adaptation is not dependent on sys-
temic IGF-​1, but rather the interplay across biocompartments. Chronic RT studies examining
resting circulating IGF-​1 concentrations have shown reductions, no change, and elevations with
no change or reductions in IGFBP-​1 and IGFBP-​3 (41). No significant changes in IGF-​1,
IGFBP-​1, or IGFBP-​3 were found in subjects who were considered extreme responders to 16
weeks of RT versus nonresponders, although a trend was shown where IGFBP-​3 in extreme
responders tended to be lower than nonresponders (58). Resistance-​trained men were shown to
have higher resting IGF-​1 concentrations than untrained men (64).Thus, the contributions and
ramifications of single-​data point circulating IGF-​1 chronic measures remain unclear.
Of greater significance may be the training responses of locally produced IGF-​1 isoforms.
RE of sufficient intensity and volume increases IGF-​1 and MGF mRNA for up to 48 hours
post RE (41,59). IGF-​1 and MGF mRNA have increased 2 hours post exercise (but not 6
hours) after a single bout of moderate-​(65% of 1RM; 18–​20 repetitions) and moderately high-​
intensity (85% of 1RM; 8–​10 repetitions) RE (82). Other studies show MGF acts independ-
ently, is expressed earlier than other IGF-​1 isoforms in response to RE, and may have greater
anabolic potency (15). Overloaded muscle and mechanical damage from RT are potent stimuli
for local production of IGF-​1.

IGF-​1 receptor and intracellular signaling

The endocrine, autocrine, and paracrine effects of IGF-​1 are mediated through binding to
the IGF-​1 receptor (IGF-​1R), a ligand-​activated receptor tyrosine kinase. The IGF-​1R gene
is mapped to chromosome 15q25–​26. Two receptor types have been identified (type 1 and 2).
IGF-​1 binds with greater affinity to type 1 than to type 2 or the insulin receptor whereas IGF-​2
binds with greater to affinity to the type 2 receptor (see Figure 21.7). IGF-​1R phosphorylation
may be sensitive to RE favoring high-​volume over high-​intensity protocols 1 hour after exer-
cise (16) and moderate-​and moderately high-​intensity RE has been shown to increase IGF-​1R
mRNA 2 hours after RE (82).

Glucocorticoids
Glucocorticoids (GCs) are steroids synthesized primarily in the adrenal cortex (with minimal
amounts produced in other tissues such as the brain, thymus, skin, and gastrointestinal tract).
The primary glucocorticoid is cortisol (~95% of GC activity), synthesized from pregnenolone
via P450c17 17α-​ hydroxylation and a few other intermediate steps. Cortisol secretion is
under neuroendocrine control via hypothalamic secretion of corticotropin-​releasing hormone
(CRH), the activation of gene transcription of pro-​opiomelanocortin (POMC) in the anterior

302
303

Neuroendocrine responses to training

IGF-1

α α
β β
P P
Shc PIP 2
IRS-1
PI3K
PIP 3
GRB2 SOS [Ca++]

PDK1
Ras Calcineurin
GSK3
TSC1 AKT Fox01 atrogenes
TSC2
Raf
NFATc1 p27

MEK
GATA-2 Rheb mTORC1 eIF-4E

ERK1/2 (MAPK)
p70S6k
Cyclin D1 c-myc p90rsk
Protein Synthesis
Proliferation Differentiation

Figure 21.7 Intramuscular IGF-​1 signaling. IGF-​1R contains two extracellular α subunits (IGF-​1
binding sites) and two transmembrane β subunits that upon binding to IGF-​1 autophosphorylate the
tyrosine residues which stimulates intermediates such as the IRS family (IRS-​1 in skeletal muscle) and Shc
proteins. The two main intracellular signaling pathways involve the phosphatidyl-​inositol 3-​kinase (PI3)/​
Akt kinase pathway and the mitogen-​activated protein kinase (MAPK)/​ERK pathway (55). A calcium-​
dependent calcineurin pathway has been identified resulting in the activation of the transcription
factor GATA-​2 whose upregulation is associated with muscle hypertrophy (59). IGF-​1 has also been
shown to increase DHP receptor concentrations in skeletal muscle suggesting that IGF-​1 may influence
excitation–​contraction coupling and subsequent recovery of muscle following damage (59). In addition,
IGF-​1 may increase cyclin D1 mRNA thereby enhancing proliferation and the myogenic response and
inhibit “atrogenes” (59). Lastly, IGF-​1 interaction with IGF-​1R is influenced by other hormone–​receptor
interactions and negatively affected by proinflammatory cytokines such as TNFα, IL-​1, and IL-​6 (1).

pituitary, and subsequent release of its fragment adrenocorticotropic hormone (ACTH) which,
in turn, induces cortisol release in the adrenal cortex. Cortisol is released in response to several
stimuli including the stress of RE. Cortisol binds mostly to corticosteroid-​binding globulin
(CRG; ~75%), albumin (~15%), while 10% of circulating cortisol is free. Similar to androgens,
GCs diffuse through the cell membrane, and bind to their nuclear receptors, but elicit catabolic
effects on skeletal muscle especially in type II muscle fibers. As for androgens, rapid nongenomic
effects of GC have also been reported (24).

Acute responses and chronic adaptations of glucocorticoids

to resistance training
Several studies have shown significant elevations in cortisol and ACTH from an acute bout
of RE with the response similar between men and women (41). The acute cortisol response

303
304

W.J. Kraemer et al.

correlates to blood lactate and protocols high in volume, moderate-​to-​high intensity with
short rest periods tend to elicit the most substantial cortisol increase (41). Resting cortisol
concentrations reflect a long-​term training stress and RT has been shown to produce no change,
reductions, and elevations (41). It appears that the acute cortisol response may reflect metabolic
stress, whereas chronic changes (or lack of change) may be involved with tissue homeostasis
involving protein metabolism.

Glucocorticoid signaling
Glucocorticord/Glucocorticord Receptor (GR) interaction increases protein degradation,
decreases protein synthesis, decreases transport of amino acids, and causes atrophy, mainly
of type muscle II fibers. Of interest is the potential competitive signaling between GC/​
GR and androgens/​AR. In fact, the homologies of the DNA-​binding and ligand-​binding
domains between the AR and GR are 79% and 50%, respectively, and both AR and GR
can bind to the same cis-​element on a shared androgen and GC DNA response element, i.e.,
5ʹ-​AGAACAnnnTGTTCT-​3ʹ (21).Thus, competitive binding in this region could help mediate,
in part, anabolic versus catabolic signaling within muscle.
Signaling pathways are vast and catabolism may occur in a variety of ways. Glucocorticoids
inhibit the anabolic actions of androgens, insulin, IGF-​ 1, and amino acids (leucine) (66).
Glucocorticoids stimulate major proteolytic pathways such as the ubiquitin proteasome, lyso-
somal (autophagy), and calpain systems. Glucocorticoids have also been shown to inhibit the PI3/​
Akt pathway (that mediates insulin and IGF-​1 signaling) via reduced IRS-​1 and increased p85α
and miR1 expression, increase FOXO-​1 and FOXO-​3a gene expression, activate “atrogenes”
such as Atrogin-​1, MuRF-​1, and Cathepsin-​L, inhibit mTORC1 via increased transcription
of REDD1 and KLF-​15, increase C/​EBPβ and p300, stimulate glycogen synthase kinase 3β
(GSK3β) (and decrease β-​catenin), inhibit production of muscle IGF-​1 isoforms, increase Pdk4
gene expression (and inhibit pyruvate dehydrogenase), reduce circulating T concentrations, and
increase myostatin (MSTN) gene expression and mRNA with a concomitant decrease in MHC
II protein expression (47, 66).

Glucocorticoid receptor
The GR gene (NR3C1) is located on chromosome 5q31–​32. It encodes the GR, which
belongs to the nuclear receptor superfamily of ligand-​dependent transcription factors. NR3C1
contains NH2-​terminal transactivation, DNA-​binding, and ligand-​binding domains as well as
a hinge region. Coactivators, chromatin remodeling enzymes, and RNA polymerase II assist
with hormone interaction and transcription processes. The GR DNA-​binding domain binds
to specific DNA sequences in the promoter region of target genes. Multiple isoforms of the
GR are expressed in various tissues (24). GC binding to the GR dissociates heat shock proteins
and the cortisol/​GR complex is chaperoned to the nucleus via a protein complex along the
microtubules. Chromatin remodeling occurs via pioneering transcription factors enabling the
hormone/​GR complex binding to the genomic target sequence. The ligand-​bound GR can
then regulate gene expression.
Several polymorphisms of the GR gene have been identified and shown to relate to various
health indices.Van Rossum and colleagues (74) examined the ER22/​23EK polymorphism and
reported associations with LBM and muscle strength in young men but not women. This poly-
morphism has been associated with GC resistance, lower cholesterol, and increased insulin

304
305

Neuroendocrine responses to training

sensitivity (75). A follow-​up study showed that the ER22/​23EK and N363S polymorphisms
modified the inverse relationships seen between serum cortisol and muscle mass and strength in
older adults (57).The Bcll polymorphism in intron 2 of the GR gene showed higher incidences
of obesity and insulin resistance (74,75). The N363S polymorphism has been associated with
lower bone mineral density and increased insulin resistance and body mass index (74,75). Ash
and colleagues (2) studied GR polymorphisms prior to and following 12 weeks of RT and
reported that men with the 2722 G>A and 1887 GG genotypes gained 16% less strength than
AA homozygotes and 6% greater strength gains than A alleles, respectively (accounting for 3.2%
and 2.4% of the variance of strength changes respectively), with no effects on muscle size. In
women, only those carrying the 1017 T>C allele gained more muscle size than CC homozygotes
(accounting for 1.7% of the variance of relative size). No genotypes were associated with base-
line strength or muscle size. Thus, some polymorphisms of GR mediate the sensitivity of the
GC response and can partially affect strength and hypertrophy adaptions to RT.
Acute changes in GR mRNA or protein content and potential long-​ term up-​or
downregulation play a role in tissue remodeling and muscle damage repair following RE and
could mediate part of the anabolic/​catabolic processes seen during RT. Willoughby et al. (83)
reported significant upregulation of GR protein content 6 and 24 hours following an eccentric
RE protocol in untrained men with an attenuated response from the same protocol 3 weeks
later. However, in RT men and women no change was found in GR content 10 and 70 minutes
following a RE protocol (78). Interestingly, women had higher GR content than men at all time
points. These studies suggest that training status plays a critical role in GR training response.
Willoughby (84) also reported higher GR content following 6 and 12 weeks of RT; however,
samples were taken in close proximity to training so the chronic adaptation is difficult to sur-
mise. In contrast, 6 weeks of intense run training in military cadets (leading to increased muscle
mass) induced a significant 6.3% reduction in GR mRNA (68). Thus, reduced GC sensitivity
may be seen over time during training but the acute response may be determined by training
status and the extent of muscle damage incurred.

Summary
Testosterone, GH superfamily, IGF-​1, and cortisol play a substantial role in mediating the
acute responses to RE and subsequent RT adaptations. Genetics substantially contributes to
these responses and adaptations via the amount of each hormone produced through regula-
tion of enzymes in biosynthetic pathways, the quantity/​activity/​affinity of transport proteins,
coregulatory proteins, hormone receptors, and molecular pathway intermediates. Several can-
didate genes have been identified that contribute to a strength and power performance pheno-
type. Further study is needed to investigate more specific performance phenotypes based on
genetic variability.

References
1. Adams GR. Insulin-​like growth factor 1 signaling in skeletal muscle and the potential for cytokine
interactions. Medicine and Science in Sports and Exercise 42: 50–​57, 2010.
2. Ash GI, Kostek MA, Lee H, et al. Glucocorticoid receptor (NR3C1) variants associate with the
muscle strength and size response to resistance training. PLoS One 11: e0148112, 2016.
3. Bajer B, Vlcek M, Galusova A, Imrich R, Penesova A. Exercise associated hormonal signals as
powerful determinants of an effective fat mass loss. Endocrine Regulation. 49: 151–​163, 2015.

305
306

W.J. Kraemer et al.

4. Baumann G. Growth hormone heterogeneity: genes, isohormones, variants, and binding proteins.
Endocrine Reviews 12: 424–​449, 1991.
5. Baumann G. Growth hormone isoforms. Growth Hormone & IGF Research 19: 333–​340, 2009.
6. Ben-​Zaken S, Meckel Y, Nemet D, Eliakim A. Can IGF-​1 polymorphism affect power and
endurance athletic performance? Growth Hormone & IGF Research 23: 175–​178, 2013.
7. Bergland R, Blume H, Hamilton A, Monica P, Paterson R. Adrenocorticotropic hormone may
be transported directly from the pituitary to the brain. Science 31: 541–​543, 1980.
8. Bray MS, Hagberg JM, Perusse L, Rankinen T, Roth SM, Wolfarth B, Bouchard C. The
human gene map for performance and health-​related fitness phenotypes: the 2006–​2007 update.
Medicine and Science in Sports and Exercise 41: 34–​72, 2009.
9. Devaney JM, Hoffman EP, Gordish-​Dressman H, Kearns A, Zambraski E, Clarkson PM.
IGF-​II gene region polymorphisms related to exertional muscle damage. Journal of Applied Physiology
102: 1815–​1823, 2007.
10. Eder IE, Culig Z, Putz T, Menardi CN, Bartsch G, Klocker H. Molecular biology of
the androgen receptor: from molecular understanding to the clinic. European Urology 40:
241–​251, 2001.
11. Eynon N, Hanson ED, Lucia A, Houwelling PJ, Garton F, North KN, Bishop DJ.
Genes for elite power and sprint performance: ACTN3 leads the way. Sports Medicine
43: 803–​8 17, 2013.
12. Folland JP, McCauley TM, Phypers C, Hanson B, Mastana SS. The relationship of testos-
terone and AR CAG repeat genotype with knee extensor muscle function of young and older men.
Experimental Gerontology 47: 437–​443, 2012.
13. Gaylin BD. Growth hormone releasing hormone receptor. Receptors Channels 8: 155–​162, 2002.
14. Gelman EP. Molecular biology of the androgen receptor. Journal of Clinical Oncology
20: 3001–​3015, 2002.
15. Goldspink G, Wessner B, Tschan H, Bachl N. Growth factors, muscle function, and doping.
Endocrinology Metabolism Clinics of North America 39: 169–​181, 2010.
16. Gonzalez AM, Hoffman JR, Townsend JR, et al. Intramuscular anabolic signaling and endocrine
response following high volume and high intensity resistance exercise protocols in trained men.
Physiological Reports 3: e12466, 2015.
17. Gonzalez AM, Hoffman JR, Stout JR, Fukuda DH, Willoughby DS. Intramuscular anabolic
signaling and endocrine response following resistance exercise: implications for muscle hypertrophy.
Sports Medicine 46: 671–​685, 2016.
18. Gordon SE, Kraemer WJ, Vos NH, Lynch JM, Knuttgen HG. Effect of acid-​base balance on
the growth hormone response to acute high-​intensity cycle exercise. Journal of Applied Physiology
76: 821–​829, 1994.
19. Gordon SE, Kraemer WJ, Looney DP, Flanagan SD, Comstock BA, Hymer WC. The influ-
ence of age and exercise modality on growth hormone bioactivity in women. Growth Hormone and
IGF Research 24: 95–​103, 2014.
20. Grindeland RE, Kraemer WJ, Hymer WC. Two types of rat pituitary somatotrophs secrete
growth hormone with different biological and immunological profiles. Growth Hormone and IGF
Research 36: 52–​56, 2017.
21. Harada N, Inui H, and Yamaji R. Competitive and compensatory effects of androgen signaling and
glucocorticoid signaling. Receptors and Clinical Investigation 2: e785, 2015.
22. Harridge SDR. Plasticity of human skeletal muscle: gene expression to in vivo function. Experimental
Physiology 92: 783–​797, 2007.
23. Harridge SDR, Velloso CP. IGF-​1 and GH: potential use in gene doping. Growth Hormone and
IGF Research 19: 378–​382, 2009.
24. Hartmann K, Koenen M, Schauer S, Wittig-​Blaich S, Ahmad M, Baschant U, Tuckerman
JP. Molecular actions of glucocorticoids in cartilage and bone during health, disease, and steroid
therapy. Physiological Reviews 96: 409–​447, 2016.
25. Hoffman JR, Kraemer WJ, Bhasin S, Storer T, Ratamess NA, Haff GG, Willoughby DS,
Rogol AD. Position stand on androgen and human growth hormone use. Journal of Strength and
Conditioning Research 23: S1–​S59, 2009.
26. Hooper DR, Kraemer WJ, Focht BC, Volek JS, DuPont WH, Caldwell LK, Maresh CM.
Endocrinological roles for testosterone in resistance exercise responses and adaptations. Sports Medicine
47: 1709–​1720, 2017.

306
307

Neuroendocrine responses to training

27. Huuskonen A, Lappalainen J, Oksala N, Santtila M, Häkkinen K, Kyrolainen H, Atalay M.

Common genetic variation in the IGF1 associates with maximal force output. Medicine and Science in
Sports and Exercise 43: 2368–​2374, 2011.
28. Hymer WC, Grindeland RE, Nindl BC, Kraemer WJ. Growth hormone variants and human
exercise In: The Endocrine System in Sports and Exercise, edited by Kraemer WJ, Rogol, AD. Malden,
MA: Blackwell Publishers, p. 77–​93, 2005.
29. Hymer W, Kirshnan K, Kraemer W, Welsch J, Lanham W. Mammalian pituitary growth hor-
mone: applications of free flow electrophoresis. Electrophoresis 21: 311–​317, 2000.
30. Hymer WC, Kraemer WJ, Nindl BC, et al. Characteristics of circulating growth hormone in
women after acute heavy resistance exercise. American Journal of Physiology, Endocrinology Metabolism
281: E878–​E887, 2001.
31. Kempna P, Marti N, Udhane S, Fluck CE. Regulation of androgen biosynthesis –​a short review
and preliminary results from the hyperandrogenic starvation NCI-​H295R cell model. Molecular and
Cellular Endocrinology 408: 124–​132, 2015.
32. Kemppainen JA, Lane MV, Sar M, Wilson EM. Androgen receptor phosphorylation, turnover,
nuclear transport, and transcriptional activation. Specificity for steroids and antihormones. Journal of
Biological Chemistry 267: 968–​974, 1992.
33. Kopchick JJ. Lessons learned from studies with the growth hormone receptor. Growth Hormone and
IGF Research 28: 21–​25, 2016.
34. Kostek MC, Delmonico MJ, Reichel JB, Roth SM, Douglass L, Ferrell RE, Hurley BF. Muscle
strength response to strength training is influenced by insulin-​like growth factor 1 genotype in older
adults. Journal of Applied Physiology 98: 2147–​2154, 2005.
35. Kostek MC, Devaney JM, Gordish-​Dressman H, et al. A polymorphism near IGF1 is associated
with body composition and muscle function in women from the Health, Aging, and Body
Composition Study. European Journal of Applied Physiology 110: 315–​324, 2010.
36. Kraemer WJ, Noble BJ, Culver BW, Clark MJ. Physiologic responses to heavy-​resistance exercise
with very short rest periods. International Journal of Sports Medicine 8: 247–​252, 1987.
37. Kraemer WJ, Marchitelli L, Gordon SE, et al. Hormonal and growth factor responses to heavy
resistance exercise protocols. Journal of Applied Physiology 69: 1442–​1450, 1990.
38. Kraemer WJ, Fleck SJ, Dziados JE, et al. Changes in hormonal concentrations after different
heavy-​resistance exercise protocols in women. Journal of Applied Physiology 75: 594–​604, 1993.
39. Kraemer WJ, Rubin MR, Häkkinen K, et al. Influence of muscle strength and total work on
exercise-​induced plasma growth hormone isoforms in women. Journal of Science and Medicine in Sport
6: 295–​306, 2003.
40. Kraemer WJ, Ratamess NA. Fundamentals of resistance training: progression and exercise pre-
scription. Medicine and Science in Sports and Exercise 36: 674–​688, 2004.
41. Kraemer WJ and Ratamess NA. Hormonal responses and adaptations to resistance exercise and
training. Sports Medicine 35: 339–​361, 2005.
42. Kraemer WJ, Nindl BC, Marx JO, et al. Chronic resistance training in women potentiates growth
hormone in vivo bioactivity: characterization of molecular mass variants. American Journal of Physiology,
Endocrinology, Metabolism 291: E1177–​E1187, 2006.
43. Kraemer WJ, Dunn-​Lewis C, Comstock BA, Thomas GA, Clark JE, Nindl BC. Growth
hormone, exercise, and athletic performance: a continued evolution of complexity. Current Sports
Medicine Reports 9: 242–​252, 2010.
44. Kraemer WJ, Kennett MJ, Mastro AM, et al. Bioactive growth hormone in older men and
women: its relationship to immune markers and healthspan. Growth Hormone and IGF Research
34: 45–​54, 2017.
45. Kraemer WJ, Ratamess NA, Flanagan SD, Shurley JP, Todd JS, Todd TC. Understanding the
science of resistance training: an evolutionary perspective. Sports Medicine 47: 2415–​2435, 2017.
46. Kraemer WJ, Ratamess NA, Nindl BC. Recovery responses of testosterone, growth hormone, and
IGF-​1 after resistance exercise. Journal of Applied Physiology 122: 549–​558, 2017.
47. Kuo T, Harris CA, Wang JC. Metabolic functions of glucocorticoid receptor in skeletal muscle.
Molecular and Cellular Endocrinology 380: 79–​88, 2013.
48. Kupr B, Schnyder S, Handschin C. Role of nuclear receptors in exercise-​ induced muscle
adaptations. Cold Spring Harbor Perspectives in Medicine 7: a029835, 2017.
49. MacLean HE, Chiu WSM, Notini AJ, et al. Impaired skeletal muscle development and function in
male, but not female, genomic androgen receptor knockout mice. FASEB Journal 22: 2676–​2689, 2008.

307
308

W.J. Kraemer et al.

50. McCall GE, Goulet C, Grindeland RE, Hodgson JA, Bigbee AJ, Edgerton VR. Bed rest
suppresses bioassayable growth hormone release in response to muscle activity. Journal of Applied
Physiology 83: 2086–​2090, 1997.
51. Mizrachi D, Auchus RJ. Androgens, estrogens, and hydroxysteroid dehydrogenases. Molecular and
Cellular Endocrinology 301: 37–​42, 2009.
52. Nielsen TL, Hagen C, Wraae K, Bathum L, Larsen R, and Brixen K. The impact of the CAG
repeat polymorphism of the androgen receptor gene and adipose tissues in 20–​29-​year-​old Danish
men: Odense Androgen Study. European Journal of Endocrinology 162: 795–​804, 2010.
53. Nindl BC, Hymer WC, Deaver DR, Kraemer WJ. Growth hormone pulsatility profile
characteristics following acute heavy resistance exercise. Journal of Applied Physiology 91: 163–​172, 2001.
54. Nindl BC, Kraemer WJ, Marx JO, Arciero PJ, Dohi K, Kellogg MD, Loomis GA. Overnight
responses of the circulating IGF-​I system after acute, heavy-​resistance exercise. Journal of Applied
Physiology 90: 1319–​1326, 2001.
55. Nindl BC. Exercise modulation of growth hormone isoforms: current knowledge and future
directions for the exercise endocrinologist. British Journal of Sports Medicine 41: 346–​348, 2007.
56. Nindl BC, Urso ML, Pierce JR, et al. IGF-​I measurement across blood, interstitial fluid, and muscle
biocompartments following explosive, high-​power exercise. American Journal of Physiology –​Regulatory
Integrated Comparative Physiology 303: R1080–​R1089, 2012.
57. Peeters GMEE, van Schoor NM, van Rossum EFC, Visser M, Lips P.The relationship between
cortisol, muscle mass and muscle strength in older persons and the role of genetic variations in the
glucocorticoid receptor. Clinical Endocrinology (Oxford) 69: 673–​682, 2008.
58. Petrella JK, Kim J, Mayhew DL, Cross JM, Bamman MM. Potent myofiber hypertrophy during
resistance training in humans is associated with satellite cell-​mediated myonuclear addition: a cluster
analysis. Journal of Applied Physiology 104: 1736–​1742, 2008.
59. Philippou A, Halapas A, Maridaki M, Koutsilieris M. Type I insulin-​like growth factor receptor
signaling in skeletal muscle regeneration and hypertrophy. Journal of Musculoskeletal Neuronal Interaction
7: 208–​218, 2007.
60. Puthucheary Z, Skipworth JRA, Rawal J, Loosemore M, Van Someren K, Montgomery HE.
Genetic influences in sport and physical performance. Sports Medicine 41: 845–​859, 2011.
61. Ratamess NA, Kraemer WJ, Volek JS, et al. Androgen receptor content following heavy resistance
exercise in men. Journal of Steroid Biochemistry Molecular Biology 93: 35–​42, 2005.
62. Roy AK, Tyagi RK, Song CS, Lavrovsky Y, Ahn SC, Oh TS, Chatterjee B. Androgen
receptor: structural domains and functional dynamics after ligand-​receptor interaction. Annals of the
NY Academy of Science 949: 44–​57, 2001.
63. Rubin MR, Kraemer WJ, Kraemer RR, et al. Responses of growth hormone aggregates to
different intermittent exercise intensities. European Journal of Applied Physiology 89: 166–​170, 2003.
64. Rubin MR, Kraemer WJ, Maresh CM, et al. Response of high-​affinity growth hormone binding
protein to acute heavy resistance exercise in resistance-​trained and untrained men. Medicine and Science
in Sports and Exercise 37: 395–​403, 2005.
65. Sato K, Iemitsu M, Matsutani K, Kurihara T, Hamaoka T, Fujita S. Resistance training restores
muscle sex steroid hormone steroidogenesis in older men. FASEB Journal 28: 1891–​1897, 2014.
66. Schakman O, Kalista S, Barbe C, Loumaye A, Thissen JP. Glucocorticoid-​induced skeletal
muscle atrophy. The International Journal of Biochemistry and Cell Biology 45: 2163–​2172, 2013.
67. Schilbach K, Bidlingmaier M. Growth hormone binding protein –​physiological and analytical
aspects. Best Practice & Research Clinical Endocrinology & Metabolism 29: 671–​683, 2015.
68. Silva TS, Longui CA, Rocha MN, et al. Prolonged physical training decreases mRNA levels of
glucocorticoid receptor and inflammatory genes. Hormone Research in Paediatrics 74: 6–​14, 2010.
69. Simmons ZL, Roney JR. Variation in CAG repeat length of the androgen receptor gene predicts
variables associated with intrasexual competitiveness in human males. Hormones and Behavior
60: 306–​312, 2011.
70. Sood S, Hanson ED, Delmonico MJ, et al. Does insulin-​like growth factor 1 genotype influence
muscle power response to strength training in older men and women? European Journal of Applied
Physiology 112: 743–​753, 2012.
71. Spiering BA, Kraemer WJ, Anderson JM, et al. Resistance exercise biology: manipulation of
resistance exercise programme variables determines the responses of cellular and molecular signaling
pathways. Sports Medicine 38: 527–​540, 2008.

308
309

Neuroendocrine responses to training

72. Teles MG, Silveira LF, Tusset C, Latronico AC. New genetic factors implicated in human GnRH-​
dependent precocious puberty: the role of kisspeptin system. Molecular and Cellular Endocrinology
346: 84–​90, 2011.
73. Thomas GA, Kraemer WJ, Kennett MJ, et al. Immunoreactive and bioactive growth hor-
mone responses to resistance exercise in men who are lean or obese. Journal of Applied Physiology
111: 465–​472, 2011.
74. Van Rossum EF, Voorhoeve PG, te Velde SJ, et al.The ER22/​23EK polymorphism in the gluco-
corticoid receptor gene is associated with a beneficial body composition and muscle strength in
young adults. Journal of Clinical Endocrinology and Metabolism 89: 4004–​4009, 2004.
75. Van Rossum EF, Lamberts SW. Polymorphisms in the glucocorticoid receptor gene and their
associations with metabolic parameters and body composition. Recent Progress in Hormone Research
59: 333–​357, 2004.
76. Veldhuis JD, Anderson SM, Shah N, et al. A neurophysiological regulation and target-​tissue impact
of the pulsatile mode of growth hormone secretion in the human. Growth Hormone and IGF Research
11 Suppl A: S25–​S37, 2001.
77. Vingren JL, Kraemer WJ, Hatfield DL, et al. Effect of resistance exercise on muscle steroidogenesis.
Journal of Applied Physiology 105: 1754–​1760, 2008.
78. Vingren JL, Kraemer WJ, Hatfield DL, et al. Effect of resistance exercise on muscle steroid
receptor protein content in strength-​trained men and women. Steroids 74: 1033–​1039, 2009.
79. Vingren JL, Kraemer WJ, Ratamess NA, Anderson JM, Volek JS, Maresh CM. Testosterone
physiology in resistance exercise and training: the up-​stream regulatory elements. Sports Medicine
40: 1037–​1053, 2010.
80. Vo N, Goodman RH. CREB-​binding protein and p300 in transcriptional regulation. Journal of
Biological Chemistry 276: 13505–​13508, 2001.
81. Walsh S, Zmuda JM, Cauley JA, et al. Androgen receptor CAG repeat polymorphism is associated
with fat-​free mass in men. Journal of Applied Physiology 98: 132–​137, 2005.
82. Wilborn CD, Taylor LW, Greenwood M, Kreider RB, Willoughby DS. Effects of different
intensities of resistance exercise on regulators of myogenesis. Journal of Strength and Conditioning
Research 23: 2179–​2187, 2009.
83. Willoughby DS, Taylor M, Taylor L. Glucocorticoid receptor and ubiquitin expression after
repeated eccentric exercise. Medicine and Science in Sports and Exercise 35 : 2023–​2031, 2003.
84. Willoughby DS. Effects of heavy resistance training on myostatin mRNA and protein expression.
Medicine and Science in Sports and Exercise 36: 574–​582, 2004.
85. Zhang Y, Jiang J, Black RA, Baumann G, Frank SJ. Tumor necrosis factor-​ α converting
enzyme (TACE) is a growth hormone-​ binding protein (GHBP) sheddase: the metalloprotease
TACE/​ADAM-​17 is critical for (PMA-​induced) GH receptor proteolysis and GHBP generation.
Endocrinology 141: 4342–​4348, 2000.
86. Zhou ZX, Lane MV, Kemppainen JA, French FS, Wilson EM. Specificity of ligand-​dependent
androgen receptor stabilization: receptor domain interactions influence ligand dissociation and
receptor stability. Molecular Endocrinology 9: 208–​218, 1995.

309
310

22
MYOSTATIN’S ROLE IN
GENETIC CONTROL OF
MUSCLE SIZE AND STRENGTH
Dustin S. Hittel

Introduction
Previous chapters have described the genetic, molecular, and neuroendocrine underpinnings of
skeletal muscle size and strength and their response to exercise training. In this chapter, we discuss
the specific role of myostatin, a protein that encompasses these three levels of control over muscle
morphology and yet underscores the challenges and opportunities facing the nascent field of
sport and exercise systems genetics. The transforming growth factor-​beta (TGF-​β) superfamily
includes a diverse group of growth and differentiation factors that regulate embryonic develop-
ment and subsequently, tissue homeostasis in adult animals (59). In 1997, Alexandra McPherron
and Se-​Jin Lee identified growth/​differentiation factor-​8 (GDF-​8), a novel muscle-​specific TGF-​
β family member that is expressed during embryogenesis and continues to be expressed by cells
fated to the muscle lineage throughout development and into adulthood (13, 33). McPherron
and Lee went on to characterize the developmental role of GDF-​8 by disrupting the gene in
mice, using targeted inactivation resulting in extremely hypermuscular animals with very low
body fat.These findings determined that prenatal loss of GDF-​8 liberates skeletal muscle growth
through a combination of hypertrophy (an increase in the diameter of individual skeletal muscle
fibers) and hyperplasia (an increase in the number of skeletal muscle fibers) (4, 6, 19, 32, 33, 37).
Anecdotally, this gene was renamed myostatin (MSTN) in homage to a drug first appearing in
“The Incredible Hulk” television series, but most likely owes this shared etymology with the
Greek description of its role in skeletal muscle homeostasis (myo-​(of muscle) stasis-​(inactivity/​
inhibition)).
The discovery of myostatin subsequently revealed the genetic basis for naturally occurring
“double-​muscling” phenotypes of significant agricultural interest such as the Piedmontese,
Asturiana, Marchigiana, and Belgian Blue cattle breeds (34), with additional natural mutations
associated with excess muscle mass in racing dogs (40) and domestic sheep (10). Similarly,
humans harboring natural mutations in the myostatin gene exhibit hypermuscularity and
low body fat without any obvious deleterious effects (34, 50), which indicates that the
role of myostatin in regulating muscle growth development is conserved across species.
Hypermuscularity was also evident in adult mice treated with a variety of myostatin-​depleting
biological drugs (9), whereas postnatal myostatin overexpression is associated with skel-
etal muscle wasting so dramatic it resembles cachexia (44, 64). Taken together, these studies

310
311

Myostatin and muscle size and strength

identify the primary role for myostatin as a powerful negative regulator of skeletal muscle
growth and development (3).
Since its discovery, myostatin has generated significant popular interest. One needs only
to conduct a cursory Internet search to find hundreds of “myostatin-​blocking” supplements
targeting athletes seeking to exceed the natural limits of human physiology (13, 57). While the
efficacy of these supplements does not survive minimal scientific scrutiny, the development of
myostatin-​blocking therapies may still hold great promise for patients suffering from muscle
wasting disease and cachexia.Apart from regulating skeletal muscle size alone in healthy mice and
humans, myostatin has also been implicated in the metabolic deterioration and muscle wasting
associated with sarcopenic obesity, heart failure, both type 1 and 2 diabetes, and cancer (13). As
such, this chapter will focus on myostatin’s impact on muscle size and strength characteristics
with particular attention paid to its role in exercise adaptation and muscle wasting. For a broader
historical perspective on the discovery and characterization of this fascinating growth factor, the
reader is referred to several excellent reviews (3, 13, 29, 52, 56).

Gene expression, secretion, and signaling

The myostatin gene (MSTN), located on chromosome 2 in humans, is organized into three
exons separated by two introns (Human Genome Assembly GRCh38.p12, chromosome
2: 190,055,697–​190,062,729). Among the other members of the TGF-​β superfamily, GDF-​
11 exhibits the highest sequence homology with myostatin, with 90% identity of the carboxy
terminal region. Interestingly, parabiosis experiments, the surgical union of two mice allowing
sharing of the blood circulation, identified GDF-​11 as a potential therapeutic for age-​related
cardiac and skeletal muscle decrements (21). However, recent experiments showing that
supraphysiological levels of GDF-​11 cause wasting of both skeletal and cardiac muscle indicate
a deleterious role in muscle wasting diseases (21).
Because of the potentially devastating effects of unchecked or aberrant expression (64),
the spatiotemporal regulation of myostatin signaling is highly coordinated (11). As with most
TGF-​β family members, myostatin is initially synthesized as an inactive pro-​protein (375 amino
acids in humans) with an amino-​terminal signal peptide (23 amino acids), pro-​domain (242
amino acids), and a carboxy-​terminal (108-​amino acid) growth factor domain (11).The primary
sequence of myostatin is remarkably consistent among vertebrates (~96% identity), supporting
a conserved physiological role. After synthesis, the pro-​protein forms a disulfide dimer that
is subsequently cleaved by furin-​like pro-​protein convertase (furin) into an amino-​terminal
inhibitory pro-​domain and a dimeric carboxy-​terminal active domain (22). Secreted myostatin
then circulates as a latent multiprotein complex consisting of the carboxy dimer noncovalently
bound to the pro-​ domain and/​or other circulating inhibitory proteins such as follistatin,
follistatin-​like-​3 (FSTL3), and G protein-​coupled receptor associated sorting proteins 1 and 2
(GASP-​1/​2) (22). In addition, decorin, a matrix-​associated proteoglycan, also binds myostatin
which likewise prevents the interaction of the peptide with its receptor, whereas TCAP/​
telethonin/​limb-​girdle muscular dystrophy 2G, a prominent sarcomeric protein, traps myostatin
in the Golgi apparatus, possibly preventing its secretion in response to mechanical disruption
of muscle sarcomeres (13). Similarly, a pool of unprocessed extracellular pro-​myostatin is bound
to a large latent complex leading to extracellular sequestration and prevention of furin cleavage
(58). Further evidence for proteolytic regulation of myostatin occurs during myoblast differen-
tiation, where the expression of furin mRNA is negatively regulated by myostatin resulting in
slower processing of pro-​myostatin into its biologically active form (31).This negative-​feedback

311
 312

Figure 22.1 The activation (→) and inhibition (-​-​|) of myostatin signaling by exercise, disease, and its rescue by pharmacological targeting. (A) A schematic
representing how exercise and disease reciprocally affect the expression of myostatin resulting in adaptive or maladaptive changes in skeletal muscle size and
strength. Also shown are natural and synthetic (pharmacological) protein modulators of myostatin signaling that dictate the remodeling of healthy muscle in
response to exercise and potentially reverse muscle atrophy associated with disease. The blow-​up panel (B) describes the primary molecular processes in myostatin-​
involved muscle atrophy where dashed lines (-​-​-​-​) indicate the presence of intermediate steps not included in the figure. Myostatin signaling begins with binding
to ActRIIB, which in turn recruits activin type 1 receptors and induces the phosphorylation of SMAD2/​3, leading to its binding with SMAD4 and translocation of
the complex to the nucleus where it blocks the transcription of genes responsible for the myogenesis (MYOD, PAX7) and increases the expression of FOXO1 and
SMAD7. SMAD7 attenuates myostatin signaling by blocking the interaction of SMAD2/​3 with activin type 1 receptors as well as preventing the formation of the
SMAD 2/​3 and SMAD4 complex. The phosphorylation of SMAD1/​5/​8 via the BMP signaling pathway also blocks this complex by competing for SMAD4. AKT
phosphorylation occurs in the response to insulin and IGF-​1 stimulation. The active/​phosphorylated form of AKT in turn activates mTOR leading to increased
protein synthesis and the inhibition of FOXO1 by phosphorylation. Under conditions of high myostatin signaling, AKT phosphorylation is inhibited, leading
to decreased mTOR activity and to decreased protein synthesis. In addition, activated FOXO1 translocates to the nucleus and induces the transcription of E3
ubiquitin ligases MURF-​1 and atrogin-​1, increasing protein degradation.
313

Myostatin and muscle size and strength

mechanism is thought to regulate myostatin levels during prenatal development to allow for
coordinated muscle development (52). Finally, myostatin-​binding proteins, regardless of the type
or route of delivery, produce a hypermuscular phenotype in adult mice that indicates a common
mechanism in vivo (Figure 22.1A) (11).
Recent structural characterization of the latent myostatin complex suggests that it circulates
in serum in a “spring-​loaded” state that can be quickly triggered into activation in the appro-
priate physiological context (55). Biologically active myostatin is liberated from its inhibitory
protein complex by the BMP1/​Tolloid (TLD) metalloprotease located in the extracellular
matrix of muscle, heart (in humans), and possibly other tissues (55, 61). Free from inhibitory
proteins, the approximately 25 kDa mature myostatin dimer binds to the cell surface activin
receptor type IIa or IIb (ActIIRA or ActIIRB) and two of either activin type 1 (ALK4) or
TGF-​β type 1 (ALK5) receptors to form a heterotetrametic complex, which in turn induces
the phosphorylation of the Sma-​and Mad-​related (SMAD) family of transcription factors
via the canonical TGF-​β signaling pathway (11, 58). SMAD proteins are transcription factors
that regulate the expression of hundreds of genes, including those involved in muscle growth
(development and exercise) and wasting (atrophy and disease) (29). In addition to the regulation
of gene transcription, the phosphorylation of SMAD2/​3 also inhibits AKT/​protein kinase B
phosphorylation leading to low mechanistic/​mammalian target of rapamycin (mTOR) activity
and hence lower rates of protein synthesis (29). Conversely, the phospho-​AKT-​mediated acti-
vation of mTOR with myostatin inhibition accounts for only half of the resulting increase in
muscle mass, revealing the equal importance of the regulation of atrophy-​promoting genes
(Figure 22.1B) (29).
During myogenesis, the myostatin-​induced phosphorylation of SMAD2/​3 facilitates binding
to SMAD4 and subsequently translocation to the nucleus, where it regulates the transcription
of myogenic factors that enable myogenic precursor cells to exit the cell cycle and differen-
tiate into myofibers (15). There is also mounting evidence that mitogenic signaling through the
bone morphogenetic protein (BMP) receptor modulates SMAD2/​3 signaling by competing
for SMAD4 via SMAD1/​5/​8 phosphorylation (20). For instance, when myostatin signaling is
high, SMAD4 binds to SMAD2/​3 and then translocates to the nucleus where it can decrease
the expression of genes needed for muscle growth and/​or increase the expression of genes that
drive muscle wasting (Figure 22.1B). However, when BMP signaling is high, SMAD4 prefer-
entially binds SMAD1/​5/​8 resulting in the increased expression of muscle growth genes and/​
or the attenuation of myostatin-​induced gene expression (15). Furthermore, animals in which
SMAD1/​5/​8 are constitutively activated are resistant to myostatin-​induced muscle atrophy
suggesting that BMP signaling is dominant over myostatin signaling. Together, these studies
demonstrate that the ability of myostatin to regulate muscle growth is dependent on competi-
tion for SMAD4 binding.
Another level of control over myostatin signaling is mediated by SMAD7, which is able
to block the phosphorylation of SMAD2/​3 by competing with the receptor, blocking
access to the SMAD proteins (63). Reciprocal autoregulation is also evident as SMAD7
mRNA expression is induced by myostatin, whereas myostatin mRNA is decreased by
SMAD7 in skeletal muscle (Figure 22.1B) (63). Taken together, these findings describe the
highly redundant, complex, and elegant regulation of myostatin signaling in the growth and
development of healthy skeletal muscle and provide clues about its dysregulation in muscle
wasting disease states.

313
314

D.S. Hittel

Resistance exercise, size, and strength

Myostatin inhibition was initially considered to be a potential resistance exercise mimetic
because of the hypermuscular phenotype it generates across multiple species (3). However, it
soon became apparent that muscle from engineered myostatin knockout animals was of poor
quality, exhibiting compromised force production, fewer mitochondria, and weaker tendons (5,
36). Notwithstanding these observations in mice, early human studies found that both acute
and chronic resistance exercise elicited significant decreases in skeletal muscle myostatin mRNA
levels (Figure 22.1A) (46). For example, myostatin mRNA decreased by 2.2-​fold in muscle
biopsies taken 4 hours after an acute bout of knee extension resistance training and remained
44% below baseline levels in biopsies taken at 24 hours (43). Contradicting these findings,
muscle myostatin mRNA, as well as muscle and serum protein levels, were all increased acutely
(~15 minutes) after the last bout of a resistance training program.The reason for the discrepancy
may reflect variations in/​lack of standardization of the training regimens for gene and protein
expression studies, as well as differences in biopsy sampling time. Further muddying the waters,
a positive correlation between skeletal muscle mass and both local and circulating myostatin
levels, the so-​called myostatin paradox, is counterintuitive for a negative muscle regulator and
likely reflects the larger pool of protein-​bound inactive and active myostatin needed to maintain
a larger muscle mass (17). Given the extensive post-​translational regulation of myostatin, others
have attempted to determine whether muscle and circulating protein levels are changed by
resistance training. For instance, Walker et al. showed that plasma myostatin protein levels were
decreased by approximately 20% following 10 weeks of elbow flexor exercise (54) and Saremi
and coworkers identified a modest but significant 10% decrease in serum myostatin levels
following 12 weeks of arm-​and leg-​press resistance training (48). However, Kim et al. reported
no change in serum myostatin levels following 16 weeks of knee extensor resistance training
and reported high variability in serum myostatin levels even between untrained subjects (25).
Notably, the generation of specific antibodies against GDF8/​myostatin has been hampered
due to cross-​reactivity with other TGF-​β family members such as the highly homologous
GDF-​11.This may account for the high degree of variability in the immunodetection of mature
and latent forms of myostatin in serum, plasma and tissue homogenates. This may also account
for the withdrawal of the ActRIIB receptor-​Fc ligand-​trap from phase II trials given that most
attempts to block myostatin have been hampered by the cross-​reactivity with structurally related
TGF-​β superfamily growth factors (11). The differences between these studies may also reflect
the confounding effects of myostatin efflux and/​or clearance from the systemic circulation
between subjects or between resistance exercise paradigms (3).Thus, it does not appear that cir-
culating myostatin levels alone can be used as a reliable biomarker for training status nor does it
accurately reflect the mass of exercised muscle. Moreover, the exact role of changes in myostatin
expression in response to resistance exercise is not clear. But in a cluster analysis on resistance-​
trained humans, while myostatin mRNA levels significantly decreased in muscle biopsies 24
hours after a single knee extensor resistance exercise bout and 24 hours after completing the
final bout of a 16-​week training regimen, the decrease was the same for non-​, modest, and
extreme responders with respect to changes in muscle fiber size, suggesting that the magnitude
of decrease in myostatin expression was not associated with greater hypertrophic growth in
response to resistance training (25). Thus, it is not clear to what extent decreases in myostatin
expression may be a prerequisite for the muscle fiber hypertrophy accompanying resistance
training in humans. While myostatin signaling clearly plays an important role in muscle growth
and development, its role in load-​induced muscle growth remains unclear. As described above,
myostatin mRNA levels do not predict load-​induced skeletal muscle hypertrophy. Highlighting

314
315

Myostatin and muscle size and strength

yet another level of control over myostatin signaling, the laboratory of Barr and colleagues show
the transcriptional inhibitor Notch is activated following resistance exercise and translocates to
the nucleus where it prevents the transcriptional repression of SMAD2/​3. In this way, SMAD
activity can be regulated locally within the exercised muscle, resulting in a muscle-​specific signal
for load-​induced skeletal muscle hypertrophy (29).
Recently, Sharples and colleagues investigated genome-​wide DNA methylation and gene
expression after muscle hypertrophy followed by detraining and later reloading (51). They
discovered increased frequency of hypomethylation across the genome after reloading versus
earlier loading that corresponded to increased gene expression, suggesting that muscle maintains
an epigenetic “memory” of hypertrophy. Although myostatin was not among these genes, nor
has it been shown to directly influence their expression, Fan et al. determined that sulforaphane,
a compound that is abundant in cruciferous vegetables, is a histone deacetylase and a DNA
methyltransferase inhibitor, and significantly represses myostatin expression. This epigenetic
repression of myostatin was attributed in part to hypoacetylation of the MyoD binding sites in
the myostatin promoter (14). Another epigenetic regulator of myostatin is SIRT6, a member
of the sirtuin family of protein deacetylases that regulate gene expression in response to the
energy status of the cell. Interestingly, SIRT6 knockout mice display a myostatin-​dependent
loss of muscle, fat, and bone density, typical characteristics of cachexia and starvation. It was
subsequently shown that SIRT6 depletion in cardiac and skeletal muscle cells promotes
myostatin expression. Taken together, these findings suggest that SIRT6 may mediate the link
between global nutritional status and skeletal muscle mass through the epigenetic regulation of
myostatin (47).
The relationship between myostatin and muscle strength remains equivocal. The ratio of
amino acid-​changing versus nonchanging variants in the human myostatin gene are greater
than expected, suggesting that they have been subject to recent positive selection. This, in turn,
suggests that human variations in the myostatin gene are associated with functional differences
(49). This is somewhat borne out in the observation that myostatin variants are associated
with baseline muscle strength among African Americans (27). However, other variants in
the myostatin gene are associated with extreme longevity, lower muscle strength, and obesity,
suggesting that the selective pressure may be thrifty rather than strength and size promoting
(49). Indeed, the high energetic cost to maintain a large muscle mass may explain, in part, the
infrequency of inactivating myostatin mutations in the human population. For instance, one
of the only humans with a confirmed natural myostatin-​null genotype exhibited a muscle
mass twice that of sex-​and age-​matched controls (50). Although this individual was otherwise
healthy, heterozygotes in his pedigree also exhibited increased muscle size and strength, but to
a lesser degree.
Finally, genetic studies of the canine myostatin gene by Mosher et al. have shown it to be
responsible for the double-​muscling phenotype seen in “bully” whippets (39). Furthermore, the
loss of one functional myostatin allele confers a competitive advantage to racing whippets (over
~300 metres), providing evidence that loss of myostatin function can enhance athletic perform-
ance. The degree to which loss of function mutations in myostatin contributes to elite athletic
performance phenotypes in humans will become apparent in the current era of population-​
scale genome sequencing.

Endurance exercise
Studies of the relationship between myostatin and endurance exercise adaptation have generally
demonstrated a decrease in gene and protein expression following exercise (Figure 22.1A) (3).

315
316

D.S. Hittel

For example, 12 weeks of cycling endurance exercise training decreased myostatin mRNA by
50% in human muscle biopsies (26). In addition, muscle and plasma myostatin protein levels
decreased after 6 months of low-​intensity aerobic training in prediabetic subjects (23). Myostatin
mRNA levels also decrease in response to a single bout of endurance exercise training. For
instance, myostatin mRNA levels were decreased approximately three-​to fourfold in gastrocne-
mius biopsies after a single bout of running for 30 minutes at 75% VO2max (28). High-​intensity
interval training (HIIT) can elicit similar adaptations to endurance exercise and may offer a
time-​saving alternative to traditional endurance exercise (16). Early evidence seems to indicate
that HIIT exercises result in lower myostatin mRNA to levels similar to those seen after longer
endurance training; however, more studies are needed to support these findings (16, 42). Taken
together, these results suggest that a decrease in myostatin expression in humans is a hallmark of
both acute and chronic endurance exercise.
Although muscle from myostatin-​ null animals is grossly hypertrophied, it exhibits
compromised force production and fewer mitochondria (5, 36). It was therefore not sur-
prising that myostatin null mice show reduced involuntary treadmill time to exhaustion
compared to wild-​type mice (30). One possible reason for decreased endurance perform-
ance in myostatin null mice is a greater percentage of fatigable fast-​twitch glycolytic fibers
compared to wild-​type animals (18). Despite this, muscle from myostatin null mice were
able to increase mitochondrial enzymes in response to run and swim training (30) showing
that myostatin inactivation does not adversely affect the metabolic adaptations accompanying
endurance exercise training (3).
In summary, a decrease in myostatin mRNA and protein levels in muscle is a characteristic
of endurance exercise training in humans and rodents regardless of the exercise modality, inten-
sity, or dose. However, germline inactivation of myostatin limits aerobic exercise performance
in mice, possibly due to a larger proportion of glycolytic fibers, while myostatin inactivation
combined with exercise training in older adult mice appears to increase endurance exercise
performance. Together, these findings indicate that myostatin plays a central role in integrating
anabolic and catabolic responses to exercise (Figure 22.1B).

Obesity, insulin resistance, and diabetes

There is mounting evidence that myostatin has effects on metabolism that are associated with
obesity and insulin resistance (3). For instance, muscle and circulating myostatin protein levels
are increased in obese human subjects, as is myostatin secretion from myotubes derived from
myoblasts isolated from muscle biopsies of obese compared to nonobese women (24). Conversely,
myostatin mRNA levels decreased in muscle biopsies from obese human patients following
weight loss due to either biliopancreatic diversion or gastric bypass surgery (3). In addition, the
manipulation of myostatin signaling can dramatically impact the development of obesity in mice.
For instance, muscle-​specific overexpression of myostatin caused a decrease in muscle mass and
an increase in epididymal fat pad mass similar to sarcopenic obesity (44).This may be explained,
in part, by the role of myostatin in the regulation of fat mass. Following the pioneering work of
Allen et al. (2), others have identified the expression of myostatin and ActIIRB mRNA in adi-
pose tissue, although it has not been established if adipose-​derived myostatin is secreted into the
circulating endocrine pool of latent and active myostatin. That being said, possessing an intact
activin signaling pathway predisposes adipose cells and tissue to regulation by myostatin. For
instance, myostatin stimulates 3T3-​L1 adipocyte proliferation by increasing the expression of
cell-​proliferation-​related genes and yet inhibits their differentiation resulting in deceased lipid

316
317

Myostatin and muscle size and strength

accumulation and promoting the expression of lipolytic enzymes (62). Further, myostatin has
recently been shown to inhibit the differentiation of brown preadipocytes and that the amount
of subcutaneous fat and leaf fat of myostatin null pigs were significantly decreased mainly due
to the browning of subcutaneous adipose (8). Therefore, excess myostatin signaling promotes
both the loss of muscle mass and accumulation of poor quality adipose tissue resulting in a
diabetogenic metabolic profile.
Conversely, many studies have shown that inactivation of myostatin can prevent the devel-
opment of obesity in mice. For instance, crossing myostatin null mice with genetically obese
ob/​ ob mice attenuated the increased adipose mass, hyperglycemia, hyperlipidemia, and
hyperinsulinemia typically observed in these animals (35). Whereas, postnatal injection of
mice with the ActRIIB-​Fc ligand trap increased muscle mass and decreased fat mass in mice
fed a high-​fat diet (1). Taken together, these studies clearly demonstrate that the inhibition
of myostatin can greatly arrest or reverse the development of obesity and its adverse health
consequences in mice.
Recent evidence suggests that myostatin may play a role in the development of diabetes
in addition to, and perhaps independent of, its effect on obesity. For instance, transcriptomic
analysis revealed that myostatin mRNA levels were elevated in skeletal muscle biopsies from
type 2 diabetics, as well as from nonobese hyperinsulinemic relatives of type 2 diabetics (41). In
addition, both muscle and plasma myostatin protein levels in insulin-​resistant middle-​aged men
were decreased by aerobic exercise training and were strongly correlated with insulin sensitivity
(23). These subjects experienced decreases in circulating myostatin, insulin, and glucose levels
that were not accompanied by any change in fat mass or body mass index following exercise
training, strongly suggesting that these changes were not secondary to a change in adipose mass.
Furthermore, Bernardo et al. (7) showed improvements in fed and fasted blood glucose levels
without any change in adipose mass in ob/​ob mice injected for 6 weeks with a neutralizing
antibody to myostatin. Lastly, injection of recombinant myostatin decreased insulin sensitivity
in healthy male mice and in mice harbouring a loss-​of-​function mutation to myostatin on a
high-​fat diet without a corresponding change in body mass in either study (23, 60). Together
these studies indicate that increased myostatin expression is inversely related to insulin sensi-
tivity independent of obesity status.

Atrophy and cachexia

Loss of muscle mass is observed after trauma, with neuromuscular disorders and catabolic
diseases such as cancer, diabetes, renal failure, cardiorespiratory insufficiency, and sepsis (53).
Furthermore, muscle atrophy also occurs with starvation, immobilization, microgravity, and
sarcopenia, the age-​related loss of skeletal muscle that is associated with frailty (Figure 22.1B)
(38). Regardless of the pathophysiological or environmental causes, myostatin-​ associated
wasting of skeletal muscle results from both decreased protein synthesis and increased ubiquitin-​
mediated protein degradation of contractile proteins (52). Myostatin also inhibits paired box 7
(PAX7) expression resulting in the inhibition of myogenic precursor cell proliferation and sat-
ellite cell renewal and thereby blunting the cellular basis for muscle regeneration. Myostatin also
induces the AKT-​mediated activation of the transcription factor forkhead box O1 (FOXO1)
which in turn, increases the expression of the ubiquitin E3 ligases atrogin-​1 and MuRF1.
This results in the ubiquination and degradation of sarcomeric proteins when combined with
myostatin/​mTORC1-​associated suppression of muscle structural protein expression results in
the rapid loss of skeletal muscle mass (Figure 22.1B).

317
318

D.S. Hittel

Cachexia is a multifactorial syndrome associated with chronic disease that is characterized

by skeletal muscle wasting with associated asthenia (loss of muscle strength), loss of body fat,
and a dramatic catabolic shift in lipid, carbohydrate, and protein metabolism (52, 53). Cancer-​
cachexia is exhibited by about 80% of all patients who possess advanced tumors of pancre-
atic, lung, and gastric origin. Not surprisingly, elevated levels of circulating myostatin correlate
with the progression of cancer-​associated cachexia in part because several tumor cell lines
produce and secrete myostatin (53). Given the highly visible and personally devastating effects
of cancer cachexia, it is a high priority target for myostatin-​neutralizing therapies in humans
(Figure 22.1A).
Myostatin upregulation is observed in muscle of hypoxemic patients with severe chronic
obstructive pulmonary disease (COPD). Myostatin is similarly increased with hypoxia-​induced
atrophy in rats and in human myotubes treated with hypoxia-​mimicking agents (45). It is there-
fore not surprising that approximately 40% of people with COPD eventually develop signifi-
cant muscle wasting with a dramatic decline in oxidative muscle phenotype of the remaining
muscle, resulting in reduced energy output loss of function (52). In addition, muscle myostatin-​
mediated expression of the ubiquitin proteasome E3 ligases atrogin-​1 and MuRF-​1 have been
show to be the primary drivers of muscle wasting in COPD patients (Figure 22.1B) compared
with healthy controls suggesting a mechanistic link that could be targeted with antimyostatin
therapeutics.
Chronic inflammation of the liver caused by alcohol, diet, or disease eventually results in
cirrhosis and loss of skeletal muscle mass secondary to hyperammonemia (12). As with cancer
and COPD, mounting evidence implicates myostatin is a key factor involved in muscle wasting
seen in patients with liver disease. Indeed, higher levels of circulating myostatin were noted in
patients with end-​stage liver disease and in skeletal muscle of patients with cirrhosis (39).
As with cancer, COPD, and liver disease, one of the comorbidities associated with chronic
kidney disease is the rapid onset of muscle wasting (52). Increased protein breakdown, insulin
resistance, excessive glucocorticoid production, and a significant increase in the expression of
muscle myostatin track with the loss of muscle mass in these patients. Subsequent pharma-
cological inhibition of myostatin reversed the loss of body weight and muscle mass in mice,
making chronic kidney disease an ideal target for myostatin-​neutralizing therapies in humans.

Myostatin inhibitors
Although the targeted inhibition of myostatin has been successfully used to increase muscle
growth in food animals and in the burgeoning “clean meat” market, its therapeutic use in
humans faces considerable technical and ethical challenges. For instance, several myostatin-​
neutralizing biological agents have been withdrawn due to safety concerns and failure to meet
primary clinical endpoints (9).This is due in part to the cross-​reactivity of myostatin antagonists
with structurally related growth factors. While this underscores the challenges faced when
developing new drugs, a number of recent studies may provide new insights for the design of
the next generation of myostatin inhibitors. First, to investigate the molecular mechanisms by
which pro-​myostatin remains latent, Cotton et al. have resolved the structure of unprocessed
pro-​myostatin and analyzed the properties of the protein in its latent and active forms (11).
A high-​resolution structure of myostatin may guide the development of inhibitors with lower
cross-​reactivity and higher specificity. Second, blockade of individual ActRII receptors using
anti-​ActRIIA or anti-​ActRIIB antibodies has been shown to produce only a small (~15%)
increase in muscle mass in animals (38). However, Morvan et al. demonstrate that maximal
anabolic/​hypertrophic response is achieved with simultaneous blockade of both receptors (38).

318
319

Myostatin and muscle size and strength

Third, an alternative spliced 129-​amino acid isoform of avian myostatin inhibits proteolytic
processing of pro-​myostatin, and thereby reduces the release of mature myostatin in cell culture
and in vivo (9). Myostatin inhibition using such peptides is an emerging area of interest in both
agricultural animals to increase meat production as well as human medicine (Figure 22.1A).
Although myostatin inhibitors are a promising new class of therapeutics for the treatment
of muscle wasting, they are also very tempting as performance-​enhancing drugs and cosmetic
agents. While no myostatin inhibitors have obtained clinical approval, the proactive develop-
ment of detection methods for emerging doping agents represent a key aspect of prevention.
For instance, Bimagrumab (Novartis) is a human anti-​ActRII antibody which was found to
increase muscle mass and function by blocking ActRII signaling. As it has considerable potential
for being misused as a doping agent in sports, a rapid mass spectrometric detection assay has been
developed in concert with the drug manufacturer to discourage use and abuse by athletes (57).

Summary
Since the discovery of myostatin over 20 years ago, considerable progress has been made in
understanding its impact on the growth and maintenance of skeletal muscle mass. In keeping
with its negative role in myogenesis, myostatin expression is highly regulated at several levels
including epigenetic, transcriptional, post-​transcriptional, and post-​translational. Mutations
in and around the myostatin gene have taught us much about the role of lean body mass
in combating obesity as well as the neuroendocrine response to exercise training and dis-
ease. Recent insights about the structure and regulation of myostatin have revealed novel
exploits for the development of the next generation of myostatin antagonists to combat
muscle wasting disease and cachexia. Finally, besides its impact on muscle, myostatin plays
an important role in metabolism and as a facilitator of muscle–​organ crosstalk in both in
health and disease. Future studies of myostatin will tease out the molecular mechanisms by
which myostatin balances anabolic and catabolic responses of skeletal muscle in both health
and disease.

References
1. Akpan I, Goncalves MD, Dhir R et al. The effects of a soluble activin type IIB receptor on obesity
and insulin sensitivity. Int J Obes (Lond). 2009;33(11):1265–​73.
2. Allen DL, Cleary AS, Speaker KJ et al. Myostatin, activin receptor IIb, and follistatin-​like-​3 gene
expression are altered in adipose tissue and skeletal muscle of obese mice. Am J Physiol Endocrinol
Metab. 2008;294(5):E918–​27.
3. Allen DL, Hittel DS, McPherron AC. Expression and function of myostatin in obesity, diabetes,
and exercise adaptation. Med Sci Sports Exerc. 2011;43(10):1828–​35.
4. Amthor H, Macharia R, Navarrete R et al. Lack of myostatin results in excessive muscle growth
but impaired force generation. Proc Natl Acad Sci U S A. 2007;104(6):1835–​40.
5. Amthor H, Macharia R, Navarrete R et al. Lack of myostatin results in excessive muscle growth
but impaired force generation. Proc Natl Acad Sci U S A. 2007;104(6):1835–​40.
6. Bellinge RH, Liberles DA, Iaschi SP, O’Brien PA, Tay GK. Myostatin and its implications on
animal breeding: a review. Anim Genet. 2005;36(1):1–​6.
7. Bernardo BL, Wachtmann TS, Cosgrove PG et al. Postnatal PPARdelta activation and myostatin
inhibition exert distinct yet complimentary effects on the metabolic profile of obese insulin-​resistant
mice. PloS One. 2010;5(6):e11307.
8. Cai C, Qian L, Jiang S et al. Loss-​of-​function myostatin mutation increases insulin sensitivity and
browning of white fat in Meishan pigs. Oncotarget. 2017;8(21):34911–​22.
9. Chen PR, Lee K. Invited review. Inhibitors of myostatin as methods of enhancing muscle growth
and development. J Anim Sci. 2016;94(8):3125–​34.

319
320

D.S. Hittel

10. Clop A, Marcq F, Takeda H et al. A mutation creating a potential illegitimate microRNA target
site in the myostatin gene affects muscularity in sheep. Nat Genet. 2006;38(7):813–​8.
11. Cotton TR, Fischer G, Wang X et al. Structure of the human myostatin precursor and determinants
of growth factor latency. EMBO J. 37: 367–​383, 2018.
12. Dasarathy S. Myostatin and beyond in cirrhosis: all roads lead to sarcopenia. J Cachexia Sarcopenia
Muscle. 2017;8(6):864–​9.
13. Dschietzig TB. Myostatin –​from the mighty mouse to cardiovascular disease and cachexia. Clin
Chim Acta. 2014;433:216–​24.
14. Fan H, Zhang R, Tesfaye D et al. Sulforaphane causes a major epigenetic repression of myostatin
in porcine satellite cells. Epigenetics. 2012;7(12):1379–​90.
15. Gaarenstroom T, Hill CS. TGF-​beta signaling to chromatin: how Smads regulate transcription
during self-​renewal and differentiation. Semin Cell Dev Biol. 2014;32:107–​18.
16. Gillen JB, Gibala MJ. Is high-​intensity interval training a time-​efficient exercise strategy to improve
health and fitness? Appl Physiol Nutr Metab. 2014;39(3):409–​12.
17. Gilson H, Schakman O, Combaret L et al. Myostatin gene deletion prevents glucocorticoid-​
induced muscle atrophy. Endocrinology. 2007;148(1):452–​60.
18. Girgenrath S, Song K, Whittemore LA. Loss of myostatin expression alters fiber-​type distribution
and expression of myosin heavy chain isoforms in slow-​and fast-​type skeletal muscle. Muscle Nerve.
2005;31(1):34–​40.
19. Girgenrath S, Song K, Whittemore LA. Loss of myostatin expression alters fiber-​type distribution
and expression of myosin heavy chain isoforms in slow-​and fast-​type skeletal muscle. Muscle Nerve.
2005;31(1):34–​40.
20. Goodman CA, Hornberger TA. New roles for Smad signaling and phosphatidic acid in the regu-
lation of skeletal muscle mass. F1000Prime Rep. 2014;6:20.
21. Hammers DW, Merscham-​Banda M, Hsiao JY, Engst S, Hartman JJ, Sweeney
HL. Supraphysiological levels of GDF11 induce striated muscle atrophy. EMBO Mol Med.
2017;9(4):531–​44.
22. Hill JJ, Davies MV, Pearson AA et al.The myostatin propeptide and the follistatin-​related gene are
inhibitory binding proteins of myostatin in normal serum. J Biol Chem. 2002;277(43):40735–​41.
23. Hittel DS, Axelson M, Sarna N, Shearer J, Huffman KM, Kraus WE. Myostatin decreases with
aerobic exercise and associates with insulin resistance. Med Sci Sports Exerc. 2010;42(11):2023–​9.
24. Hittel DS, Berggren JR, Shearer J, Boyle K, Houmard JA. Increased secretion and expression of
myostatin in skeletal muscle from extremely obese women. Diabetes. 2009;58(1):30–​8.
25. Kim JS, Petrella JK, Cross JM, Bamman MM. Load-​mediated downregulation of myostatin
mRNA is not sufficient to promote myofiber hypertrophy in humans: a cluster analysis. J Appl Physiol
(1985). 2007;103(5):1488–​95.
26. Konopka AR, Douglass MD, Kaminsky LA et al. Molecular adaptations to aerobic exercise
training in skeletal muscle of older women. J Gerontol A Biol Sci Med Sci. 2010;65(11):1201–​7.
27. Kostek MA, Angelopoulos TJ, Clarkson PM et al. Myostatin and follistatin polymorphisms
interact with muscle phenotypes and ethnicity. Med Sci Sports Exerc. 2009;41(5):1063–​71.
28. Louis E, Raue U, Yang Y, Jemiolo B, Trappe S. Time course of proteolytic, cytokine, and
myostatin gene expression after acute exercise in human skeletal muscle. J Appl Physiol (1985).
2007;103(5):1744–​51.
29. Marcotte GR, West DW, Baar K. The molecular basis for load-​induced skeletal muscle hyper-
trophy. Calcif Tissue Int. 2015;96(3):196–​210.
30. Matsakas A, Mouisel E, Amthor H, Patel K. Myostatin knockout mice increase oxidative muscle
phenotype as an adaptive response to exercise. J Muscle Res Cell Motil. 2010;31(2):111–​25.
31. McFarlane C, Langley B, Thomas M et al. Proteolytic processing of myostatin is auto-​regulated
during myogenesis. Dev Biol. 2005;283(1):58–​69.
32. McPherron AC, Huynh TV, Lee SJ. Redundancy of myostatin and growth/​differentiation factor
11 function. BMC Dev Biol. 2009;9:24.
33. McPherron AC, Lawler AM, Lee SJ. Regulation of skeletal muscle mass in mice by a new TGF-​
beta superfamily member. Nature. 1997;387(6628):83–​90.
34. McPherron AC, Lee SJ. Double muscling in cattle due to mutations in the myostatin gene. Proc Natl
Acad Sci U S A. 1997;94(23):12457–​61.
35. McPherron AC, Lee SJ. Suppression of body fat accumulation in myostatin-​deficient mice. J Clin
Invest. 2002;109(5):595–​601.

320
321

Myostatin and muscle size and strength

36. Mendias CL, Bakhurin KI, Faulkner JA.Tendons of myostatin-​deficient mice are small, brittle, and
hypocellular. Proc Natl Acad Sci U S A. 2008;105(1):388–​93.
37. Mendias CL, Marcin JE, Calerdon DR, Faulkner JA. Contractile properties of EDL and soleus
muscles of myostatin-​deficient mice. J Appl Physiol. 2006;101(3):898–​905.
38. Morvan F, Rondeau JM, Zou C et al. Blockade of activin type II receptors with a dual anti-​
ActRIIA/​IIB antibody is critical to promote maximal skeletal muscle hypertrophy. Proc Natl Acad Sci
U S A. 2017;114(47):12448–​53.
39. Mosher DS, Quignon P, Bustamante CD et al. A mutation in the myostatin gene increases muscle
mass and enhances racing performance in heterozygote dogs. PLoS Genet. 2007;3(5):e79.
40. Mosher DS, Quignon P, Bustamante CD et al. A mutation in the myostatin gene increases muscle
mass and enhances racing performance in heterozygote dogs. PLoS Genet. 2007;3(5):e79.
41. Palsgaard J, Brons C, Friedrichsen M et al. Gene expression in skeletal muscle biopsies from
people with type 2 diabetes and relatives: differential regulation of insulin signaling pathways. PloS
One. 2009;4(8):e6575.
42. Pugh JK, Faulkner SH, Jackson AP, King JA, Nimmo MA. Acute molecular responses to con-
current resistance and high-​intensity interval exercise in untrained skeletal muscle. Physiol Rep.
2015;3(4).
43. Raue U, Slivka D, Jemiolo B, Hollon C, Trappe S. Myogenic gene expression at rest and after
a bout of resistance exercise in young (18–​30 yr) and old (80–​89 yr) women. J Appl Physiol (1985).
2006;101(1):53–​9.
44. Reisz-​Porszasz S, Bhasin S, Artaza JN et al. Lower skeletal muscle mass in male trans-
genic mice with muscle-​ specific overexpression of myostatin. Am J Physiol Endocrinol Metab.
2003;285(4):E876–​88.
45. Rodriguez J, Vernus B, Chelh I et al. Myostatin and the skeletal muscle atrophy and hypertrophy
signaling pathways. Cell Mol Life Sci. 2014;71(22):4361–​71.
46. Roth SM, Martel GF, Ferrell RE, Metter EJ, Hurley BF, Rogers MA. Myostatin gene expression
is reduced in humans with heavy-​resistance strength training: a brief communication. Exp Biol Med
(Maywood). 2003;228(6):706–​9.
47. Samant SA, Kanwal A, Pillai VB, Bao R, Gupta MP. The histone deacetylase SIRT6 blocks
myostatin expression and development of muscle atrophy. Sci Rep. 2017;7(1):11877.
48. Saremi A, Gharakhanloo R, Sharghi S, Gharaati MR, Larijani B, Omidfar K. Effects
of oral creatine and resistance training on serum myostatin and GASP-​ 1. Mol Cell Endocrinol.
2010;317(1–​2):25–​30.
49. Saunders MA, Good JM, Lawrence EC, Ferrell RE, Li WH, Nachman MW. Human adaptive
evolution at Myostatin (GDF8), a regulator of muscle growth. Am J Hum Genet. 2006;79(6):1089–​97.
50. Schuelke M, Wagner KR, Stolz LE et al. Myostatin mutation associated with gross muscle hyper-
trophy in a child. N Engl J Med. 2004;350(26):2682–​8.
51. Seaborne RA, Strauss J, Cocks M et al. Human skeletal muscle possesses an epigenetic memory of
hypertrophy. Sci Rep. 2018;8(1):1898.
52. Sharma M, McFarlane C, Kambadur R, Kukreti H, Bonala S, Srinivasan S. Myostatin: expanding
horizons. IUBMB Life. 2015;67(8):589–​600.
53. Smith RC, Lin BK. Myostatin inhibitors as therapies for muscle wasting associated with cancer and
other disorders. Curr Opin Support Palliat Care. 2013;7(4):352–​60.
54. Walker KS, Kambadur R, Sharma M, Smith HK. Resistance training alters plasma myostatin but
not IGF-​1 in healthy men. Med Sci Sports Exerc. 2004;36(5):787–​93.
55. Walker RG, McCoy JC, Czepnik M et al. Molecular characterization of latent GDF8 reveals
mechanisms of activation. Proc Natl Acad Sci U S A. 2018.
56. Walker RG, Poggioli T, Katsimpardi L et al. Biochemistry and biology of GDF11
and myostatin: similarities, differences, and questions for future investigation. Circ Res.
2016;118(7):1125–​41.
57. Walpurgis K, Thomas A, Schanzer W, Thevis M. Myostatin inhibitors in sports drug
testing: detection of myostatin-​neutralizing antibodies in plasma/​serum by affinity purification and
Western blotting. Proteomics Clin Appl. 2016;10(2):195–​205.
58. Walton KL, Johnson KE, Harrison CA. Targeting TGF-​beta mediated SMAD signaling for the
prevention of fibrosis. Front Pharmacol. 2017;8:461.
59. Wharton K, Derynck R. TGFbeta family signaling: novel insights in development and disease.
Development. 2009;136(22):3691–​7.

321
322

D.S. Hittel

60. Wilkes JJ, Lloyd DJ, Gekakis N. Loss-​of-​function mutation in myostatin reduces tumor necrosis
factor alpha production and protects liver against obesity-​ induced insulin resistance. Diabetes.
2009;58(5):1133–​43.
61. Wolfman NM, McPherron AC, Pappano WN et al. Activation of latent myostatin by the BMP-​1/​
tolloid family of metalloproteinases. Proc Natl Acad Sci U S A. 2003;100(26):15842–​6.
62. Zhu HJ, Pan H, Zhang XZ et al. The effect of myostatin on proliferation and lipid accumulation
in 3T3-​L1 preadipocytes. J Mol Endocrinol. 2015;54(3):217–​26.
63. Zhu X, Topouzis S, Liang LF, Stotish RL. Myostatin signaling through Smad2, Smad3 and Smad4
is regulated by the inhibitory Smad7 by a negative feedback mechanism. Cytokine. 2004;26(6):262–​72.
64. Zimmers TA, Davies MV, Koniaris LG et al. Induction of cachexia in mice by systemically
administered myostatin. Science. 2002;296(5572):1486–​8.

322
323

23
Α LPHA-​ACTININ-​3 ’S ROLE
IN THE GENETIC CONTROL
OF MUSCLE STRENGTH AND
PERFORMANCE
Jane T. Seto, Fleur C. Garton, Kathryn N. North, and
Peter J. Houweling

Introduction
The α-​actinin-​3 gene (ACTN3) R577X polymorphism is one of the most highly studied gen-
etic variations associated with human skeletal muscle function and performance. More than 1
billion people worldwide are completely deficient in α-​actinin-​3 due to homozygosity for the
null polymorphism (ACTN3 577XX). Deficiency of α-​actinin-​3 is associated with reduced
muscle force and power, and frequency of the ACTN3 577XX genotype is underrepresented
in elite sprint and power athletes compared to population controls. The ACTN3 577X allele
has undergone strong, recent positive selection during modern human migration out of Africa,
suggesting that inheritance of this allele provides a survival advantage for endurance in colder
climes. The combination of human studies and research in the Actn3 knockout (KO) mouse
model to date has solidified the association between ACTN3 R577X and human muscle per-
formance and begun to unravel how ACTN3 modifies fast-​twitch fiber characteristics and
muscle function. In this chapter, we summarize the current research in ACTN3 R577X and
discuss the molecular mechanisms that underlie the consequences of α-​actinin-​3 deficiency,
with new insights into implications for the general population beyond athletic performance –​in
healthy ageing, muscle disuse, and disease.

The α-​actinins and skeletal muscle

The α-​actinins are a family of four actin-​binding proteins (α-​actinin 1–​4) that have evolved
from repeated gene duplication events to perform similar roles in different cell types (35). The
skeletal muscle-​specific isoforms α-​actinin-​2 and -​3 (encoded by ACTN2 and ACTN3) are
highly conserved (81% identical and 91% similar) and are major components of the Z-​disk
within the skeletal muscle contractile apparatus (sarcomeres). Both sarcomeric α-​actinins have
a N-​terminal actin-​binding domain, a rod domain with four spectrin repeats, and a C-​terminal
region containing calcium-​binding EF hand motifs (Figure 23.1) (39). α-​Actinin-​2 is expressed
in all human skeletal muscle fibers, whereas α-​actinin-​3 has evolved a specialized expres-
sion pattern and is only present in type 2 (fast, glycolytic) muscle fibers (50). The sarcomeric
α-​actinins interact with a vast network of structural, signaling, and metabolic proteins associated

323
324

J.T. Seto et al.

Figure 23.1 Localization and structure of the sarcomeric α-​actinins in skeletal muscle. A) The sarco­meric
α‑actinins are found at the Z-disc (black), where they anchor actin-containing thin filaments (grey) from
adjacent sarcomeres. The sarcomeric α-actinins form head to tail dimers to cross-link actin at the Z-disc.
Their structure consists of an actin-binding domain (ABD), a rod domain with four spectrin repeats
(R1-R4) and an EF-hand domain (EF). Image adapted from (24) and (35).

with the Z-​disk (reviewed in 33) and are known to cross-​link and stabilize actin thin filaments
at the Z-​disk during muscle contractions.

The ACTN3 R577X polymorphism is common in humans

The role of sarcomeric α-​actinins in the maintenance of ordered myofibrillar arrays in skeletal
muscle make them prime candidate genes for human muscle disease. In 1999, we identified a
common null polymorphism in ACTN3 (a C>T transition in exon 16), which converts an
arginine (R) to a premature stop (X) codon at residue 577 (R577X; rs1815739). Homozygosity
for the X-​allele results in the complete loss of α-​actinin-​3 but is not associated with overt
muscle disease (50, 51).
To determine the frequency of the ACTN3 577X allele in the general population, we
initially genotyped 485 DNA samples covering four main diversity groups of Asia/​Americas,
Australasia, Africa, and Europe. The X-​allele frequency differs throughout the world, ranging
from 10% in African populations to 45% in European and 54% in Asian populations (41).
We estimate that approximately 18% of the world’s population are completely deficient in α-​
actinin-​3.With a population size of about 7.6 billion people (in 2017), this equates to about 1.4
billion individuals worldwide being completely deficient in α-​actinin-​3. The high incidence of

324
325

Alpha-actinin-3’s role in muscle strength

ACTN3 577XX globally suggests that α-​actinin-​3 is potentially redundant –​i.e., many of its
functions in muscle can be compensated for by the closely related protein α-​actinin-​2. However,
the limited expression of α-​actinin-​3 in fast, glycolytic muscle fibers and sequence differences
from α-​actinin-​2 suggest that α-​actinin-​3 may have evolved to play a specialized role in fast
fibers that cannot be completely compensated for by α-​actinin-​2.

ACTN3 577X has undergone strong, recent positive selection

The varied allele frequencies observed between the different ethnic groups and specialized
protein expression in fast glycolytic muscle fibers suggest that selective forces may be acting
on the ACTN3 577X allele during modern human evolution. To explore this, we analyzed
ACTN3 DNA sequence and long-​range linkage disequilibrium data around the R577X allele
in individuals of European, East Asian, and African ancestry (41). Our analyses found low rates
of DNA substitutions and high long-​range linkage disequilibrium for recombination among
X-​allele-​containing haplotypes compared with the R-​allele in Europeans and Asians, consistent
with strong, recent positive selection of the 577X allele in these populations. This suggests that
the 577X allele provides an advantage to modern humans adapting to the Eurasian environ-
ment, which we estimate to have occurred around 15,000 years ago in Europe and around
30,000 years ago in East Asia. The increase in ACTN3 577XX genotype also correlated with a
higher global latitude gradient and reduced species richness (18), suggesting that environmental
variables related to temperature (cold tolerance), and diet (feast/​famine) may influence the
observed R577X genotype frequencies worldwide.

The absence of α-​actinin-​3 is detrimental to elite sprint/​power performance

Since α-​actinin-​3 shows specialized expression in the fast, glycolytic muscle fibers and deficiency
does not cause a known muscle disease, we hypothesized that deficiency of α-​actinin-​3 impacts
muscle function at the extremes of muscle performance. In collaboration with the Australian
Institute of Sports, we examined the frequency distribution of ACTN3 R577X in elite sprint and
endurance athletes compared to population controls. Elite athletes were defined as those who had
qualified to represent Australia internationally either at the Commonwealth/​Olympic Games or
world championships. A total of 458 elite Australian Caucasian athletes from 14 different sports
and 436 healthy Caucasian controls were genotyped for ACTN3 R577X (74). Analysis of the
ACTN3 R577X genotype in a subset of 107 specialist sprint/​power athletes compared to 194
specialist endurance athletes revealed differences in ACTN3 genotype frequency. Remarkably,
sprint/​power athletes had an extremely low frequency of 577XX (5% compared to 18% in
controls) and a high frequency of RR (50% compared to 30% in controls) (χ2=19.70, P <0.0001)
(Figure 23.2), suggesting that α-​actinin-​3 deficiency is detrimental to elite sprint/​power per-
formance. In contrast, female elite endurance athletes had higher frequencies of 577XX geno-
type compared to controls (χ2=6.15, P <0.05), although this effect was not significant when the
whole population was examined. Importantly, genotype profiles in sprint/​power and endurance
athletes deviated in opposite directions and differed from each other (χ2=19.45, P <0.001) (74).
Our finding that α-​actinin-​3 deficiency is detrimental in elite sprint/​power performance
in Caucasians has been independently replicated in 16 elite athlete populations from around
the world (Table 23.1). These studies show that male and female sprint/​power athletes (across
different sport and backgrounds) have a lower frequency of the ACTN3 577XX genotype
compared to population controls. There are some exceptions, with several studies finding no
association between ACTN3 R577X and strength/​power athletic status or performance (23,

325
326

Table 23.1 Sprint/​power elite athlete association studies

Origin Sports/​events ACTN3 XX genotype % Reference

Athlete vs. controls

Caucasian
Australian Judo, speed skating, swimming, track M: XX: 8% vs. XX: 16% 74
and field, track cycling F: XX 0% vs. 16%

Finnish Track and field M & F: XX 0% vs. 9% 48

Greek Jumping decathlon, throwing, track M & F: XX 16% vs. 18% 52

and field
Sprinters (100–​400 m) M & F: XX 9% vs. 18%

Israeli Sprinters (track 100–​200 m) M & F: XX 14% vs. 18% 15

Sprinters (track 100–​200 m) M & F: XX 16.7% vs. 2
18.4%

Italian Artistic gymnasts M & F: XX 3% vs. 19% 43

Polish Track sprinters, swimming, weight M & F: XX 8% vs. 15% 9

lifting

Spanish Track and field M & F: XX 12% vs. 15% 61

Russian Alpine skiing and jumping, gymnastics, M: XX 6% vs. 16% 13

body building, figure skating, ice F: 6.5% vs. 13%
hockey, power lifting, soccer, speed
skating, swimming, trowing, track
and field, volleyball, weight lifting,
and wrestling

Asian
Japanese Wrestling M: XX 11% vs. 28% 30
International sprinters ≤400 m M & F: XX 10% vs. 26% 44
Track and field M & F: RR vs. XX; OR 31
1.59, 95% CI 1.16–​
2.18; P=0.003

Korean Gymnasts, sprinters, throwers short M: XX 21% vs. 18% 28

distance speed skaters, weightlifters F: XX 11% vs. 18%
and martial arts (Taekwondo)
Weightlifters, sprinters (≤400 m), speed M & F: XX 11.6% vs. 32
skaters (≤1500 m) and swimmers 19.1%
(≤100 m)

Taiwanese Swimmers ≤400 m M: XX 14% vs. 20% 7

F: XX 11% vs. 20%
327

Alpha-actinin-3’s role in muscle strength

Table 23.1 (Cont.)

Origin Sports/​events ACTN3 XX genotype % Reference

Athlete vs. controls
Chinese Sprinter (100–​200 m), cyclists (500 M & F: 5% vs. 34% 76
m), throwers, jumpers, shot put,
weightlifting

African
Jamaican Track and field M & F: XX 3% vs. 2% 64

Nigerian Sprinters (track 100–​400 m) and M & F: 0% vs. 3% 75

jumpers

African Track and field M & F: 2% vs. 4% 64

American
Bodybuilding and power lifting M & F: 7% vs. 16% 59

Figure 23.2 Frequency of the ACTN3 genotype in controls and elite sprint and endurance athletes. Sample
size of each group is shown in parentheses. Female and male athletes who had competed at an Olympic
level are shown as separate groups for both sprint/power and endurance events. Image adapted from (74).

64, 65). In addition, the ACTN3 R577X association does not appear to influence sprint/​power
performance in African athletes. Three African population cohorts (Jamaican and US), two of
which included World Champion/​Olympic medalists, showed low frequencies of the 577XX
genotype in both athletes (2–​7%) and controls (2–​4%) (59, 64, 75). This precludes the identi-
fication of significantly altered frequencies compared to controls but can be interpreted as the
African population being more suited to sprint performance (74).
To date, four meta-​analyses (1, 20, 38, 73) have explored the association between ACTN3
R577X with sprint, power, and endurance phenotypes in athletes, and are unanimous that
the absence of α-​actinin-​3 is detrimental to sprint/​power performance in Caucasian/​Asian
athletes –​demonstrating that the ACTN3 R577X association with sprint/​power perform-
ance is one of the most highly replicable genetic modifiers of muscle strength and power. In
our most recent meta-​analysis which encompassed raw data from 12 independent studies with
similar caliber athletes who competed in similar sprint/​power events, we found strong evidence

327
328

J.T. Seto et al.

for a consistent homozygote effect (RR vs. XX; overall per-​allele odds ratio (OR)=1.4, 95%
confidence interval (CI) 1.3–​1.6; between-​study standard deviation (s.d.) τ=0.12, 95% CI
0.01–​0.31) but there was substantial heterogeneity in the dominance effect (overall OR=0.98,
95% CI 0.73–​1.3; between-​study s.d. τ=0.40, 95% CI 0.18–​0.72), indicative of the effect of
heterozygotes being highly variable (20). Therefore, while the absence of α-​actinin-​3 (ACTN3
XX) is strongly correlated with reduced elite sprint/​power performance, our study indicates
that a nonsingle genetic model can explain the association between ACTN3 R577X and elite
sprint/​power performance due to the variable effect of the heterozygotes.

The Actn3 knockout mouse model

α-​Actinin-​3 deficiency “slows” fast-​twitch muscle fiber characteristics
Our group established an α-​actinin-​3 (Actn3) KO mouse model in order to understand the
mechanisms underlying the positive selection of the 577X allele and the effects of α-​actinin-​3
deficiency on muscle performance. Actn3 KO mice recapitulated human α-​actinin-​3 deficiency
and demonstrated reduced grip strength (7.4% lower), hindlimb muscle force generation (10.9%
lower), and enhanced endurance running performance as compared to wild-​type (WT) controls
(6, 40, 41). Electron microscopy shows normal Z-​disk ultrastructure and sarcomeric organization
in Actn3 KO mouse muscles (Figure 23.3). Unlike humans, α-​actinin-​2 is not expressed in all
muscle fibers in WT mice (45), but α-​actinin-​2 is upregulated and expressed in all fiber types in
Actn3 KO muscles, similar to the expression pattern seen in human ACTN3 XX (Figure 23.4).

Figure 23.3 α-​Actinin-​3 deficiency in skeletal muscle. Skeletal muscle cross sections from mouse (WT,
Actn3 KO) and human immuno-stained for α-actinin-3 (top row) and α-actinin-2 (bottom row). α-Actinin-3
deficiency results in up-regulation of α-actinin-2 so that it is expressed in all muscle fibres. Image adapted
from (41).

328
329

Alpha-actinin-3’s role in muscle strength

Figure 23.4 Z-​disk proteins are upregulated in Actn3 knockout mice. A) α-Actinin-2, desmin, myotilin,
γ-filamin are up-regulated in Actn3 KO muscle compared to WT muscle; α-actin and myosin as shown
as loading controls. B) The expression of the 78kDa isoform of Z-band alternatively spliced PDZ-motif
protein (ZASP) is also up-regulated in Actn3 KO muscles, however expression of the ZASP related, actin-
associated LIM protein (Pdlim3/ALP) is not altered in KO muscles relative to WT control. Myosin is
shown as loading control. Image adapted from (66).

Consistent with reduced muscle force generation, analyses of lean mass by dual-​energy X-​ray
absorptiometry and isolated muscles from WT and Actn3 KO mice showed that KO muscles are
reduced in mass and have smaller fast-​twitch fibers.There was no change in fiber type proportions, but
analysis of the overall muscle cross-​sectional area (CSA) dominated by each fiber type (considering
both fiber size and number) showed reduction in 2B CSA and increased 2X CSA in KO muscles
relative to WT (21, 40, 67). A number of studies in human cohorts found similar genotype effects.
The ACTN3 577XX genotype was associated with reductions in lean mass, fat mass, and BMI
(72), as well as thigh muscle CSA (77). α-​Actinin-​3 deficiency was also found to be associated with
reduced proportions of fast fibers (71), while other studies assessing active individuals found no asso-
ciation between ACTN3 genotype and fiber type distribution or fiber size (19, 49).
Enzymatic activities in the anaerobic and aerobic pathways involved in glucose breakdown
are also different in Actn3 KO muscles relative to WT (40, 41).The activity of anaerobic enzyme
lactate dehydrogenase (LDH) was reduced by approximately 16% in Actn3 KO muscle, while
mitochondrial enzymes citrate synthase, succinate dehydrogenase, and cytochrome c oxidase,
were increased by 22–​39%. Activities of fatty acid oxidation enzymes, β-​hydroxyacyl-​CoA
dehydrogenase and medium chain acyl-​CoA dehydrogenase were also 30–​42% higher in Actn3
KO muscles. These results indicate that deficiency of α-​actinin-​3 causes fast-​twitch muscle
fibers to switch to greater reliance on oxidative rather than glycolytic metabolism, consistent
with characteristics of a slower-​twitch muscle.
The absence of α-​actinin-​3 also altered the twitch characteristics of fast-​twitch muscles.
Assessment of the fast fiber-​r ich, extensor digitorum longus muscles from Actn3 KO and WT
mice showed that Actn3 KO muscles have longer twitch half-​relaxation times and enhanced
force recovery following contraction-​induced muscle fatigue (6). Single muscle fiber analyses
further showed that contractile changes were caused by slower calcium uptake (calcium loading
into sarcoplasmic reticulum) in Actn3 KO fast-​twitch fibers compared to WT fibers (5).Together
with reduced fast-​twitch fiber size and shifts towards oxidative metabolism, these qualities make

329
330

J.T. Seto et al.

α-​actinin-​3-​deficient fast-​twitch muscles less favorable for rapid, repetitive muscle contractions
necessary for sprint/​power generation, but beneficial for endurance performance.

ACTN3 genotype and the response to exercise training

Skeletal muscle is capable of adapting to environmental stresses and physiological demands.
The baseline phenotypes of the Actn3 KO mouse confirm that α-​actinin-​3 plays a key role in
the determination of fast-​twitch fiber size, fiber type CSA, muscle mass, and metabolism, so it
follows that the Actn3 genotype could also influence how muscle adapts to exercise training.
Indeed, elite athletes represent the most well-​trained cohorts with decade/​s of persistent and
specific training –​it is conceivable therefore that the effect of the ACTN3 genotype on elite
athletic performance may be enhanced by training.
Consistent with our hypothesis, the shift in fast-​twitch fiber properties towards slow-​twitch
not only predisposed Actn3 KO mice for endurance performance, but also primed their respon-
siveness to endurance training. In response to a progressive treadmill running program over 4
weeks, where trained WT mice could run two times further than their untrained counterparts
in a maximal incremental run to fatigue (similar to a maximal oxygen consumption (VO2max)
test), trained Actn3 KO mice were able to run 3.3 times further compared to untrained KO
mice (67). Analyses of untrained and trained muscles showed that trained Actn3 KO muscles
had greater 2B to 2X CSA type shifts and higher oxidative enzyme activity compared to trained
WT, consistent with enhanced response to endurance training.
The effect of the ACTN3 genotype on endurance training responsiveness in athletes has so
far not been replicated but improved endurance capacity in the general population has been
observed in a number of studies (Table 23.2). Similarly, the effect of the ACTN3 genotype on
the response to various resistance training programs has been examined. While some found
increased strength gains with 577XX, others observed no clear association with the ACTN3
genotype (Table 23.2). Overall, there appears to be a positive association between the ACTN3
577R allele with higher response to resistance training.This could be reflective of the fast fiber-​
specific expression of α-​actinin-​3 and its role in maintaining muscle strength. The response to
resistance training has yet to be examined in the Actn3 mouse model but future investigation
could shed light on the mechanism for this training effect in humans. An extensive review of
this topic was performed by Garton et al. (19).

α-​Actinin-​3 influences on muscle strength and performance

To properly understand the phenotypic effects of the Actn3 KO mouse model, and indeed,
α-​actinin-​3 deficiency in humans, the presence and function of α-​actinin-​2 in skeletal muscle
must also be considered. Total sarcomeric α-​ actinin content in skeletal muscle is closely
regulated; induced overexpression of one sarcomeric α-​actinin leads to downregulation of the
other (20, 66).The effects of the Actn3 KO mouse and ACTN3 577XX humans must therefore
be interpreted as a consequence of the loss of α-​actinin-​3 and simultaneous upregulation of
α-​actinin-​2 in fast-​twitch muscle fibers.
The diverse phenotypes of the Actn3 KO mouse emphasize the involvement of sarcomeric
α-​actinins in various systems that govern muscle biology, strength, and performance. As
discussed, the sarcomeric α-​actinins are components of a sensor machinery through direct pro-
tein interactions at both the membrane and the Z-​disk (reviewed in 35). Although α-​actinin-​2
and α-​actinin-​3 share 80% identity in amino acid sequence (45), functional divergence and
fiber type specificity between these closely related proteins are likely mediated by differences in

330
331

Table 23.2 ACTN3 genotype and training association studies (endurance and resistance training)

Origin Sex N Tested Significant associations Reference

Age (years)

Endurance training
Brazil M 150 Maximal incremental ACTN3 XX genotype 53
running test, is over-​represented in
ventilatory threshold groups with higher
(VT) and respiratory VT(P <0.0001)
compensation point and respiratory
(RCP) compensation (P
<0.0001)
Estonia M&F 58 5-​year endurance No significant 42
15–​19 41 males cross-​country skiing difference
17 females performance
Brazil M 206 18 weeks of endurance XX show higher VO2 at 69
25 ± 4 running training anaerobic threshold,
RCP, and exercise
peak compared to
RR (P <0.003).
Following training the
differences between
XX and RR were no
longer observed
Resistance training
Spain F 139 2 weeks of high-​speed Significant genotype-​ 56
65.5 + 8.2 power training on training interaction
maximal strength (P <0.05) for
(1RM) of the arm all muscular
and leg muscles, performance indices
muscle power
performance
(counter-​movement
jump), and functional
capacity (sit-​to-​stand
test)
UK M 51 9-​week knee extension ACTN3 R-​allele 14
20.3 + 3.1 resistance training carriers had greater
(RT) quadriceps femoris
muscle volume (Vm),
maximum isoinertial
strength (1RM), and
maximum power
(W max) than XX
homozygotes at
baseline (all P <0.05),
but responses to RT
were independent of
ACTN3 genotype (all
P >0.05)
(continued)
332

Table 23.2 (Cont.)

Origin Sex N Tested Significant associations Reference

Age (years)
Brazil F 246 Quadriceps isokinetic No significant 36
67 strength pre and post difference
24 wks of resistance
training
Brazil M 141 1RM bench press, No significant 22
23–​31 knee extensors peak difference
torque (60°/​s), and
knee extensors before
and after 11 wks of
training
40 Ultrasound muscle RR/​RX increased
thickness before muscle thickness after
and after 11 wks of training
training
Taiwan M 50 12 wks swim training, In young males; RR 7
11–​13 25 m and 100 m had fastest mean
swim time pre and before and after
post, body mass training (no P value).
index, % body fat, in RX and XX greater
young males=serum improvement in 25 m
thyroid, testosterone swim performance vs.
and growth RR genotype
M 38 hormones No significant
21–​30 difference
US M 71 Knee extensor Post-​training, trend for 11
56–​74 concentric peak relative peak power
power before and change in RR was
after 10 wks of higher than XX
unilateral knee (P=0.07)
training
F 86 Higher peak power
64 and1RM in XX
women at baseline
vs. RX/​RR. Post-​
training, relative peak
power change in RR
was higher than the
XX (P <0.05)
333

Alpha-actinin-3’s role in muscle strength

binding affinity or selectivity among interaction partners for a specific α-​actinin isoform, which
could culminate in altered protein complex stability. Based on the known functions of the α-​
actinin interacting partners, the roles of α-​actinin-​3 (and α-​actinin-​2) can be broadly classified
into the following categories: structural, signaling, metabolic, and calcium handling.

α-​Actinin-​3 enhances Z-​disk stability

Assessment of Actn3 WT and KO muscle response to lengthening contractions showed that
α-​actinin-​3 plays a major structural role in the Z-​disk. In response to muscle damage induced
using repeated eccentric contractions at 30% stretch (equivalent to downhill skiing), Actn3 KO
muscles showed approximately 2.5 times greater force deficits compared to WT (66), indicating
that the absence of α-​actinin-​3 from fast-​twitch fibers results in increased susceptibility to
eccentric muscle damage. The mechanism underlying the altered Z-​disk stability in Actn3 KO
muscles points to an alteration in Z-​disk protein composition and interaction. Compensatory
upregulation of α-​actinin-​2 in Actn3 KO mouse muscles is accompanied by increases in
myofibrillar Z-​disk proteins γ-​filamin, myotilin, and Z-​band alternatively spliced PDZ-​motif
protein (ZASP), as well as the intermediate filament protein desmin (66) (Figure 23.4). These
proteins are expressed in both skeletal and cardiac muscle and are known to contribute to
Z-​disk stability and elasticity. In addition, ZASP, calsarcin-​2, titin, and vinculin were shown
to preferentially bind to α-​actinin-​2, while myotilin showed equivalent affinity for α-​actinin-​
2 and α-​actinin-​3 (66, 67). The interaction between ZASP, titin, and α-​actinin-​2 is critical
for the maintenance of Z-​disk integrity during muscle contractions, and titin and vinculin
are both involved in the modulation of Z-​disk width (33). Together, these findings indicate
that α-​actinin-​3 plays an integral role in stabilizing the Z-​disk during lengthening muscle
contractions –​a role that cannot be fully compensated by increased expression of α-​actinin-​2
and other Z-​disk proteins and tighter protein interactions. As Z-​disk rigidity also influences
muscle force transmission, changes in protein composition and interactions may also contribute
to the reduction in force generation in α-​actinin-​3-​deficient muscles.

Sarcomeric α-​actinins regulate calcineurin signaling

Calsarcin-​2 (also called myozenin 1 or FATZ) is exclusively expressed in fast-​twitch fibers of
skeletal muscle and inhibits calcineurin –​a calcium/​calmodulin-​dependent serine, threonine
phosphatase that mediates transcription of the slow, oxidative myogenic program in muscle
(17). The calsarcin-​2 KO mouse demonstrates enhanced calcineurin/​NFAT activity, resulting
in increases in oxidative metabolism, reduced fast-​twitch muscle mass, and increased endur-
ance running capacity –​phenotypes similar to the Actn3 KO mouse model (16). To date, the
calsarcin-​2 KO and the Actn3 KO are the only two examples where loss of function resulted in
enhanced oxidative phenotype.
This led to our hypothesis that enhanced calcineurin signaling may also occur down-
stream of α-​actinin-​3 deficiency (67). Assessment of muscles from the Actn3 KO mouse
model showed significant upregulation of calcineurin activity and expression of the down-
stream regulator of calcineurin (RCAN1–​4) with α-​actinin-​3 deficiency; these results were
also confirmed in human muscle samples. There were no changes in calsarcin-​2 expression in
WT and Actn3 KO muscles, but as discussed, calsarcin-​2 preferentially binds to α-​actinin-​2.
The calsarcins have overlapping regions that facilitate interaction with α-​actinin, γ-​filamin,
telethonin, and calcineurin that could result in competition for binding with calsarcin (34).

333
334

J.T. Seto et al.

Further investigation of these protein dynamics by co-​expressing calsarcin-​2, α-​actinin-​2, and

calcineurin in COS cells and co-​immunoprecipitation showed that calcineurin is outcompeted
for binding with calsarcin-​2 in the presence of increasing α-​actinin-​2 concentrations –​ resulting
in greater release of calcineurin from the inhibitory effects of calsarcin-​2 (67). In the context
of α-​actinin-​3 deficiency where α-​actinin-​2 is overexpressed, the dynamics of these protein
interactions thus explain the enhanced calcineurin activity and resultant shifts towards oxi-
dative metabolism in fast-​twitch fibers of Actn3 KO and ACTN3 577XX muscles. Increased
calcineurin activity at baseline also contributed to the Actn3 KO muscle adaptive response to
endurance training (67).

α-​Actinin-​3 regulates muscle metabolism through modulation of

glycogen phosphorylase activity
The sarcomeric α-​actinins also form complexes with a number of metabolic enzymes at the
Z-​disk, such as glycogen phosphorylase (GPh) (8, 56), fructose 1,6-​bisphosphatase, and aldolase
(57), suggesting that the sarcomeric α-​actinins are likely involved in metabolic sensing with
respect to glyconeogenesis and glycogenolysis (34). In the absence of α-​actinin-​3, GPh activity
is significantly reduced by 50%, resulting in almost twice the glycogen content of WT muscles
at baseline –​suggesting that glycogen breakdown is dysregulated/​slower in KO muscles (56).
The increase in glycogen content and reduction in GPh activity occurs concomitantly with
the changes in glycolytic and mitochondrial enzyme activity and are apparent in Actn3 KO
mouse muscles from at least 4 weeks of age. Together, these results suggest that reduced GPh
activity due to α-​actinin-​3 deficiency necessitates fast-​twitch muscle fibers to switch to greater
reliance on oxidative metabolism in order to generate ATP at a rate that is sufficient to sustain
requirements for rapid force generation. The mechanism by which α-​actinin-​3 influences GPh
activity is yet to be fully uncovered, but proteomic analyses suggest that α-​actinin-​3 plays a role
in the post-​translational modification of GPh (56).
Muscle metabolism may indeed be the trait that is most sensitive to changes in α-​actinin-​3
expression. In our most recent study, we explored the effect of postnatal α-​actinin-​3 replace-
ment by “gene doping” Actn3 KO muscles to restore α-​actinin-​3 expression to WT levels.
Contrary to expectations, there were no effects on muscle force generation, muscle mass, fiber
size or fiber type, and no changes to Z-​disk protein composition or calcineurin signaling (20),
but muscle metabolism and response to contraction-​induced fatigue were successfully restored
to WT levels in rescued Actn3 KO mice (Figure 23.5). These results suggest that the effect of
α-​actinin-​3 on oxidative metabolism occurs independent of these other traits, or that changes
in those traits require expression of α-​actinin-​3 during muscle development. In contrast, post-
natal overexpression of α-​actinin-​3 in WT muscles caused widespread muscle damage and was
detrimental to muscle function and performance.

α-​Actinin-​3 modifies calcium handling

Intracellular calcium concentration ([Ca2+]i) varies during muscle contraction and relaxation
through calcium release and reuptake at the sarcoplasmic reticulum (SR). In addition to modu-
lating glycogenolysis, GPh-​a (the serine-​14 phosphorylated active form of GPh) is also a nega-
tive regulator of SR calcium release –​thus providing the functional crosstalk between SR
calcium release and ATP generation (25). Consistent with reduced GPh activity, both Actn3
KO muscle fibers and myotubes in culture show an approximately fourfold higher calcium leak

334
 newgenrtpdf
335
Figure 23.5 Postnatal replacement of α-​actinin-​3 in Actn3 knockout muscles restored muscle metabolism and fatigue response to wild-​type levels. Actn3 KO mice
received intramuscular injections of recombinant adeno-associated viral vectors (rAAV) carrying human ACTN3 (KO-ACTN3) in the tibialis anterior muscle (TA)
of one hindlimb, or empty vector (KO-MCS) in the TA of the contralateral limb. Results were compared to WT muscles injected with empty vector (WT-MCS)
at the equivalent dose. A) In situ physiology assessments of KO-MCS (red), KO-ACTN3 (black) and WT-MCS (blue) TA muscles. Muscles were repeatedly and
maximally stimulated at 150 Hz (1s on, 1s off) for 2 min and force loss (% maximal force) due to fatigue was recorded. Force recovery was assessed over 1, 2, 3, 5
and 10 mins. Arrows indicate time-points shown in panel B. B) KO-ACTN3 muscles showed significantly greater force loss and slower force recovery compared
to KO-MCS with a force/time profile that is more similar to WT-MCS. C) Protein expression of muscle mitochondrial complexes I-V (CI-V) showed significant
upregulation of CII and CIV in KO-MCS compared to WT-MCS, which corresponds to succinate dehydrogenase and cytochrome c oxidase subunit I. Replacement
of a-actinin-3 in KO TA muscles (KO-ACTN3) resulted in significant decreases in all complexes (I, II, III, IV and V) relative to KO-MCS, to be equal or below
WT-MCS levels. Image adapted from (20).
336

J.T. Seto et al.

A B
10 s
[Ca2+]i as % of original
150%
10 s
0s 0s 40 s
40 s 60 s
100%
60 s
50%

0%
500 ms
500 ms

Figure 23.6 WT and Actn3 knockout [Ca2+]i transients at 0, 10, 40, and 60 seconds during contraction-​
induced muscle fatigue. WT muscle fibres show the characteristic pattern of [Ca2+]i during repeated
tetanic stimulation – initially rising, then gradually falling. In contrast, KO fibres maintained [Ca2+]i longer
into the fatigue run. Image adapted from (24).

Figure 23.7 Mechanistic insights from Actn3 knockout mouse model highlighting the structural, metabolic,
signaling, and calcium handling pathways. Alterations in all four pathways have been described as a result of
α-actinin-3 deficiency. The arrows indicate the cascades between each pathway as well as the phenotype and
mechanism thought to be responsible for alterations in each pathway. Image adapted from (33).

compared to WT (24, 56). Calcium reuptake is also approximately threefold higher in KO fibers
during twitch relaxation, such that on balance [Ca2+]i declines more quickly in KO than in WT
(24). Actn3 KO muscles also show slower rate of decay for tetanic [Ca2+]i compared to WT, in
line with increased resistance to fatigue (6, 24, 40) (Figure 23.6). Major proteins associated with
the rise of calcium transients (dihydropyridine receptor voltage sensor (DHPR); and ryanodine-​
receptor Ca2+-​release channel (RyR1)), and decay of calcium transients (which includes the
calcium buffer parvalbumin, the SR calcium pump SERCA1, and the intraluminal calcium
buffers calsequestrin and sarcalumenin) were examined in WT and Actn3 muscles. Consistent
with increased calcium reuptake, Actn3 KO muscles showed significant increases in SERCA1,

336
337

Alpha-actinin-3’s role in muscle strength

calsequestrin, and sarcalumenin. Collectively, these results indicate that α-​actinin-​3 regulates
muscle contractile properties through modulating calcium dynamics.

Mechanistic insights from the Actn3 KO mouse

Our work in the Actn3 KO mouse is unravelling the complex interplay of the structural,
metabolic, signaling, and calcium handling functions of α-​actinin-​3 and the underlying con-
sequence of α-​actinin-​3 deficiency that led to the changes in muscle function and perform-
ance (Figure 23.7). The primary absence of α-​actinin-​3 from fast-​twitch muscle fibers leads
to changes in the recruitment of structural and metabolic proteins at the Z-​disk, affecting
Z-​disk rigidity and force transmission through altered protein composition and binding. This,
in turn, enhances calcineurin signaling, as increased α-​actinin-​2 sequesters calsarcin-​2 from
its calcineurin inhibitory function. Increased calcineurin activity promotes transcription of
slow-​twitch genes, which together with reduced GPh activity, drives the metabolic shift
toward oxidative metabolism. As part of a continuous cycle, increased mitochondrial respir-
ation supplies the necessary ATP to support increased SERCA1 calcium uptake; the increased
demand for ATP signals for greater calcium leak, which in turn stimulates the activities of the
tricarboxylic acid cycle enzymes and increases ATP production. Increased calcium leak also
induces greater calcium/​calmodulin-​dependent calcineurin activity, further promoting the
maintenance of the slow myogenic program and improving muscle fatigue resistance.
Increased heat generated by ATP hydrolysis as a result of increased SR calcium leakage
and pumping in α-​actinin-​3-​deficient muscles would be advantageous in cold environments.
Together with improved fatigue resistance and increased mitochondrial activity, Actn3 KO
muscles thus exhibit all the hallmarks of muscles exposed to cold (4), suggesting that Actn3
KO mice are pre-​acclimatized to cold. In addition, the shift towards oxidative muscle metab-
olism promotes energy efficiency that is beneficial for survival under conditions of famine and
enhances endurance capacity. Our findings from the mouse model therefore lend biological
support to the positive selection of the ACTN3 577X allele during modern human migration
out of Africa into colder climates.

Beyond athletics –​the role of α-​actinin-​3 in muscle ageing,

disuse, and disease
While the absence of α-​actinin-​3 does not in itself cause disease, there is accumulating evi-
dence to show that the ACTN3 genotype is associated with increased morbidity in people
who are frail. Several studies have examined the impact of ACTN3 R577X on skeletal muscle
traits and function in the aging population (55). With some exceptions, the general consensus
from human and mouse studies is that the effect of the ACTN3 genotype on some muscle
traits and physical performance measures is lifelong, with α-​actinin-​3 deficiency generally
associated with lower muscle mass and strength and higher sarcopenia risk. On this basis,
the effect of the ACTN3 genotype on resistance training response in the elderly is being
scrutinized as a way to identify at-​r isk groups and personalize strategies for sarcopenia pre-
vention and treatment.
The absence of α-​actinin-​3 also influences how skeletal muscle adapts to disuse. Using the
Actn3 KO mouse model to assess the effects of denervation and immobilization, we found that
Actn3 KO mice exhibit significantly lower levels of muscle and fast 2B fiber atrophy, suggesting
that α-​actinin-​3 deficiency affords a level of protection against muscle wasting (21). Additionally,
KO muscles show reduced threshold for fast-​to-​slow fiber CSA shifts (with immobilization)

337
338

J.T. Seto et al.

Table 23.3 ACTN3 genotype and inherited muscle diseases

Population description (number) N Ages Study results Reference

(years)
Sex

Merosin congenital muscular 54 5–​25 No correlation was found with degree of 70

dystrophy (11) M&F muscle degeneration or clinical course
Severe childhood autosomal
recessive MD (11)
Mild limb girdle (13)
Neuromuscular disease patients 217 M&F No difference in the frequency of XX 51
and controls genotypes among individuals with
dystrophic, myopathic, neurogenic, or
normal muscle histology
Human muscle glycogen 99 8–​81 No significant relationships identified 60
phosphorylase deficiency M&F between clinical severity (daily activity
(MPL/​McArdle’s disease) measures) and ACTN3 genotype
40 34–​41 XX and RX had a higher VȮ2 peak than 37
M&F RR females. No differences in male
patients.
Inflammatory myopathies 37 24–​54 Enhanced proportion XX genotypes in 62
Dermatomyositis (DM) M&F affected group (P <0.001). ACTN3
(N=27) polymyositis (PM) genotype not related to severity of
(N=10) phenotype or enzyme levels (CPK,
LDH, AST, and ALT)
Glycogen storage disease type 126 58–​68 XX genotype is significantly associated with 10
II (GSDII) or Pompe disease M&F an early onset of the disease (P=0.024)
Duchenne muscular dystrophy 61 M DMD patients with the X-​allele show 26
(DMD) reduced muscle strength and a longer
10 min walk test time

and increased threshold for slow-​to-​fast CSA shifts (with denervation) compared to WT, as well
as changes in muscle metabolism and calcineurin signaling with respect to both stimuli that is
consistent with maintenance of a “slower” muscle phenotype.
We and others have also examined the role of ACTN3 R577X in the progression and
development of various muscle disorders (Table 23.3). Carriage of the ACTN3 X-​allele was
associated with increased exercise tolerance in female patients with McArdle’s disease (a disorder
of glycogen accumulation) (37), while α-​actinin-​3 deficiency was associated with increased
likelihood of developing inflammatory myopathies (63) and earlier onset of Pompe disease
(10). Recently, the ACTN3 genotype was also correlated with survival in patients with con-
gestive heart failure, with patients carrying the X-​allele having 1.72 times higher mortality than
patients with the ACTN3 577RR genotype (P=0.01) (3).

ACTN3 R577X is a modifier of Duchenne muscular dystrophy.

Duchenne muscular dystrophy (DMD) is an X-​linked inherited muscle disease caused by
mutations in the dystrophin gene that is characterized by a severe progressive muscle degen-
eration and weakness. Patients are diagnosed from an early age, typically become wheelchair

338
339

Alpha-actinin-3’s role in muscle strength

A B
mdx mdx

Figure 23.8 Actn3 knockout/​mdx muscles show slower disease progression compared to mdx. A) Fibre
branching is a hallmark of damage and regeneration. Actn3 KO/mdx muscles had markedly fewer
complex branched fibres compared to mdx muscles. B) Protein expression of a-actinin-2, RCAN1-4 and
AMPK is increased in Actn3 KO/mdx (DKO) muscles, consistent with activation of the slow myogenic
program. Image adapted from (26).

bound by age 12, and despite careful clinical management, succumb to the disease in their
late 20s. The current gold standard therapy for DMD is chronic corticosteroid (prednisone or
deflazacort) treatment (47) to slow progression of the disease. There is considerable interpatient
variability in DMD onset and progression (12), and much of this variation is thought to arise
from the presence of modifier genes that alter disease progression independently from the
causative mutations in the DMD gene.
Because ACTN3 R577X is so common and strongly influences muscle function in the
general population –​both healthy and frail –​–​we hypothesized that the ACTN3 genotype
may also be a disease modifier for DMD. In collaboration with the Cooperative International
Neuromuscular Research Group (CINRG) consortium, we examined the association of
ACTN3 R577X with a number of clinical muscle strength and performance measures in 61
young, ambulant patients with DMD (27). Consistent with our hypothesis, patients carrying
the ACTN3 577X allele showed reduced muscle strength and a longer 10-​minute walk test
time (Table 23.3). To study the functional implications of the ACTN3 genotype in DMD,
we crossed the Actn3 KO mouse with the mdx mouse (a model for DMD) to generate a
double knockout model (Actn3 KO/​mdx). Similar to ACTN3 577XX patients, Actn3 KO/​mdx
mice showed reduced muscle strength, but older Actn3 KO/​mdx mice showed reduced stretch-​
induced muscle damage, suggesting that α-​actinin-​3 deficiency protects dystrophic muscle from
contraction induced damage and thereby slows disease progression. Moreover, compared to
12-​month-​old mdx mice, muscle fibers from Actn3 KO/​mdx muscles exhibited lower incidence
and severity of fiber branching –​a hallmark of muscle damage and regeneration (Figure 23.8).
Comparison of Actn3 KO/​mdx muscles to mdx showed no difference in the expression of
utrophin, a dystrophin homolog that is highly expressed in developing muscle (68). However,
Actn3 KO/​mdx muscles showed increased oxidative muscle metabolism compared to mdx,
which was mediated by significantly higher activation of calcineurin and AMPK signaling.
Since slow fibers are less susceptible to damage in the mdx mouse (46), the predisposition of
“slower” fast-​twitch muscles with α-​actinin-​3 deficiency fibers likely protected Actn3 KO/​mdx
mice from progressive contraction-​induced damage. In addition, enhanced calcineurin signaling
is associated with increased regeneration (16). Our combined studies in mice and humans lend
support to our hypothesis that the ACTN3 R577X genotype is a modifier of clinical outcomes
in DMD patients.

339
340

J.T. Seto et al.

Therapeutic implications of Actn3 knockout mouse findings for

ACTN3 577XX patients
The findings from our muscular dystrophy studies have important implications for clinical
management of DMD patients moving forward. Current treatment strategies that aim to
replace dystrophin and induce muscle hypertrophy are successful in ameliorating disease (58),
but our results from α-​actinin-​3-​deficient muscle suggest that co-​delivery of pharmacological
interventions that promote slow muscle programming (29) may provide better long-​term pro-
tective benefits for patients with DMD. Similar strategies could be employed for patients with
muscle disuse, since increased shifts towards slow, oxidative metabolism mediated by enhanced
calcineurin activity were common in our studies of muscle disuse and muscular dystrophy using
the Actn3 KO mouse model. Continued research in these areas are required and considering
the ACTN3 genotype of patients is likely to inform on their prognosis and develop a more
complete clinical picture.

Conclusion
Deficiency of α-​actinin-​3 is common and its detrimental effect on elite sprint and power per-
formance has been repeatedly demonstrated. The effect of α-​actinin-​3 deficiency on more vul-
nerable populations, such as those who are elderly or suffer from muscle disuse or disease, will
benefit greatly from further research, especially for conditions that could be treated or ameliorated
using exercise training interventions, and for identifying patients who are at greater risk for
steeper decline. Phenotypic analysis of the Actn3 KO mouse model has revealed valuable insights
regarding the specialized role of α-​actinin-​3 and its cooperative functions with α-​actinin-​2 in
regulating muscle structure, cell signaling, metabolism, and calcium handling. Further research
into the mechanisms behind α-​actinin-​3 deficiency could aid identification of therapeutic targets
for personalized, precision medicine and provide better understanding towards the positive selec-
tion of the ACTN3 577X allele and its implications for health in modern society.

References
1. Alfred T, et al. ACTN3 genotype, athletic status, and life course physical capability: meta-​analysis of
the published literature and findings from nine studies. Hum Mutat 32: 1008–​1018, 2011.
2. Ben-​Zaken S, et al. ACTN3 polymorphism: comparison between elite swimmers and runners. Sports
Med Open 1: 13, 2015.
3. Bernardez-​Pereira S, et al. ACTN3 R577X polymorphism and long-​term survival in patients with
chronic heart failure. BMC Cardiovasc Disord 14: 90, 2014.
4. Bruton JD, et al. Increased fatigue resistance linked to Ca2+-​stimulated mitochondrial biogenesis in
muscle fibres of cold-​acclimated mice. J Physiol 588: 4275–​4288, 2010.
5. Chan S, et al. Properties of extensor digitorum longus muscle and skinned fibers from adult and aged
male and female Actn3 knockout mice. Muscle Nerve 43: 37–​48, 2011.
6. Chan S, et al. A gene for speed: contractile properties of isolated whole EDL muscle from an alpha-​
actinin-​3 knockout mouse. Am J Physiol 295: C897–​C904, 2008.
7. Chiu LL, et al. ACTN3 genotype and swimming performance in Taiwan. Int J Sports Med
32: 476–​480, 2011.
8. Chowrashi P, et al. Amorphin is phosphorylase; phosphorylase is an alpha-​actinin-​binding protein.
Cell Motil Cytoskeleton 53: 125–​135, 2002.
9. Cieszczyk P, et al. Association of the ACTN3 R577X polymorphism in Polish power-​orientated
athletes. J Hum Kinet 28: 55–​61, 2011.

340
341

Alpha-actinin-3’s role in muscle strength

10. De Filippi P, et al. Genotype-​phenotype correlation in Pompe disease, a step forward. Orphanet J
Rare Dis 9: 102, 2014.
11. Delmonico MJ, et al. Alpha-​actinin-​3 (ACTN3) R577X polymorphism influences knee extensor
peak power response to strength training in older men and women. J Gerontol A Biol Sci Med Sci
62: 206–​212, 2007.
12. Desguerre I, et al. Clinical heterogeneity of Duchenne muscular dystrophy (DMD): definition of
sub-​phenotypes and predictive criteria by long-​term follow-​up. PLoS One 4: e4347, 2009.
13. Druzhevskaya AM, et al. Association of the ACTN3 R577X polymorphism with power athlete
status in Russians. Eur J Appl Physiol 103: 631–​634, 2008.
14. Erskine RM, et al. The individual and combined influence of ACE and ACTN3 genotypes on
muscle phenotypes before and after strength training. Scand J Med Sci Sports 24: 642–​648, 2014.
15. Eynon N, et al. ACTN3 R577X polymorphism and Israeli top-​level athletes. Int J Sports Med
30: 695–​698, 2009.
16. Frey N, et al. Calsarcin-​2 deficiency increases exercise capacity in mice through calcineurin/​NFAT
activation. J Clin Invest 118: 3598–​3608, 2008.
17. Frey N, et al. Calsarcins, a novel family of sarcomeric calcineurin-​binding proteins. Proc Natl Acad Sci
U S A 97: 14632–​14637, 2000.
18. Friedlander SM, et al. ACTN3 Allele frequency in humans covaries with global latitudinal gradient.
PLoS One 8: e52282, 2013.
19. Garton FC and North KN. The effect of heterozygosity for the ACTN3 null allele on human
muscle performance. Med Sci Sports Exerc 2015.
20. Garton FC, et al. The effect of ACTN3 gene doping on skeletal muscle performance. Am J Hum
Genet 102: 845–​857, 2018.
21. Garton FC, et al. alpha-​Actinin-​3 deficiency alters muscle adaptation in response to denervation and
immobilization. Hum Mol Genet 23: 1879–​1893, 2014.
22. Gentil P, et al. ACTN3 R577X polymorphism and neuromuscular response to resistance training. J
Sports Sci Med 10: 393–​399, 2011.
23. Gineviciene V, et al. Association analysis of ACE, ACTN3 and PPARGC1A gene polymorphisms
in two cohorts of European strength and power athletes. Biol Sport 33: 199–​206, 2016.
24. Head SI, et al. Altered Ca2+ kinetics associated with alpha-​actinin-​3 deficiency may explain positive
selection for ACTN3 null allele in human evolution. PLoS Genet 11: e1004862, 2015.
25. Hirata Y, et al. Mastoparan binds to glycogen phosphorylase to regulate sarcoplasmic reticular Ca2+
release in skeletal muscle. Biochem J 371: 81–​88, 2003.
26. Hogarth MW, et al. Analysis of the ACTN3 heterozygous genotype suggests that alpha-​actinin-​3
controls sarcomeric composition and muscle function in a dose-​dependent fashion. Hum Mol Genet
25: 866–​877, 2016.
27. Hogarth MW, et al. Evidence for ACTN3 as a genetic modifier of Duchenne muscular dystrophy.
Nat Comm 8: 14143, 2017.
28. Hong SSJ and Jin HJ. Assessment of association of ACTN3 genetic polymorphism with Korean
elite athletic performance. Genes Genomics 35: 617, 2013.
29. Jahnke VE, et al. Metabolic remodeling agents show beneficial effects in the dystrophin-​deficient
mdx mouse model. Skelet Muscle 2: 16, 2012.
30. Kikuchi N, et al. Higher frequency of the ACTN3 R allele + ACE DD genotype in Japanese elite
wrestlers. J Strength Cond Res 26: 3275–​3280, 2012.
31. Kikuchi N, et al. ACTN3 R577X genotype and athletic performance in a large cohort of Japanese
athletes. Eur J Sport Sci 16: 694–​701, 2016.
32. Kim H, et al. The ACTN3 R577X variant in sprint and strength performance. J Exerc Nutrition
Biochem 18: 347–​353, 2014.
33. Lee FX, et al. How does alpha-​actinin-​3 deficiency alter muscle function? Mechanistic insights into
ACTN3, the “gene for speed”. Biochim Biophys Acta 1863: 686–​693, 2016.
34. Lek M and North KN. Are biological sensors modulated by their structural scaffolds? The role of
the structural muscle proteins alpha-​actinin-​2 and alpha-​actinin-​3 as modulators of biological sensors.
FEBS Lett 584: 2974–​2980, 2010.
35. Lek M, et al. The evolution of skeletal muscle performance: gene duplication and divergence of
human sarcomeric α-​actinins. Bioessays 32: 17–​25, 2010.

341
342

J.T. Seto et al.

36. Lima RM, et al. ACE and ACTN3 genotypes in older women: muscular phenotypes. Int J Sports Med
32: 66–​72, 2011.
37. Lucia A, et al. The 577X allele of the ACTN3 gene is associated with improved exercise capacity in
women with McArdle’s disease. Neuromuscul Disord 17: 603–​610, 2007.
38. Ma F, et al. The association of sport performance with ACE and ACTN3 genetic polymorphisms: a
systematic review and meta-​analysis. PLoS One 8: e54685, 2013.
39. MacArthur DG and North KN. A gene for speed? The evolution and function of alpha-​actinin-​3.
Bioessays 26: 786–​795, 2004.
40. MacArthur DG, et al. An Actn3 knockout mouse provides mechanistic insights into the asso-
ciation between alpha-​ 3 deficiency and human athletic performance. Hum Mol Genet
actinin-​
17: 1076–​1086, 2008.
41. MacArthur DG, et al. Loss of ACTN3 gene function alters mouse muscle metabolism and shows
evidence of positive selection in humans. Nat Genet 39: 1261–​1265, 2007.
42. Magi A, et al. The association analysis between ACE and ACTN3 genes polymorphisms and endur-
ance capacity in young cross-​country skiers: longitudinal study. J Sports Sci Med 15: 287–​294, 2016.
43. Massidda M, et al. Association between the ACTN3 R577X polymorphism and artistic gymnastic
performance in Italy. Genet Test Mol Biomarkers 13: 377–​380, 2009.
44. Mikami E, et al. ACTN3 R577X genotype is associated with sprinting in elite Japanese athletes. Int
J Sports Med 35: 172–​177, 2014.
45. Mills M, et al. Differential expression of the actin-​binding proteins, alpha-​actinin-​2 and -​3, in different
species: implications for the evolution of functional redundancy. Hum Mol Genet 10: 1335–​1346, 2001.
46. Moens P, et al. Increased susceptibility of EDL muscles from mdx mice to damage induced by
contractions with stretch. J Muscle Res Cell Motil 14: 446–​451, 1993.
47. Moxley RT, 3rd, et al. Practice parameter: corticosteroid treatment of Duchenne dystrophy: report
of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice
Committee of the Child Neurology Society. Neurology 64: 13–​20, 2005.
48. Niemi AK, and Majamaa K. Mitochondrial DNA and ACTN3 genotypes in Finnish elite endur-
ance and sprint athletes. Eur J Hum Genet 13: 965–​969, 2005.
49. Norman B, et al. Strength, power, fiber types, and mRNA expression in trained men and women
with different ACTN3 R577X genotypes. J Appl Physiol (1985) 106: 959–​965, 2009.
50. North KN and Beggs AH. Deficiency of a skeletal muscle isoform of alpha-​actinin (alpha-​actinin-​
3) in merosin-​positive congenital muscular dystrophy. Neuromuscul Disord 6: 229–​235, 1996.
51. North KN, et al. A common nonsense mutation results in alpha-​actinin-​3 deficiency in the general
population. Nat Genet 21: 353–​354, 1999.
52. Papadimitriou ID, et al. The ACTN3 gene in elite Greek track and field athletes. Int J Sports Med
29: 352–​355, 2008.
53. Pasqua LA, et al. Influence of ACTN3 R577X polymorphism on ventilatory thresholds related to
endurance performance. J Sports Sci 34: 163–​170, 2016.
54. Pereira A, et al. ACE I/​D and ACTN3 R/​X polymorphisms as potential factors in modulating
exercise-​related phenotypes in older women in response to a muscle power training stimuli. Age
35: 1949–​1959, 2013.
55. Pickering C and Kiely J. ACTN3, morbidity, and healthy aging. Front Genet 9: 15, 2018.
56. Quinlan KGR, et al. Alpha-​actinin-​3 deficiency results in reduced glycogen phosphorylase activity
and altered calcium handling in skeletal muscle. Hum Mol Genet 19: 1335–​1346, 2010.
57. Rakus D, et al. Interaction between muscle aldolase and muscle fructose 1,6-​bisphosphatase results
in the substrate channeling. Biochemistry 43: 14948–​14957, 2004.
58. Rodino-​Klapac LR, et al. Micro-​dystrophin and follistatin co-​delivery restores muscle function in
aged DMD model. Hum Mol Genet 22: 4929–​4937, 2013.
59 Roth SM, et al. The ACTN3 R577X nonsense allele is under-​represented in elite-​level strength
athletes. Eur J Hum Genet 16: 391–​394, 2008.
60. Rubio JC, et al. Genotype modulators of clinical severity in McArdle disease. Neurosci Lett
422: 217–​222, 2007.
61. Ruiz JR, et al. Can we identify a power-​ oriented polygenic profile? J Appl Physiol (1985)
108: 561–​566, 2010.
62. Sandoval-​García F, et al. The ACTN3 R577X polymorphism is associated with inflammatory
myopathies in a Mexican population. Scand J Rheumatol 41: 396–​400, 2012.

342
343

Alpha-actinin-3’s role in muscle strength

63. Sandoval-​Garcia F, et al. The ACTN3 R577X polymorphism is associated with inflammatory
myopathies in a Mexican population. Scand J Rheumatol 41: 396–​400, 2012.
64. Scott RA, et al. ACTN3 and ACE genotypes in elite Jamaican and US sprinters. Med Sci Sports Exerc
42: 107–​112, 2010.
65. Sessa F, et al. Gene polymorphisms and sport attitude in Italian athletes. Genet Test Mol Biomarkers
15: 285–​290, 2011.
66. Seto JT, et al. Deficiency of α-​actinin-​3 is associated with increased susceptibility to contraction-​
induced damage and skeletal muscle remodeling. Hum Mol Genet 20: 2914–​2927, 2011.
67. Seto JT, et al.ACTN3 genotype influences muscle performance through the regulation of calcineurin
signaling. J Clin Invest 123: 4255–​4263, 2013.
68. Seto JT, et al. Gene replacement therapies for duchenne muscular dystrophy using adeno-​associated
viral vectors. Curr Gene Ther 12: 139–​151, 2012.
69. Silva MS, et al. Elimination of influences of the ACTN3 R577X variant on oxygen uptake by
endurance training in healthy individuals. Int J Sports Physiol Perform 10: 636–​641, 2015.
70. Vainzof M, et al. Deficiency of alpha-​actinin-​3 (ACTN3) occurs in different forms of muscular dys-
trophy. Neuropediatrics 28: 223–​228, 1997.
71. Vincent B, et al. The ACTN3 (R577X) genotype is associated with fiber type distribution. Physiol
Genomics 32: 58–​63, 2007.
72. Walsh S, et al. ACTN3 genotype is associated with muscle phenotypes in women across the adult age
span. J Appl Physiol 105: 1486–​1491, 2008.
73. Weyerstrass J, et al. Nine genetic polymorphisms associated with power athlete status –​a meta-​
analysis. J Sci Med Sport 21: 213–​220, 2018.
74. Yang N, et al. ACTN3 genotype is associated with human elite athletic performance. Am J Hum
Genet 73: 627–​631, 2003.
75. Yang N, et al. The ACTN3 R577X polymorphism in East and West African athletes. Med Sci Sports
Exerc 39: 1985–​1988, 2007.
76. Yang R, et al. ACTN3 R577X gene variant is associated with muscle-​related phenotypes in elite
Chinese sprint/​power athletes. J Strength Cond Res 31: 1107–​1115, 2017.
77. Zempo H, et al. ACTN3 polymorphism affects thigh muscle area. Int J Sports Med 31: 138–​142, 2010.

343
344
345

SECTION 5

Systems genetics of sports

performance

As we transition into the later sections of the book, we build on the foundational information
of the previous chapters and now add substantial complexity. Whereas the previous chapters
have looked at the systems genetics underlying particular phenotypes (e.g., physical activity,
cardiorespiratory fitness, and muscle size), Section 5 now moves to similar questions related to
sport and exercise performance. Performance is remarkably challenging to measure, and by def-
inition elite-​level performance is rare. The complexity of performance is in part due to the fact
that it is not a trait per se, but an outcome of many traits working in combination. Moreover,
performance is generally tied to competition, which additionally imparts elements of environ-
mental conditions and chance that laboratory-​based trait measurements are meant to minimize.
The authors contributing to Section 5 are experts in a variety of related fields and provide
unique insights into the role of systems genetics on different aspects of sport and exercise per-
formance. In Chapter 24, Dr. Nir Eynon and his team from Victoria University in Australia
provide a review of the genetic aspects of sport performance and lay the foundation for the
section. Chapter 25, contributed by Drs. Colin Moran, Alun Williams, and Guan Wang from the
UK, delves into the use of elite athletes as models for genetic analysis. These unique performers
who rise to the very top of their sport can provide unique insights into what makes them so
good, yet their small numbers challenge our statistical models. In Chapter 26, Drs. José Maia
and Peter Katzmarzyk look at how our familial relations, specifically in the form of twin and
family-​based genetic studies, can further inform our understanding of genetic contributions to
performance and related traits. Chapters 27 and 28 examine a frequent side effect of engage-
ment in sport: injury.While certain sports lend themselves to greater rates of injury, genetic sus-
ceptibility has emerged as a contributing factor for certain individuals. Chapter 27, by Drs. Ryan
Tierney and Jane McDevitt from Temple University in the US, considers the systems genetics
of brain injury, in particular concussion, which has gained public attention across the world in
recent years. Chapter 28 focuses below the neck, with Dr. Malcolm Collins and his team from
South Africa specifically examining the role of genetic susceptibility to soft tissue injuries, such
as tendon and ligament tears. As genetic testing costs shrink and our understanding of genetic
susceptibility improves, the push for genetic testing in these areas will increase with the hopes
of reducing injury rates in athletes (as we explore more broadly in Section 6).

345
346

Systems genetics of sports performance

We end Section 5 with a chapter focused on the broader theme of sex and performance.
While we know that performance in men’s and women’s sport differs in many areas, justifying
sex separation in the vast majority of sports leagues, what is less certain is the basis for these
underlying differences. In Chapter 29, Dr. Mindy Millard-​Stafford and Matthew Wittbrodt
from the US review the evidence for sex differences in sport performance and outline the latest
hypotheses for why these differences exist.
In total, Section 5 not only provides a bookend for the prior sections of the book, by pulling
in the trait-​specific analyses of preceding chapters and viewing them through the complex lens
of performance, but also opens a door to Section 6, which will examine the ethical questions
inherent in advancing sport performance with our growing understanding of systems genetics.
We have many gaps in our knowledge, as outlined in these insightful chapters, but as those gaps
are narrowed, athletes, coaches, and others will use this information to their advantage to try to
pursue performance improvements.

346
347

24
SUMMARY FINDINGS ON
GENETICS AND SPORT
PERFORMANCE
Macsue Jacques, Shanie Landen, Sarah Voisin, and Nir Eynon

Introduction
In the past decades, the genetic basis for athletic performance and response to exercise training
(trainability) has become a topic of great interest. Twin and familial studies have shown that
part of the interindividual variability reported in sports science studies can be explained by the
genetic makeup of individuals (16). Although humans’ DNA sequences are similar at 99.9%,
genetic mutations are found across the genome at a rate of approximately 1 change per 1000
base pairs. Some of these mutations consist of the replacement of a single nucleotide by another
and are referred to as single nucleotide polymorphisms (SNPs). Those SNPs can be rare (occur
at a frequency <1%) or common (occur at a frequency >1%) within a given population. Each
human carries 3–​4 million common variants and 200,000–​500,000 rare variants, and some of
these variants are involved in athletic performance and trainability. To discover those variants,
initially candidate gene approaches and then later genome-​wide association studies (GWAS)
were conducted (Figure 24.1). To date, more than 200 genetic variants have been associated
with athletic performance or trainability in at least one study (1, 10, 14, 22, 52, 53, 65). These
findings suggest that athletic performance and trainability are not determined by a single gene
but by a plethora of genes and are therefore considered complex traits. In this chapter, we will
summarize the findings on genetics and sport performance and we will discuss some of the
current research efforts that are undergoing to further characterize “exercise genes.”

Exercise performance and heritability

It is common for more than one family member to be an athlete.There is a long list of successful
sibling athletes (i.e., the Williams sisters in tennis), and parent and child athletes (i.e., John Kelly
and John Kelly Jr. in rowing). Examples like these highlight not only the significance of family
support in producing a successful athlete, but also the potential role of genetic predisposition
to exceptional athletic performance. In science, the genetic predisposition for a given trait is
measured by a quantity called “heritability.” Heritability is defined as the proportion of variance
in a trait that is explained by heritable factors, for a given population at a given moment in time
(7). Heritability in exercise performance was first introduced by twin and family studies in the
1970s–​1990s. A pioneering study on 25 pairs of twins found that about 93% of the variation

347
 newgenrtpdf
348
Figure 24.1 Timeline of genetic studies in sports science.
349

Genetics and sport performance

in VO2max is explained by genetic factors (26). However, this study assumed similar environ-
mental interaction between twin pairs and the absence of gene–​environment correlation, which
is not necessarily true and may overestimate heritability (19). High heritability estimates were
also found for skeletal muscle fiber composition (h2=96.5%) (26), maximal power (h2=97%),
and maximal isometric force (h2=83%) (23). Subsequent to these pioneering studies, other
twin and family studies also reported high heritability estimates for VO2max (77% (17), 69–​87%
(31), and 71% (38)). A meta-​analysis of eight twin studies found a weighted VO2max herit-
ability estimate of 72% (50). In addition to those, several other studies in different areas have
attempted to estimate heritability in different cohorts and phenotypic traits (28, 29, 34, 35, 57,
58). Although heritability studies did not require genetic testing to create estimations, they
played an important role in the full comprehension of the heritability concept, often misun-
derstood (19). For example, when said that VO2max h2=93%, is almost always misinterpreted to
mean that 93% of this phenotype is genetically determined and only 7% would be affected to
environment stimuli. Heritability only has a meaning at the population level and is irrelevant
at an individual level. In addition, heritability varies depending on the studied population (e.g.,
young, old, healthy, and diseased) and time as the environment can change with time. As pre-
viously written:

[Heritability] describes the extent to which heredity affects the variation of a given
attribute in a given population exposed to common environmental influences at a given
time. A high heritable attribute does not mean that a phenotype is predetermined and
the environment has no effect. It only indicates that the observed individual differences
in the given attribute are due to genetic differences and are highly predictable. (19)

Following heritability estimates for aerobic and muscular fitness at a single timepoint, the herit-
ability of responses to exercise training (trainability) was also calculated in the large-​scale HEalth,
RIsk factors, exercise Training And GEnetics (HERITAGE) Family Study in the 1990s. After
adjusting for confounders such as age, sex, baseline VO2max, body mass, and composition, the
HERITAGE Family Study reported that approximately 47–​50% of the variance in VO2max
improvements following exercise training was due to genetic factors (3). The variance in indi-
vidual training response between families was 2.5 times higher than within families. As a result
of family and twin studies reporting consistently high heritability estimates, a hunt for the genes
responsible for both athletic performance and trainability began.

Linkage analyses and the candidate gene approach

The pivotal HERITAGE Family Study (4) and two other twin and family studies (36, 56)
performed linkage analysis to detect the chromosomal location of genes associated with athletic
status, physical activity level, and trainability. They identified markers on several chromosomes,
which were associated with their traits of interest, but few of them were in common. The
completion of the Human Genome Project in 2001 (59) allowed for the development of gen-
etic studies based on genes with a specific function. This method, called the “candidate gene
approach,” was a hypothesis-​driven approach whereby the frequency of mutations in genes
with a potentially relevant function for exercise was compared between athletes and controls, or
associated with aerobic and muscular fitness measures (20). A recent review reported that about
200 SNPs have been associated with performance traits in at least one study, and around 25
SNPs with athletic status (16, 61, 65).These SNPs were found to have an extremely small effect
size, which means that their individual contribution to the variability in performance traits and

349
350

M. Jacques et al.

athletic status was tiny, and therefore hard to detect. In addition, only two SNPs showed con-
sistent replication across different studies and populations.
One of them is located in the ACTN3 gene. The alpha-​actinin-​3 protein, encoded by
the ACTN3 gene, is almost exclusively expressed in the sarcomeres of fast glycolytic type II
fibers that generate powerful, explosive contractions (67). The unique expression pattern and
sequence conservation of alpha-​actinin-​3 over 300 million years in humans, suggest that it has
an important function in fast, glycolytic (type IIX) muscle fibers (33). A common SNP in this
gene causes the substitution of an R (arginine) for a stop codon (X) at amino acid position
577, resulting in complete deficiency of the alpha-​actinin-​3 protein (ACTN3 XX variant) (67).
The ACTN3 R577X variant has been extensively studied in sports performance. The earliest
findings reported that fewer Australian sprint elite athletes than healthy controls carried the
XX genotype (66). Many replication studies in different cohorts of varying ethnicities found a
systematically lower frequency of the XX genotype in strength and power athletes compared
with controls and endurance athletes (13, 15, 21, 40, 42, 49, 67). Overall, nearly all studies on
ACTN3 and sports performance reported the same findings: the RR genotype is associated
with strength and muscle power, while the XX genotype tends to be associated with endurance
performance, but this association is less pronounced (12, 15, 18, 24, 25, 30, 41, 43, 54, 55).
Another gene that has showed consistent replication in exercise science is the angiotensin-​
converting enzyme (ACE) gene. ACE converts the angiotensin I hormone to another form
named angiotensin II (46). Angiotensin II helps regulate blood pressure and may also influence
skeletal muscle function, although this role is not completely understood (9). A polymorphism
in the ACE gene that consists of an insertion (I) or a deletion (D) of a segment of DNA (with
a length of 287 bp of nucleotides), alters the levels of the ACE enzyme in blood. Individuals can
have either zero (DD genotype), one (ID genotype), or two (II genotype) insertions at the ACE
I/​D polymorphism. Of the three genotypes, DD, is associated with the highest levels of ACE in
blood (64). DD carriers have a higher proportion of fast-​twitch muscle fibers and greater speed
(44, 52). However, findings on ACE are more conflicting, and less convincing than those on
ACTN3.The initial findings that II carriers are better at endurance while DD carriers are better
at power/​strength (37, 46) was not replicated in some studies (8, 51), perhaps due to different
cohorts or small sample sizes.
In addition to ACTN3 and ACE, many other genes with diverse functions have been
associated with exercise-​related traits without showing consistent replication. Some are involved
in skeletal muscle function, while others play roles in the production of cellular energy or com-
munication between nerve cells. From the candidate gene approach, the large-​scale screening of
millions of SNPs and their association with exercise-​related traits in genome-​wide association
studies (GWAS) has greatly advanced the field.

Genome-​wide association studies

With advances in microarray-​ based high-​ throughput technologies, screening hundreds of
thousands, and even millions of SNPs simultaneously has been made possible. The GWAS
approach is an unbiased, hypothesis-​free design that has led to substantial progress in the field of
disease genetics (32) and more recently in the field of sport performance. In fact, this method
has identified a plethora of genes, whose variants can be related to physical performance,
achievement, and sports results. Unfortunately, this method also revealed that genes influencing
exercise performance and their relationships are more complex than previously thought. The
HERITAGE Family Study was the first GWAS to identify several SNPs related to VO2max
trainability in an unbiased, hypothesis-​free manner (5). However, replications of those SNPs

350
351

Genetics and sport performance

were unsuccessful (19). This lack of replication may be explained by a lack of statistical power
in replication studies, as SNPs that influence exercise-​related traits have typically very small
effect sizes. In other words, those SNPs have such a small effect on the phenotype of interest
that it is challenging to detect their effect, thus leading to negative results. A recent GWAS was
conducted on a total of 1520 elite athletes and 2760 controls spanning eight different cohorts
of athletes from Australia, Ethiopia, Japan, Kenya, Poland, Russia, and Spain (47). Forty-​five
promising SNPs were found during the discovery phase but failed to pass the replication phase.
The study was likely underpowered to identify alleles with small effect sizes, perhaps due to
the fact that the technology used was an earlier-​generation microarray covering only 195,000
gene markers (48). Another GWAS of 492 strength/​power and 227 endurance athletes found
that the rare T allele of rs939787 in the dystrophin gene (DMD) was overrepresented among
strength/​power athletes compared with endurance athletes (25% vs. 8.8%) (39). Even more
recently, 16 novel genetic loci were associated with handgrip strength (a simple measure of
functional strength) in a large-​scale GWAS of 195,180 individuals; however, those interesting
findings await replication (63). Although candidate gene and GWAS studies have increased our
understanding of the genetic contribution to athletic performance, especially with regard to
effect sizes, those initial findings need to be replicated. In addition, the molecular mechanisms
by which those SNPs act need to be uncovered to have a clear picture of how the genetic
makeup shapes the athletic potential and exercise adaptation.

Conclusion and recommendations for future research

To date, neither candidate genes nor genome-​wide associations have convincingly validated
any of the target genes discovered by the pioneering HERITAGE Family Study (5), and the
largest GWAS combining more than 1500 athletes found no evidence of a common DNA
variant profile specific to world-​class endurance athletes (47). The limited progress in genetics
and sports performance achieved today is due to small and mostly heterogeneous cohorts,
resulting in doubtful and conflicting findings (19, 62). There is a necessity for larger collective
work, with well-​defined phenotypes of interest, tightly controlled interventions, and subsequent
replication studies to produce robust results. Sports scientists have answered this call by creating
the “Athlome Project Consortium” (www.athlomeconsortium.org). This multicentered, inter-
national collaborative action intends to create a large databank, with enough expertise and state-​
of-​the-​art “omics” technologies from around the world.This project aims to expand knowledge
on genetic variants involved in sports performance, trainability, and injury predisposition (more
about this project can be found in a recent review, see 62).
Bouchard also argued that there is a need for a paradigm shift in the field (2). Research should
be conducted using an unbiased approach, using the full power of genomics, epigenomics,
and transcriptomics together with large-​scale cohorts and validation studies. Fortunately, the
Athlome Consortium meets those suggested criteria, and the project published a GWAS on
eight different cohorts of athletes (48) and announced the sequencing of 1000 of the world’s
greatest athletes as part of the 1000 Athlome Project (45). The field has progressed from herit-
ability models to comprehensive sequencing and genetic screening, but to fully comprehend the
mechanism of action of exercise-​related variants, it is essential to consider interactions between
those genetic variants and the environment. Perhaps the epigenetic response to exercise training
will help scientists understand changes in gene function that cannot be explained by changes in
DNA sequences (60). Epigenetics is an attractive mechanism to explain the paradoxical findings
of identical twins who differ in heritable traits (6, 27). The field of sports and epigenetics is still
in its infancy, and the mechanisms that modulate gene expression are not well understood (11).

351
352

M. Jacques et al.

Great challenges lay ahead in the field of sports and genomics/​epigenomics, but it is an exciting
future to dissect the role of epigenetic and genetic modifications on sports performance and
trainability.

References
1. Ben-​Zaken S, Meckel Y, Nemet D, Eliakim A. IGF-​I receptor 275124A>C (rs1464430) poly-
morphism and athletic performance. J Sci Med Sport 18: 323–​327, 2015.
2. Bouchard C. Exercise genomics—​a paradigm shift is needed: a commentary. Br J Sports Med
49: 1492–​1496, 2015.
3. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, Perusse L, Leon AS, Rao DC.
Familial aggregation of VO2 max response to exercise training: results from the HERITAGE Family
Study. J Appl Physiol 87: 1003–​1008, 1999.
4. Bouchard C, Rankinen T, Chagnon YC, Rice T, Pérusse L, Gagnon J, Borecki I, An P, Leon
AS, Skinner JS, Wilmore JH, Province M, Rao DC. Genomic scan for maximal oxygen uptake
and its response to training in the HERITAGE Family Study. J Appl Physiol 88: 551–​559, 2000.
5. Bouchard C, Sarzynski M a, Rice TK, Kraus WE, Church TS, Sung YJ, Rao DC, Rankinen
T. Genomic predictors of the maximal O₂ uptake response to standardized exercise training programs.
J Appl Physiol 110: 1160–​1170, 2011.
6. Bouchard T, Lykken D, McGue M, Segal N, Tellegen A. Sources of human psychological
differences: the Minnesota Study of Twins Reared Apart. Science 250: 223–​228, 1990.
7. Cesarini D, Visscher PM. Genetics and educational attainment. NPJ Sci Learn 2: 4, 2017.
8. Day SH, Gohlke P, Dhamrait SS, Williams AG. No correlation between circulating ACE activity
and VO2max or mechanical efficiency in women. Eur J Appl Physiol 99: 11–​18, 2007.
9. Djarova T, Bardarev D, Boyanov D, Kaneva R, Atanasov P. Performance enhancing gen-
etic variants, oxygen uptake, heart rate, blood pressure and body mass index of elite high altitude
mountaineers. Acta Physiol Hung 100: 289–​301, 2013.
10. Egorova ES, Borisova A V, Mustafina LJ, Arkhipova AA, Gabbasov RT, Druzhevskaya
AM, Astratenkova I V, Ahmetov II. The polygenic profile of Russian football players. J Sports Sci
32: 1286–​1293, 2014.
11. Ehlert T, Simon P, Moser DA. Epigenetics in sports. Sport Med 43: 93–​110, 2013.
12. Eynon N, Banting LK, Ruiz JR, Cieszczyk P, Dyatlov DA, Maciejewska-​Karlowska A,
Sawczuk M, Pushkarev VP, Kulikov LM, Pushkarev ED, Femia P, Stepto NK, Bishop DJ,
Lucia A. ACTN3 R577X polymorphism and team-​sport performance: a study involving three
European cohorts. J Sci Med Sport 17: 102–​106, 2014.
13. Eynon N, Duarte JA, Oliveira J, Sagiv M, Yamin C, Meckel Y, Sagiv M, Goldhammer E.
ACTN3 R577X polymorphism and Israeli top-​level athletes. Int J Sports Med 30: 695–​698, 2009.
14. Eynon N, Hanson ED, Lucia A, Houweling PJ, Garton F, North KN, Bishop DJ. Genes for
elite power and sprint performance: ACTN3 leads the way. Sport Med 43: 803–​817, 2013.
15. Eynon N, Ruiz JR, Femia P, Pushkarev VP, Cieszczyk P, Maciejewska-​Karlowska A,
Sawczuk M, Dyatlov DA, Lekontsev EV, Kulikov LM, Birk R, Bishop DJ, Lucia A. The
ACTN3 R577X polymorphism across three groups of elite male European athletes. PLoS One
7, 2012.
16. Eynon N, Ruiz JR, Oliveira J, Duarte JA, Birk R, Lucia A. Genes and elite athletes: a roadmap
for future research. J Physiol 589: 3063–​3070, 2011.
17. Fagard R, Bielen E, Amery A. Heritability of aerobic power and anaerobic energy generation
during exercise. J Appl Physiol 70: 357–​362, 1991.
18. Garatachea N, Verde Z, Santos-​Lozano A, Yvert T, Rodriguez-​Romo G, Sarasa FJ,
Hernández-​Sánchez S, Santiago C, Lucia A. ACTN3 R577X polymorphism and explosive leg-​
muscle power in elite basketball players. Int J Sports Physiol Perform 9: 226–​232, 2014.
19. Georgiades E, Klissouras V, Baulch J, Wang G, Pitsiladis Y.Why nature prevails over nurture in
the making of the elite athlete. BMC Genomics 18: 835, 2017.
20. Gibson WT. Core concepts in human genetics: understanding the complex phenotype of sport per-
formance and susceptibility to sport injury. Med Sport Sci 61: 1–​14, 2016.
21. Ginevičienė V, Pranculis A. Genetic variation of the human ACE and ACTN3 genes and their asso-
ciation with functional muscle properties in Lithuanian elite athletes. Kaunas Lith 47: 284–​90, 2010.

352
353

Genetics and sport performance

22. Grealy R, Smith CLE, Chen T, Hiller D, Haseler LJ, Griffiths LR. The genetics of endur-
ance: frequency of the ACTN3 R577X variant in Ironman World Championship athletes. J Sci Med
Sport 16: 365–​371, 2013.
23. Jones B, Klissouras V. Genetic variation in the force-​velocity relation of human muscle. In: Sport
and Human Genetics, edited by Malina RM, Bouchard C. Champaign, IL: Human Kinetics,
p. 155–​163, 1986.
24. Kikuchi N, Yoshida S, Min SK, Lee K, Sakamaki-​Sunaga M, Okamoto T, Nakazato K. The
ACTN3 R577X genotype is associated with muscle function in a Japanese population. Appl Physiol
Nutr Metab 322: 1–​7, 2014.
25. Kim H, Song K, Kim C. The ACTN3 R577X variant in sprint and strength performance. J Exerc
Nutrition Biochem 18: 347–​353, 2014.
26. Klissouras V. Heritability of adaptive variation. J Appl Physiol 31: 338–​344, 1971.
27. Klissouras V, Casini B, Di Salvo V, Faina M, Marini C, Pigozzi F, Pittaluga M, Spataro
A, Taddei F, Parisi P. Genes and Olympic performance: a co-​ twin study. Int J Sports Med
22: 250–​255, 2001.
28. Komi PV, Klissouras V, Karvinen E. Genetic variation in neuromuscular performance. Int Z
Angew Physiol 31: 289–​304, 1973.
29. Komi PV, Viitasalo JH, Havu M, Thorstensson A, Sjodin B, Karlsson J. Skeletal muscle fibers
and muscle enzyme activities in monozygous and dizygous twins of both sexes. Acta Physiol Scand
100: 385–​392, 1977.
30. MacArthur DG, North KN. ACTN3: a genetic influence on muscle function and athletic per-
formance. Exerc Sport Sci Rev 35: 30–​34, 2007.
31. Maes HH, Beunen GP, Vlietinck RF, Neale MC, Thomis M, Vanden Eynde B, Lysens R,
Simons J, Derom C, Derom R. Inheritance of physical fitness in 10-​yr-​old twins and their parents.
Med Sci Sport Exerc 28: 1479–​1491, 1996.
32. McCarthy MI, MacArthur DG. Human disease genomics: from variants to biology. Genome Biol
18: 18–​20, 2017.
33. Mills M. Differential expression of the actin-​ binding proteins, alpha-​ actinin-​
2 and -​ 3, in
different species: implications for the evolution of functional redundancy. Hum Mol Genet
10: 1335–​1346, 2001.
34. Missitzi J, Geladas N, Klissouras V. Heritability in neuromuscular coordination: implications for
motor control strategies. Med Sci Sports Exerc 36: 233–​240, 2004.
35. Missitzi J, Gentner R, Misitzi A, Geladas N, Politis P, Klissouras V, Classen J. Heritability of
motor control and motor learning. Physiol Rep 1: e00188, 2013.
36. De Moor MHM, Spector TD, Cherkas LF, Falchi M, Hottenga JJ, Boomsma DI, De Geus
EJC. Genome-​wide linkage scan for athlete status in 700 British female DZ twin pairs. Twin Res Hum
Genet 10: 812–​820, 2007.
37. Moran CN, Vassilopoulos C, Tsiokanos A, Jamurtas AZ, Bailey MES, Montgomery HE,
Wilson RH, Pitsiladis YP.The associations of ACE polymorphisms with physical, physiological and
skill parameters in adolescents. Eur J Hum Genet 14: 332–​339, 2006.
38. Mustelin L, Latvala A, Pietiläinen KH, Piirilä P, Sovijärvi AR, Kujala UM, Rissanen A,
Kaprio J. Associations between sports participation, cardiorespiratory fitness, and adiposity in young
adult twins. J Appl Physiol 110: 681–​686, 2011.
39. Naumov VA, Ahmetov II, Larin AK, Generozov EV, Kulemin NA, Ospanova EA,
Pavlenko AV, Kostryukova ES, Alexeev DG, Govorun VM. Genome-​wide association ana-
lysis identifies a locus on DMD (dystrophin) gene for power athlete status in Russians. Eur J Hum
Genet 22: 502, 2014.
40. Niemi A-​K, Majamaa K. Mitochondrial DNA and ACTN3 genotypes in Finnish elite endurance
and sprint athletes. Eur J Hum Genet 13: 965–​969, 2005.
41. Orysiak J, Busko K, Michalski R, Mazur-​Różycka J, Gajewski J, Malczewska-​Lenczowska
J, Sitkowski D, Pokrywka A. Relationship between ACTN3 R577X polymorphism and maximal
power output in elite Polish athletes. Medicina (Kaunas) 50: 303–​308, 2014.
42. Papadimitriou ID, Papadopoulos C, Kouvatsi A, Triantaphyllidis C.The ACTN3 gene in elite
Greek track and field athletes. Int J Sports Med 29: 352–​5, 2008.
43. Pasqua LA, Bueno S, Matsuda M, Marquezini MV, Lima-​Silva AE, Saldiva PHN, Bertuzzi
R. The genetics of human running: ACTN3 polymorphism as an evolutionary tool improving the
energy economy during locomotion. Ann Hum Biol 4460: 1–​6, 2015.

353
354

M. Jacques et al.

44. Pereira A, Costa AM, Izquierdo M, Silva AJ, Bastos E, Marques MC. ACE I/​D and ACTN3
R/​X polymorphisms as potential factors in modulating exercise-​related phenotypes in older women
in response to a muscle power training stimuli. Age (Omaha) 35: 1949–​1959, 2013.
45. Pitsiladis YP, Tanaka M, Eynon N, Bouchard C, North KN, Williams AG, Collins M, Moran
CN, Britton SL, Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov II, de Geus E, Alsayrafi
M. Athlome Project Consortium: a concerted effort to discover genomic and other “omic” markers
of athletic performance. Physiol Genomics 48: 183–​190, 2016.
46. Puthucheary Z, Skipworth JRA, Rawal J, Loosemore M, Van Someren K, Montgomery HE.
The ACE gene and human performance: 12 years on. Sports Med 41: 433–​48, 2011.
47. Rankinen T, Fuku N, Wolfarth B, Wang G, Sarzynski MA, Alexeev DG, Ahmetov II,
Boulay MR, Cieszczyk P, Eynon N, Filipenko ML, Garton FC, Generozov EV, Govorun
VM, Houweling PJ, Kawahara T, Kostryukova ES, Kulemin NA, Larin AK, Maciejewska-​
Karlowska A, Miyachi M, Muniesa CA, Murakami H, Ospanova EA, Padmanabhan S,
Pavlenko AV, Pyankova ON, Santiago C, Sawczuk M, Scott RA, Uyba VV, Yvert T, Perusse
L, Ghosh S, Rauramaa R, North KN, Lucia A, Pitsiladis Y, Bouchard C. No evidence of a
common DNA variant profile specific to world class endurance athletes. PLoS One 11: 1–​24, 2016.
48. Rankinen T, Fuku N, Wolfarth B, Wang G, Sarzynski MA, Alexeev DG, Ahmetov II,
Boulay MR, Cieszczyk P, Eynon N, Filipenko ML, Garton FC, Generozov EV, Govorun
VM, Houweling PJ, Kawahara T, Kostryukova ES, Kulemin NA, Larin AK, Maciejewska-​
Karłowska A, Miyachi M, Muniesa CA, Murakami H, Ospanova EA, Padmanabhan
S, Pavlenko AV, Pyankova ON, Santiago C, Sawczuk M, Scott RA, Uyba VV, Yvert T,
Perusse L, Ghosh S, Rauramaa R, North KN, Lucia A, Pitsiladis Y, Bouchard C. No evi-
dence of a common DNA variant profile specific to world class endurance athletes. PLoS One
11: e0147330, 2016.
49. Roth SM, Walsh S, Liu D, Metter EJ, Ferrucci L, Hurley BF. The ACTN3 R577X nonsense
allele is under-​represented in elite-​level strength athletes. Eur J Hum Genet 16: 391–​394, 2008.
50. Schutte NM, Nederend I, Hudziak JJ, Bartels M, de Geus EJC. Twin-​sibling study and meta-​
analysis on the heritability of maximal oxygen consumption. Physiol Genomics 48: 210–​219, 2016.
51. Scott RA, Moran C, Wilson RH, Onywera V, Boit MK, Goodwin WH, Gohlke P, Payne J,
Montgomery H, Pitsiladis YP. No association between angiotensin converting enzyme (ACE)
gene variation and endurance athlete status in Kenyans. Comp Biochem Physiol A Mol Integr Physiol
141: 169–​175, 2005.
52. Scott RA, Irving R, Irwin L, Morrison E, Charlton V, Austin K, Tladi D, Deason M, Headley
SA, Kolkhorst FW, Yang N, North K, Pitsiladis YP.ACTN3 and ACE genotypes in elite Jamaican
and US sprinters. Med Sci Sport Exerc 42: 107–​112, 2010.
53. Sessa F, Chetta M, Petito A, Franzetti M, Bafunno V, Pisanelli D, Sarno M, Iuso S,
Margaglione M. Gene polymorphisms and sport attitude in Italian athletes. Genet Test Mol Biomarkers
15: 285–​290, 2011.
54. Seto JT, Quinlan KGR, Lek M, Zheng XF, Garton F, Macarthur DG, Hogarth MW,
Houweling PJ, Gregorevic P, Turner N, Cooney GJ, Yang N, North KN. ACTN3 geno-
type influences muscle performance through the regulation of calcineurin signaling. J Clin Invest
123: 4255–​4263, 2013.
55. Shang X, Zhang F, Zhang L, Huang C. ACTN3 R577X polymorphism and performance
phenotypes in young Chinese male soldiers. J Sports Sci 30: 255–​260, 2012.
56. Simonen RL, Rankinen T, Perusse L, Rice T, Rao DC, Chagnon Y, Bouchard C. Genome-​
wide linkage scan for physical activity levels in the Quebec Family study. Med Sci Sports Exerc
35: 1355–​1359, 2003.
57. Stunkard AJ, Harris JR, Pedersen NL, McClearn GE. The body-​mass index of twins who have
been reared apart. N Engl J Med 322: 1483–​1487, 1990.
58. Thomis MA, Van Leemputte M, Maes HH, Blimkie CJR, Claessens AL, Marchal G, Willems
E, Vlietinck RF, Beunen GP. Multivariate genetic analysis of maximal isometric muscle force at
different elbow angles. J Appl Physiol 82: 959–​967, 1997.
59. Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M,
Evans CA, Holt RA, Gocayne JD, Amanatides P, Ballew RM, Huson DH, Wortman JR,
Zhang Q, Kodira CD, Zheng XH, Chen L, Skupski M, Subramanian G, Thomas PD, Zhang
J, Gabor Miklos GL, Nelson C, Broder S, Clark AG, Nadeau J, McKusick VA, Zinder N,
Levine AJ, Roberts RJ, Simon M, Slayman C, Hunkapiller M, Bolanos R, Delcher A, Dew

354
355

Genetics and sport performance

I, Fasulo D, Flanigan M, Florea L, Halpern A, Hannenhalli S, Kravitz S, Levy S, Mobarry

C, Reinert K, Remington K, Abu-​Threideh J, Beasley E, Biddick K, Bonazzi V, Brandon
R, Cargill M, Chandramouliswaran I, Charlab R, Chaturvedi K, Deng Z, Di Francesco V,
Dunn P, Eilbeck K, Evangelista C, Gabrielian AE, Gan W, Ge W, Gong F, Gu Z, Guan P,
Heiman TJ, Higgins ME, Ji RR, Ke Z, Ketchum KA, Lai Z, Lei Y, Li Z, Li J, Liang Y, Lin X,
Lu F, Merkulov G V, Milshina N, Moore HM, Naik AK, Narayan VA, Neelam B, Nusskern
D, Rusch DB, Salzberg S, Shao W, Shue B, Sun J, Wang Z, Wang A, Wang X, Wang J, Wei
M, Wides R, Xiao C, Yan C, Yao A, Ye J, Zhan M, Zhang W, Zhang H, Zhao Q, Zheng L,
Zhong F, Zhong W, Zhu S, Zhao S, Gilbert D, Baumhueter S, Spier G, Carter C, Cravchik
A, Woodage T, Ali F, An H, Awe A, Baldwin D, Baden H, Barnstead M, Barrow I, Beeson
K, Busam D, Carver A, Center A, Cheng ML, Curry L, Danaher S, Davenport L, Desilets R,
Dietz S, Dodson K, Doup L, Ferriera S, Garg N, Gluecksmann A, Hart B, Haynes J, Haynes C,
Heiner C, Hladun S, Hostin D, Houck J, Howland T, Ibegwam C, Johnson J, Kalush F, Kline
L, Koduru S, Love A, Mann F, May D, McCawley S, McIntosh T, McMullen I, Moy M, Moy
L, Murphy B, Nelson K, Pfannkoch C, Pratts E, Puri V, Qureshi H, Reardon M, Rodriguez
R, Rogers YH, Romblad D, Ruhfel B, Scott R, Sitter C, Smallwood M, Stewart E, Strong
R, Suh E, Thomas R, Tint NN, Tse S, Vech C, Wang G, Wetter J, Williams S, Williams M,
Windsor S, Winn-​Deen E, Wolfe K, Zaveri J, Zaveri K, Abril JF, Guigó R, Campbell MJ,
Sjolander K V, Karlak B, Kejariwal A, Mi H, Lazareva B, Hatton T, Narechania A, Diemer
K, Muruganujan A, Guo N, Sato S, Bafna V, Istrail S, Lippert R, Schwartz R, Walenz B,
Yooseph S, Allen D, Basu A, Baxendale J, Blick L, Caminha M, Carnes-​Stine J, Caulk P,
Chiang YH, Coyne M, Dahlke C, Mays A, Dombroski M, Donnelly M, Ely D, Esparham S,
Fosler C, Gire H, Glanowski S, Glasser K, Glodek A, Gorokhov M, Graham K, Gropman
B, Harris M, Heil J, Henderson S, Hoover J, Jennings D, Jordan C, Jordan J, Kasha J, Kagan
L, Kraft C, Levitsky A, Lewis M, Liu X, Lopez J, Ma D, Majoros W, McDaniel J, Murphy S,
Newman M, Nguyen T, Nguyen N, Nodell M, Pan S, Peck J, Peterson M, Rowe W, Sanders
R, Scott J, Simpson M, Smith T, Sprague A, Stockwell T, Turner R, Venter E, Wang M,
Wen M, Wu D, Wu M, Xia A, Zandieh A, Zhu X. The sequence of the human genome. Science
291: 1304–​51, 2001.
60. Voisin S, Eynon N, Yan X, Bishop DJ. Exercise training and DNA methylation in humans. Acta
Physiol 213: 39–​59, 2015.
61. Wang G, Padmanabhan S, Wolfarth B, Fuku N, Lucia A, Ahmetov II, Cieszczyk P, Collins M,
Eynon N, Klissouras V, Williams A, Pitsiladis Y. Genomics of elite sporting performance: what
little we know and necessary advances. Adv Genet 84: 123–​149, 2013.
62. Wang G, Tanaka M, Eynon N, North KN, Williams AG, Collins M, Moran CN, Britton SL,
Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov II, de Geus E, Alsayrafi M, Pitsiladis
YP. The future of genomic research in athletic performance and adaptation to training. Med Sport Sci
61: 55–​67, 2016.
63. Willems SM, Wright DJ, Day FR, Trajanoska K, Joshi PK, Morris JA, Matteini AM, Garton
FC, Grarup N, Oskolkov N, Thalamuthu A, Mangino M, Liu J, Demirkan A, Lek M, Xu L,
Wang G, Oldmeadow C, Gaulton KJ, Lotta LA, Miyamoto-​Mikami E, Rivas MA, White
T, Loh P-​R, Aadahl M, Amin N, Attia JR, Austin K, Benyamin B, Brage S, Cheng Y-​C,
Cięszczyk P, Derave W, Eriksson K-​F, Eynon N, Linneberg A, Lucia A, Massidda M, Mitchell
BD, Miyachi M, Murakami H, Padmanabhan S, Pandey A, Papadimitriou I, Rajpal DK, Sale
C, Schnurr TM, Sessa F, Shrine N, Tobin MD, Varley I, Wain LV, Wray NR, Lindgren CM,
MacArthur DG, Waterworth DM, McCarthy MI, Pedersen O, Khaw K-​T, Kiel DP, Oei
L, Zheng H-​F, Forgetta V, Leong A, Ahmad OS, Laurin C, Mokry LE, Ross S, Elks CE,
Bowden J, Warrington NM, Murray A, Ruth KS, Tsilidis KK, Medina-​Gómez C, Estrada
K, Bis JC, Chasman DI, Demissie S, Enneman AW, Hsu Y-​H, Ingvarsson T, Kähönen M,
Kammerer C, Lacroix AZ, Li G, Liu C-​T, Liu Y, Lorentzon M, Mägi R, Mihailov E, Milani L,
Moayyeri A, Nielson CM, Sham PC, Siggeirsdotir K, Sigurdsson G, Stefansson K, Trompet
S, Thorleifsson G, Vandenput L, van der Velde N, Viikari J, Xiao S-​M, Zhao JH, Evans DS,
Cummings SR, Cauley J, Duncan EL, de Groot LCPGM, Esko T, Gudnason V, Harris TB,
Jackson RD, Jukema JW, Ikram AMA, Karasik D, Kaptoge S, Kung AWC, Lehtimäki T,
Lyytikäinen L-​P, Lips P, Luben R, Metspalu A, van Meurs JBJ, Minster RL, Orwoll E, Oei
E, Psaty BM, Raitakari OT, Ralston SW, Ridker PM, Robbins JA, Smith AV, Styrkarsdottir
U, Tranah GJ, Thorstensdottir U, Uitterlinden AG, Zmuda J, Zillikens MC, Ntzani EE,

355
356

M. Jacques et al.

Evangelou E, Ioannidis JPA, Evans DM, Ohlsson C, Pitsiladis Y, Fuku N, Franks PW, North
KN, van Duijn CM, Mather KA, Hansen T, Hansson O, Spector T, Murabito JM, Richards
JB, Rivadeneira F, Langenberg C, Perry JRB, Wareham NJ, Scott RA. Large-​scale GWAS
identifies multiple loci for hand grip strength providing biological insights into muscular fitness. Nat
Commun 8: 16015, 2017.
64. Williams AG, Day SH, Folland JP, Gohlke P, Dhamrait S, Montgomery HE. Circulating
angiotensin converting enzyme activity is correlated with muscle strength. Med Sci Sports Exerc
37: 944–​948, 2005.
65. Wolfarth B, Rankinen T, Hagberg JM, Loos RJF, Pérusse L, Roth SM, Sarzynski MA,
Bouchard C. Advances in exercise, fitness, and performance genomics in 2013. Med Sci Sports Exerc
46: 851–​859, 2014.
66. Yang N, MacArthur DG, Gulbin JP, Hahn AG, Beggs AH, Easteal S, North K. ACTN3 geno-
type is associated with human elite athletic performance. Am J Hum Genet 73: 627–​631, 2003.
67. Yang R, Shen X, Wang Y, Voisin S, Cai G, Fu Y, Xu W, Eynon N, Bishop DJ, Yan X.
ACTN3 R577X gene variant is associated with muscle-​related phenotypes in elite Chinese sprit/​
power athletes. J Strength Cond Res 31: 1107–​1115, 2016.

356
357

25
USING ELITE ATHLETES AS
A MODEL FOR GENETIC
RESEARCH
Colin N. Moran, Alun G. Williams, and Guan Wang

The general principle behind many human genetic studies is to compare the genetics of a group
of individuals with shared characteristics of interest (such as being good at sport) to a control
group lacking those characteristics.We would expect that genetic variants related to the produc-
tion of those characteristics would be more common in the former group.With as much as 66%
of athlete status being the result of heritable factors (13), elite athletes must have a genetic pro-
file that predisposes them to being good at sport, responsive to training, and resistant to injury as
well as ensuring that they train hard, eat well, and have a skilled team of coaches around them.
The more extreme and well-​defined this group of athletes is, the more pronounced we would
expect any genetic difference to be. Defining that group of individuals well is key to this process.

Defining athletes as elite for genetics research

Elite is defined as “a select group that is superior in terms of abilities or qualities to the rest of
a group or society” (37). In terms of athletes, although a consensus definition is not clear, those
achieving selection for international teams are often described as elite (50). However, athletes in
the Olympic team of any country will come in a variety of shapes and sizes, depending on their
sport of choice and particular abilities. Elite distance runners are typically small in stature, with
thin limbs, high type I muscle fiber content, and a high maximal oxygen uptake (VO2max) (21,
33, 44) (Figure 25.1). In contrast, elite weightlifters are typically large, with thick limbs, high
type II muscle fiber content, and a lower VO2max (25, 49) (Figure 25.1). Many physiological
differences can be observed between athletes specializing in different sports.When investigating
the genetic components of elite athletes’ makeup, it is important to know which sport an athlete
is considered to be elite in and what the physiological demands are of that sport.
Grouping together athletes from sports with similar physiological demands is useful.
Grouping athletes specializing in strength sports and comparing them to nonathlete controls
may tell us something about the genetics that predispose to being good at strength or power
events. However, even within one sporting discipline, there are often many different ways to get
into the elite group. Typically, athletes will be selected on performance criteria, e.g., what is the
fastest time that they have run? However, two elite endurance athletes with identical 10 km per-
formance times may have achieved them quite differently and some of those differences may be

357
358

C.N. Moran et al.

Figure 25.1 Elite distance runner and elite weight lifter: Almaz Ayana (left) and Mi-​ran Jang (right).
Both are Olympic gold medalists. Ayana pictured at the 2017 world championships in London in the
10,000 m and Jang at the 2012 London Olympics in the women’s +75 kg category, respectively.
Source: Photo of Almaz Ayana by Marco Verch. Photo of Mi-​ran Jang by Korean Olympic Committees.

underpinned by genetics. One may have a sprint finish, while the other does not. One may have
a phenomenal VO2max and an average running style, while conversely the other has an average
VO2max and an exceptional running economy.To understand the genetics of elite athletes, it is
important to recognize that athletes achieve elite performance in different ways (11, 23).
Defining elite across multiple sports by a single performance criterion is even more fraught.
The combinations of physiological abilities necessary to reach the top echelons are unlikely
to be the same in different sports (27) although some common factors exist. Many athletes
make use of this and switch sports in search of success. Indeed, if athletes do not quite achieve
medal-​winning performance in their chosen sport but do have some elite-​level physiological
characteristics, sports governing bodies now encourage them to switch sports through talent
transfer programs (e.g., 52). Of course, the amount of competition may be lower in their new
sport, but assuming that it is not and that they achieve success, one could conclude that their
combination of genetic predispositions is better suited to their new sport. An athlete could be
considered to be elite, or not, in each of several individual characteristics important for perform-
ance. However, it is also important to recognize that they will likely achieve elite performance
without being elite in every single characteristic that contributes.
Participation rates in many sports differ vastly across borders and sexes as do the standards
that would be considered to be elite. Some sports are found in very few countries having barely
spread from their country of origin, such as shinty in Scotland (9) (Figure 25.2). Being the best
shinty player in Scotland is a different achievement from being the best shinty player in virtually
any other country.When participation rates for a sport are low in a country, the individuals with
the most suitable genetic profile may not take part in that sport; they may do a different sport

358
359

Elite athletes as a model for research

Figure 25.2 Shinty skill from the Kinlochshiel versus Strathglass Balliemore Cup Final (2009).
Source: Photo by Alasdair Middleton.

or they may be inactive. When participation rates are low, there is a higher chance that the best
individuals at that sport are there by accident of their involvement and because of their effort
in training rather than their genetic predisposition. Devon Harris, Michael White, and brothers
Dudley and Chris Stokes were an elite Jamaican bobsleigh crew having competed at the 1988
Calgary Olympics (26), immortalized by the 1993 film Cool Runnings (22). However, they
may not be considered truly elite against German or Swiss norms where winter sports are more
common.There may even be other Jamaicans with a more suitable genetic profile for bobsleigh
who have never had the opportunity to try it. In case–​control studies, this can lead to misclas-
sified individuals in the control group, where those with a genetic predisposition to athleticism
are included as controls because not everyone with a genetic predisposition gets involved in
sport. Since control groups are rarely screened to exclude potential cases, this causes a, typically
modest, loss of statistical power (31). Additionally, given the global variation in allele frequencies,
controls must be ethnically matched with cases and assessed to avoid confounding by popu-
lation stratification. Conversely, even when participation rates are high, there are few sports
where the top women’s achievements would be considered elite compared to the top men in
the same sport (51). Or where junior athletes could compete with adults in their prime (6, 24).
Nonetheless, they absolutely are elite when compared to the appropriate group norms. When
we make use of an elite group, to understand how genetics influences those characteristics, we
have to take great care in defining that group and the criteria against which we judge individual
achievements.
Within sport there are a variety of ways to define an individual as elite and each comes with
its own strengths and weaknesses. Defining elite based on competitive achievements allows
rapid and relatively easy identification of appropriate individuals although it may create both
athlete and control groups with some mixed physiological abilities. Defining elite based on
physiological characteristics is more time-​consuming for both researchers and athletes; however,
it may allow more appropriate groupings to be identified if a single parameter is of primary
interest, e.g.,VO2max, because of its known relationships with morbidity and mortality (5, 15).
However, if the primary interest is elite performance in a sport per se, then we must know all of

359
360

C.N. Moran et al.

those individual characteristics to take this approach and address each one in turn. Restricting
a definition of elite to a single phenotype measured in a laboratory instead of success in that
sport would be unable to identify the full range of genetic characteristics important for that
sporting success. Ideally, we would select athletes from sports with high levels of participation
where everyone puts in an equal and large amount of effort in training and we would be careful
to compare their achievements to an appropriate set of thresholds for their age or sex group.

Elite athletes can be used to model athletic ability, health, and disease
The central component of any study design is the academic question to be addressed. Clearly
elite athletes can be used to model the genetics of being an elite athlete. Depending on which
types of athletes have been chosen for study and how elite has been defined, we may ask: Which
genetic variants are required to be an elite weightlifter? Or result in a high VO2max? However,
while understanding the required components for an elite athlete may motivate many researchers,
elite athletes make up only a small proportion of the population and have no specific associated
genetic illness that requires attention.There is no imperative to investigate their genetic makeup
and such studies are less likely than research into health or disease to attract the necessary grant
funding for expensive genetics research. Nonetheless, there is also value in using elite athletes as
a model for a number of health and disease research questions less obviously related to athletic
performance. Most obvious are the links between exercise and health or all-​cause mortality
(41, 45). However, elite athletes are extreme, not only in their abilities, but also in the level of
stress that they put on their bodies: training over and over again, experiencing muscular, bone,
metabolic, and immunological stresses, and pushing themselves daily (4, 10, 47). Diseases often
put extreme stresses on similar aspects of our physiology and with severities that vary between
individuals partly due to their underlying genetics. For instance, many cancers result in a loss
of skeletal muscle mass known as cachexia which correlates with shorter survival in patients
(2). Sarcopenia is the age-​related decline in skeletal muscle mass, which occurs at different
rates in different individuals, and leads to increased risk of falls, fracture, and frailty (43). If we
overload our bodies with energy, some individuals develop obesity and diabetes, placing a var-
iety of stresses on our metabolism (54). Understanding what happens to elite athletes and why
one copes better with particular stressors than another will give insights into common diseases
suffered by individuals all across the population distribution.

Super athletes, lifestyle disease, and ageing

Individuals with noncommunicable diseases may have even more in common with elite athletes.
Although genetic variants predispose to athletic ability (46), lifestyle is also a key component
and genetic variants that are an advantage in one situation may be neutral or even a disadvan-
tage in another. If the individual with the best genetic profile in the world for elite cycling
never gets on a bicycle, she won’t be an elite cyclist. In fact, some of the genetic variants that
predispose her to being an elite athlete may also predispose her to being unhealthy when there
is an absence of the correct lifestyle factors. For example, if part of her genetic predisposition
to cycling results from genetic variants making her highly efficient at utilizing the energy she
consumes, the lifestyle choices she makes about how to use that energy have particularly sig-
nificant impacts. If she chooses to cycle, she has an advantage. If instead she is physically inactive,
she may be at a higher risk of metabolic disease as she would still be highly efficient with the
energy she consumes. This concept is known as the thrifty genotype (35). Many variants can
have differing effects depending on the lifestyle or environment in which we chose to consider

360
361

Elite athletes as a model for research

Figure 25.3 Photograph of the Finnish skier Eero Mäntyranta (1937–​2013) during the 1964 Winter
Olympics in Innsbruck.

them.The multi-​Olympic and world championship medal-​winning Finnish cross-​country skier

Eero Mäntyranta (36) (Figure 25.3) had a rare variant (rs121917830) of the erythropoietin
receptor gene (EPOR), effectively making him hypersensitive to his body’s normal erythro-
poietin levels (34). This condition, known as erythrocytosis, greatly increased his red blood
cell count and the oxygen-​carrying capacity of his blood giving him a natural advantage as an
elite endurance athlete. However, the additional blood cells thicken the blood of carriers of
this genetic variant putting them at increased risk of having abnormal blood clots with poten-
tial life-​threatening complications (8). The consequences of these genetic variants cannot be
separated from the environments in which we place them. Investigating how the environment
that we control, such as physical activity or dietary choices, interacts with genetic variants in
elite athletes pushing themselves to their limits will help us understand how those same gen-
etic variants may interact with lifestyle choices in the general population or unhealthy people.
Additionally, there are clear and important links between components of athlete physiology
and many common diseases, such as VO2max and predisposition to chronic disease and lon-
gevity (7). Being physically active is thought to be the biological default condition for health,
and optimizes the aging process in humans (16). Elite and masters athletes are usually lifelong
exercisers, and serve as a unique biological model for studying the fundamental physiological
mechanisms underlying health and ageing. It is these relationships that should drive research in
this area, and research funding, forward. Elite athletes can be used to improve our understanding
of the genetic components of these aspects of physiology. In turn, this will lead to a better
understanding of determinants of the long-​term health of the general population and what we
might do to improve them. The impact of this type of research is likely to be much broader
than the elite athletes themselves and ultimately will tell us about genetic aspects of health and
disease.

Pluses and minuses of working with elite athletes

Of course, it is possible to investigate the genetics of athletic ability and the genetics of health and
disease in less elite individuals: those with disease, the general population, or the recreationally

361
362

C.N. Moran et al.

active. However, there are a number of advantages to working with elite athletes. As mentioned
above, they are on an extreme part of the phenotypic distribution.This can only be achieved by
having both the appropriate genetic profile and the correct environment in terms of nutrition
and training. Variation in both genetic predisposition and the appropriateness of the environ-
ment is higher in the general population or the recreationally active making it harder to tease
out the genetic components. Many performance tasks that we may use in the physiology labora-
tory are complex or difficult, such as lifting weights, meaning that lack of familiarity can create
variation in performance that is unrelated to physiological ability or genetic predisposition (42).
Similarly, measuring VO2max requires that individuals push themselves to their physiological
limits, something nonathletes may be less able to achieve (28). In essence, elite athletes are more
similar to each other, providing a cleaner phenotype and hopefully a cleaner window on to the
key aspects of their genetic profile.
Conversely, working with elite athletes presents several problems. Major genetics of common
disease initiatives in recent years have recruited increasingly large numbers of participants to
allow the detection of small genetic effects using genome-​wide association studies (GWAS).
This approach has allowed identification of more than 4.3 million common genetic variants as
important for the characteristic in question whether we expect that they might be involved or
not. Early GWAS experiments had little more than a hundred participants, but the most recent
have a few hundred thousand participants with plans for even larger future studies (29). Ideally,
we would take a similar approach to elite performance. However, by definition, elite athletes
are rare, or they would not be distinguishable from the general population, meaning that this
powerful approach is not so straightforwardly applied. Additionally, in laboratory studies we typ-
ically control the exercise and nutrition of our participants so that it is uniform. Elite athletes
are very unlikely to be willing to deviate from their own personal exercise and nutrition plans.
Similarly, no one becomes an elite athlete without spending a great deal of time training, often
more than once a day. Consequently, elite athletes are busy. They may not have time or be
willing to do detailed physiological testing in the laboratory, unless it can directly inform their
training (30).
Although there are pros and cons to working with elite athletes, there is clear value in
understanding why they are so good at sport. Understanding how we can accommodate the
other demands on their time and ensuring that we provide information useful to them will help
create an environment where they are willing to be involved in research. As with any academic
research, the place to start when designing a study is by talking to the stakeholders, primarily
the athletes and the teams of people around them, so that studies are designed to be mutually
beneficial.

Global efforts to investigate elite athletes

There are a number of projects that have set out to understand genetic aspects of elite perform-
ance in athletes. They range in size greatly, include varying degrees of elite athletes, and have
taken a number of different approaches to study design. Often there is a direct trade-​off between
the quality and detail of the phenotype data collected and the number of elite athletes willing
to take part. The largest cohorts typically are those that simply require information on pre-
ferred discipline and a saliva sample for DNA extraction. Some studies include more categorical
information, such as highest level of competition or greatest achievements without requiring
much more time from the athletes. Some studies make use of existing quantitative information.
This may be in the form of personal best competition race times or weights lifted; or it may
be testing or biochemical data that at some stage was collected from the athletes to benchmark

362
363

Elite athletes as a model for research

their performance or improve their training. Again, this requires little more of the athletes’ time,
although it will require additional ethical permission to access databases and probably more
time from someone in the support organization that collected the database. Finally, it may be
possible to collect new physiological testing information on the athletes related to their elite
performance or the performance of others in the cohort. However, this requires considerable
time from athletes and makes it harder to collect large numbers of athletes.
One of the first studies to use elite athletes to investigate the genetics of sporting per-
formance was the GenAthlete study (labs.pbrc.edu/​humangenomics/​#GENATHLETE). This
multicenter, case–​control study was launched in 1993, primarily to identify genetic variants
present at different frequencies in endurance athletes with high VO2max and untrained controls
with low to average VO2max. Male endurance athletes of national or international caliber with
a VO2max of at least 75 ml/​kg/​min and male controls were recruited from Canada, Finland,
Germany, and the US. The cohort includes 315 elite endurance athletes (mean VO2max of 79
ml/​kg/​min) and 320 matched controls (mean VO2max of 40 ml/​kg/​min). Multiple candidate
gene studies and a GWAS have been performed using the resources of GenAthlete (40). Further
studies are focusing on nuclear and mitochondrial DNA sequencing.
Subsequent studies often took a similar approach, recruiting from national or international
teams, although most do not have VO2max cut-​offs as inclusion criteria. Nonexhaustive lists of
cohorts and consortia known to include elite athletes are given in Tables 25.1 and 25.2 and the
following paragraphs.
Despite numerous publications arising from these cohorts advancing our understanding of
the genetics of sporting performance, human genetic approaches have moved on to require
larger and larger cohorts for research. Researchers in this field now recognize that elite athletes
are too rare to allow rapid collection of individual DNA cohorts that are large enough for
adequate analysis. Pooling cohorts from different populations presents technical difficulties
although these are well understood and can be overcome. One advantage is that the alleles
identified will be robust across multiple population groups. Consequently, several collaborative
consortia have been set up to allow pooling of precious resources.
The Athlome Project Consortium (www.athlomeconsortium.org/​) developed from a 2015 sym-
posium in Santorini, Greece, and attempts to bring together all current studies including other
consortia and those recently launched into one large collaborative initiative (38). It includes
studies on elite athletes, on adaptation to training in both humans and animal models, and on
exercise-​related injuries. It has four aims: 1) to establish an ethically sound international research
consortium and biobank resource systematically across individual centers; 2) to discover genetic
variants associated with exercise performance, adaptive response to exercise training and skeletal
muscle injuries using the GWAS approach, targeted sequencing, or whole-​genome sequencing;
3) to validate and replicate the genetic markers from the discovery phase across sex and ethni-
city; and 4) to conduct functional investigations following replicated findings to better under-
stand the associated biology. Athlome includes the 1000 Athlome sequencing project which aims
to sequence 1000 genomes of sprinters and distance runners of West and East African descent,
respectively. Major impacts of the consortium include setting a direction for researchers in the
field, fostering links between global researchers, and producing a widely read position statement
on the direct-​to-​consumer genetic testing kits for sporting performance (55).
The Exercise at the Limit –​Inherited Traits of Endurance (ELITE; www.med.stanford.
edu/​elite.html) consortium involves researchers and athletes from all over the globe with the
aim of mapping the role that genetic factors play in determining VO2max. The strategy is to
recruit athletes with very extreme VO2max rates, creating a clean phenotype group for com-
parison to controls. The main inclusion criterion for males is to have a VO2max of ≥75 ml/​

363
364

Table 25.1 Major study cohorts in the genetics of human elite athletic performance

Study Primary design Participants

GENATHLETE Case–​control association study 315 Male endurance athletes of national

cohort involving endurance athletes with or international caliber with a
high VO2max and untrained VO2max of at least 75 ml/​kg/​min
controls with low to average recruited from Canada, Finland,
VO2max Germany, and the US
320 Male nonathlete controls matched
to athletes
UK cohort Case–​control association study 800 Rugby athletes (includes some
(The GENESIS involving primarily rugby South African), male, elite
project, athletes (RugbyGene) and (compete in top professional
including endurance runners, but leagues)
RugbyGene) also including football and 675 Endurance runners, male and
speed/​strength/​power sports. female, international/​national/​
Phenotype data on some regional standard including
athletes include dual-​energy Olympic athletes
X-​ray absorptiometry, vertical 125 Football (soccer), male,
jump, isometric mid-​thigh international/​national/​regional
pull, independently verified standard
personal best competitive 200 Miscellaneous other athletes
performances, in-​game key including sprinters, jumpers,
performance indicators in team throwers, basketball, field hockey,
sport competitions and certain swimming, cycling, male and
categories of injury history female, international/​national/​
regional standard
1125 Nonathlete controls
Eastern Case–​control association studies 9144 Male and female, international/​
European involving 28 different athletic national/​regional/​local/​
cohorts disciplines. The phenotype data noncompetitive standard
(Belarus, on the Lithuanian cohort is 4974 Nonathlete controls
Lithuania, comprised of over 20 groups
Russia, and of phenotypic measurements
Ukraine) including anthropometrics,
isokinetic tests,VO2 max,
Wingate, cardiac size and
function and blood pressure
Polish cohort Case–​control association study 660 Male and female, international/​
involving 20 different athletic national standard
disciplines 684 Polish controls
Australian Case–​control association study 125 Endurance runners, male and
cohort female, international/​national/​
regional standard
84 Sprinters, male and female,
international/​national/​regional
standard
258 Australian controls
365

Table 25.1 (Cont.)

Study Primary design Participants

Spanish cohort Case–​control association study 100 Endurance athletes of world-​class
status, male
54 Rowers of world-​class status, male
108 All-​time best judo athletes, male
88 Swimmers, male and female,
national standard
53 Power athletes, male, international/​
national standard
343 Spanish controls
European Candidate gene case–​control study 200 Caucasian swimmers of world-​class
and Asian status, male and female
swimming 158 Japanese swimmers, male and
cohort female, national standard
649 Japanese controls
168 Taiwanese swimmers, male and
female, international/​national
standard
603 Taiwanese controls
Chinese cohort Candidate gene case–​control study 241 Chinese (Han) endurance athletes,
male and female, national
standard
504 Chinese (Han) controls
Japanese cohort Case–​control association study >4000 Endurance and sprint athletes,
male and female, international/​
national standard including track
and field athletes, swimmers and
athletes of team sports
>1000 Japanese controls
Israeli cohort Candidate gene case–​control study 74 Endurance athletes, male and
female, international/​national
standard
81 Power athletes, male and female,
international/​national standard
240 Israeli controls
East African Case–​control association study 291 Kenyan endurance athletes, male
cohort and female, international/​
national standard
85 Kenyan controls
76 Ethiopian endurance athletes, male
and female, from junior/​senior
national athletic teams
315 Ethiopian controls
(continued)
366

Table 25.1 (Cont.)

Study Primary design Participants

Jamaican and Case–​control and genome-​wide 116 Jamaican sprint athletes, male and
US cohort association studies female, international/​national
standard
311 Jamaican controls
114 African American sprint athletes,
male and female, international/​
national standard
191 African American controls

Table 25.2 Major study consortia in the genetics of human elite athletic performance (excluding recently
established consortia)

Study Primary design Participants

ATHLOME Consortium bringing >19,202 Total athletes

Consortium together many >14,733 Total controls
of the individual Contributing cohorts: GENATHLETE,
cohorts and Eastern European, East African,
consortia Japanese, as well as the Netherlands
Twin Register, super-​athletes, and rat
models of exercise and health
Contributing consortia: ELITE,
GAMES, GENESMART, GENESIS,
GOINg, POWERGENE
ELITE Consortium Case–​control 800 Male (640) and female (160) endurance
association athletes with a VO2max of at least 75
study involving ml/​kg/​min (males) or 63 ml/​kg/​min
endurance (females)
athletes with high >1000 Matched controls from databases of
VO2max samples previously genotyped on same
platform
Contributing cohorts: GENATHLETE,
Eastern European, and Japanese with
many individual samples recruited
directly from around the globe
GAMES Case–​control 1520 Elite endurance athletes
Consortium association 2760 Ethnically matched controls
study involving Contributing cohorts: GENATHLETE,
endurance Australian, Ethiopian, Japanese,
athletes Kenyan, Polish, Russian, and Spanish
POWERGENE Case–​control ~500 Elite sprint/​power athletes
Consortium association study ~1400 Population matched controls
involving elite Contributing cohorts: Australian,
sprint/​power Lithuanian, Russian, Polish, South
athletes African, Spanish, UK and US with
additional samples from Belgium,
Greece, Italy, and Brazil
367

Elite athletes as a model for research

kg/​min; for females it is to have a VO2max of ≥63 ml/​kg/​min. There are currently no race
performance criteria for inclusion in the study to avoid recruiting athletes who win races by
being elite in aspects important for endurance performance other than VO2max, although, typ-
ically, the athletes are successful in endurance sports. Currently, there are 800 athlete samples
(including 160 from females) collected from around the world and the consortium is continu-
ously expanding to include new partners and regions.
GAMES is an international consortium established to compare elite endurance athletes and
ethnically matched controls in a case–​control study design. It includes 1520 endurance athletes,
including Olympic and world champions, and 2760 controls. GWASs identified 45 promising
candidate SNPs although the study was underpowered to detect small effect sizes and only one
single nucleotide polymorphism (SNP) was successfully replicated in seven additional cohorts
(40). Conversely, POWERGENE is an ongoing international consortium aiming to investigate
the genetic components of elite sprint/​power performance. Currently, it includes over 500
sprint/​power athletes and 1400 population controls from nine countries.
Other projects leverage unique aspects of the environment in certain specific sports. Rugby
players require combinations of both strength and endurance. Nonetheless, there are con-
siderable differences in the anthropometric characteristics, physical demands, and movement
patterns of players in different positions on the rugby pitch (39, 48). RugbyGene, with a current
DNA bank from over 700 elite rugby athletes, makes use of the fact that players specializing in
different positions on the rugby pitch require different combinations of these attributes yet train
together as a squad making their training intensity, training frequency, and some game demands
more similar than when strength and endurance are investigated using separate sports (17–​20).
Another strength of RugbyGene is a clear definition of eliteness in the context of a team sport,
namely regular participation in the top professional league in a major rugby playing country.
Similar research efforts in other team sports, e.g., football (soccer), should also prioritize the
broader competition for elite status that well-​financed competitive leagues provide over inter-
national competitive status that requires different performance standards in different countries,
as explained earlier.
New projects investigating aspects of elite sporting performance continue to develop, such
as GeneSMART (genes and the Skeletal Muscle Adaptive Response to Training) and GENESIS
(Genetics of Elite Status in Sport) that includes over 1700 athletes, mainly endurance runners
and the RugbyGene cohort just described, plus selected performance and injury phenotype
data for some athletes. The recently established GOINg (Genomics Of INjuries) consortium
aims to identify genetic variants that modify the risk of anterior cruciate ligament injuries.
Additionally, some very large cohorts of nonathletes exist which nonetheless contain quantita-
tive measures related to performance. For instance, the UK Biobank (www.ukbiobank.ac.uk/​)
contains handgrip strength data for over 140,000 participants. One caveat is that typically these
very large cohorts only collect simple portable measures of physiological phenotypes rather
than gold standard performance-​related measurements and the handgrip strength phenotype
has several limitations (12). Nevertheless, UK Biobank and other nonelite cohorts may provide
insight into the genes that we might expect to contain variants related to elite performance by
leveraging substantial statistical power to detect subtle influences of genetic variants.

The future of genetics research into elite athletes

To date, studies into the genetics of human elite performance have identified over 200 gene
variants involved in determining the genetic component (46). Perhaps the best known are in
alpha-​actinin-​3 (ACTN3) (56) and the angiotensin-​converting enzyme (ACE) (32). However,

367
368

C.N. Moran et al.

missing heritability is as much an issue in the genetics of elite performance as other fields and
it is debatable whether many of the variants associated with performance will stand up to the
scrutiny of further investigation. Many have been identified in a single, small candidate gene
study using only a case–​control approach. Undoubtedly, some will be successfully replicated
and we will continue to gradually build up a picture of the genetic component of elite athletic
performance. However, to move the field forward, we must reach towards the highest standards
in human genetics research. Studies must be bigger and should preferably involve both dis-
covery and replication cohorts. Several of the cohorts previously discussed are still collecting
samples and continue to grow. The recent formation of consortia, such as Athlome, ELITE, and
POWERGENE, is a recognition that the scale of the work needed requires collaboration rather
than competition between research groups, to ensure that studies have enough elite athletes
involved to provide meaningful insight into their underlying genetics.
In recent years, the techniques involved in capturing genetic variation have been evolving
and becoming more effective, comprehensive, and economic. Not only has our ability to detect
SNP variation improved, but we now additionally collect information on structural variants
such as copy number variation and heritable epigenetic marks such as DNA methylation or
noncoding RNAs. Significantly, these epigenetic marks are believed to be both heritable across
generations and malleable acutely (1, 3).Thus, in the future it may be important to also consider
the time of day and the recent activity and food consumption of elite athletes when collecting
samples from them to extract their DNA and RNA. Investigating genetic variations in multiple
ethnic groups is also vital to allow assessment of the broader relevance of a finding and identi-
fication of any alleles with population-​specific effects. This would additionally provide insights
into the heritability of a trait across populations. Understanding this diversity will lay the foun-
dation for successful genetics research into a given trait, as well as impact potential applications
in precision medicine. Future studies will integrate this complex information into a more com-
plete understanding of the heritable aspects of elite sporting performance.
The interest in this area means that new cohorts will continue to develop. Many sporting
bodies hold detailed physiological performance data on current and retired athletes collected to
help benchmark and develop the athletes in their care. With appropriate negotiation, these pro-
vide an opportunity to quickly develop large study cohorts with the addition of retrospective
DNA collection. There are also a number of ongoing studies into the growing area of athlete
health, which could have DNA collection added to them. For example, there is a growing rec-
ognition of the damage that certain sporting activities may do to aspects of long-​term health,
such as concussion or mild traumatic brain injury during normal sporting activity that may be
linked to later neurodegenerative disease (e.g., 14). Other studies collecting detailed phenotypic
information on athletes continue to develop, such as the Retired Athlete Health Project (53).
Adding DNA collection to these or other currently nongenetic studies will provide further
opportunities to develop our understanding of elite athletes.
Once variants have been conclusively related to performance, it will become important to
understand how they could functionally impact phenotypes important for performance. This
will require carefully controlled laboratory studies, including some with elite athletes, to better
understand the detail of the physiological differences resulting from the genetic differences. It
will also involve the creation of appropriate animal or cellular models of key genetic variants
for in vivo and ex vivo experiments. A combination of biological sciences and bioinformatics
will provide insights into the biomechanisms underlying health and disease through modeling
human elite athletic performance. Perhaps most importantly, they will also allow the develop-
ment of novel strategies that might mitigate some of these functional differences. Ultimately, this
will mean the development of personalized or precision training plans, diets, medical treatments,

368
369

Elite athletes as a model for research

and lifestyle advice in general and may contribute to the development of novel gene therapy
targets for muscular or respiratory conditions.

Conclusion
Investigating the genetic component of human elite athletic performance provides a unique
opportunity to tease out the consequences of a variety of rare and common genetic variants for
athletic ability as well as health and disease more generally. It will be key to develop cohorts of
truly elite individuals large enough in number, and with enough detailed physiological data to
provide the necessary resolution, to explore the findings using multidisciplinary approaches and
to advance biological and clinical research.

References
1. Baggish AL, Hale A, Weiner RB, Lewis GD, Systrom D, Wang F, Wang TJ, and Chan SY.
Dynamic regulation of circulating microRNA during acute exhaustive exercise and sustained aerobic
exercise training. J Physiol 589: 3983–​3994, 2011.
2. Baracos VE, Martin L, Korc M, Guttridge DC, and Fearon KCH. Cancer-​associated cachexia.
Nat Rev Dis Primers 4: 17105, 2018.
3. Barres R, Yan J, Egan B, Treebak JT, Rasmussen M, Fritz T, Caidahl K, Krook A, O’Gorman
DJ, and Zierath JR. Acute exercise remodels promoter methylation in human skeletal muscle. Cell
Metab 15: 405–​411, 2012.
4. Bennell KL and Brukner PD. Epidemiology and site specificity of stress fractures. Clin Sports Med
16: 179–​196, 1997.
5. Blair SN. Physical fitness and all-​cause mortality. JAMA 262: 1989.
6. Blythe DA and Kiraly FJ. Prediction and quantification of individual athletic performance of
runners. PLoS One 11: e0157257, 2016.
7. Booth FW, Roberts CK, and Laye MJ. Lack of exercise is a major cause of chronic diseases. Compr
Physiol 2: 1143–​1211, 2012.
8. Braekkan SK, Mathiesen EB, Njolstad I, Wilsgaard T, and Hansen JB. Hematocrit and risk of
venous thromboembolism in a general population.The Tromso study. Haematologica 95: 270–​275, 2010.
9. Camanachd Association. History. www.shinty.com/​mens/​about-​us/​history. [Accessed February
22, 2018.]
10. Clark A and Mach N. Exercise-​induced stress behavior, gut-​microbiota-​brain axis and diet: a sys-
tematic review for athletes. J Int Soc Sports Nutr 13: 43, 2016.
11. Coh M, Milanovic D, and Kampmiller T. Morphologic and kinematic characteristics of elite
sprinters. Coll Antropol 25: 605–​610, 2001.
12. Cronin J, Lawton T, Harris N, Kilding A, and McMaster DT. A brief review of handgrip
strength and sport performance. J Strength Cond Res 31: 3187–​3217, 2017.
13. De Moor MH, Spector TD, Cherkas LF, Falchi M, Hottenga JJ, Boomsma DI, and De Geus
EJ. Genome-​wide linkage scan for athlete status in 700 British female DZ twin pairs. Twin Res Hum
Genet 10: 812–​820, 2007.
14. Di Virgilio TG, Hunter A, Wilson L, Stewart W, Goodall S, Howatson G, Donaldson DI, and
Ietswaart M. Evidence for acute electrophysiological and cognitive changes following routine soccer
heading. EBioMedicine 13: 66–​71, 2016.
15. Fogelholm M. Physical activity, fitness and fatness: relations to mortality, morbidity and disease risk
factors. A systematic review. Obes Rev 11: 202–​221, 2010.
16. Harridge SD and Lazarus NR. Physical activity, aging, and physiological function. Physiology
(Bethesda) 32: 152–​161, 2017.
17. Heffernan SM, Kilduff LP, Day SH, Pitsiladis YP, and Williams AG. Genomics in rugby
union: a review and future prospects. Eur J Sport Sci 15: 460–​468, 2015.
18. Heffernan SM, Kilduff LP, Erskine RM, Day SH, McPhee JS, McMahon GE, Stebbings GK,
Neale JP, Lockey SJ, Ribbans WJ, Cook CJ, Vance B, Raleigh SM, Roberts C, Bennett MA,
Wang G, Collins M, Pitsiladis YP, and Williams AG. Association of ACTN3 R577X but not

369
370

C.N. Moran et al.

ACE I/​D gene variants with elite rugby union player status and playing position. Physiol Genomics
48: 196–​201, 2016.
19. Heffernan SM, Kilduff LP, Erskine RM, Day SH, Stebbings GK, Cook CJ, Raleigh SM,
Bennett MA, Wang G, Collins M, Pitsiladis YP, and Williams AG. COL5A1 gene variants pre-
viously associated with reduced soft tissue injury risk are associated with elite athlete status in rugby.
BMC Genomics 18: 820, 2017.
20. Heffernan SM, Stebbings GK, Kilduff LP, Erskine RM, Day SH, Morse CI, McPhee JS,
Cook CJ, Vance B, Ribbans WJ, Raleigh SM, Roberts C, Bennett MA, Wang G, Collins M,
Pitsiladis YP, and Williams AG. Fat mass and obesity associated (FTO) gene influences skeletal
muscle phenotypes in non-​resistance trained males and elite rugby playing position. BMC Genet
18: 4, 2017.
21. Holloszy JO and Coyle EF. Adaptations of skeletal muscle to endurance exercise and their meta-
bolic consequences. J Appl Physiol Respir Environ Exerc Physiol 56: 831–​838, 1984.
22. IMDb. Cool Runnings (1993). www.imdb.com/​title/​tt0106611/​. [Accessed February 22, 2018.]
23. Joyner MJ and Coyle EF. Endurance exercise performance: the physiology of champions. J Physiol
586: 35–​44, 2008.
24. Knechtle B, Assadi H, Lepers R, Rosemann T, and Rust CA. Relationship between age and
elite marathon race time in world single age records from 5 to 93 years. BMC Sports Sci Med Rehabil
6: 31, 2014.
25. Kraemer WJ, Fleck SJ, and Evans WJ. Strength and power training: physiological mechanisms of
adaptation. Exerc Sport Sci Rev 24: 363–​397, 1996.
26. Lamont T. Frozen in time: Jamaica’s bobsled team, Calgary Winter Olympics, 13 Feb 1988. In: The
Guardian. Online: www.theguardian.com, 2010.
27. Lorenz DS, Reiman MP, Lehecka BJ, and Naylor A.What performance characteristics determine
elite versus nonelite athletes in the same sport? Sports Health 5: 542–​547, 2013.
28. Magnan RE, Kwan BM, Ciccolo JT, Gurney B, Mermier CM, and Bryan AD. Aerobic capacity
testing with inactive individuals: the role of subjective experience. J Phys Act Health 10: 271–​279, 2013.
29. Manolio TA. In retrospect: a decade of shared genomic associations. Nature 546: 360–​361, 2017.
30. Martindale R and Nash C. Sport science relevance and application: perceptions of UK coaches. J
Sports Sci 31: 807–​819, 2013.
31. McCarthy MI, Abecasis GR, Cardon LR, Goldstein DB, Little J, Ioannidis JP, and Hirschhorn
JN. Genome-​wide association studies for complex traits: consensus, uncertainty and challenges. Nat
Rev Genet 9: 356–​369, 2008.
32. Montgomery HE, Marshall R, Hemingway H, Myerson S, Clarkson P, Dollery C, Hayward
M, Holliman DE, Jubb M, World M, Thomas EL, Brynes AE, Saeed N, Barnard M, Bell JD,
Prasad K, Rayson M, Talmud PJ, and Humphries SE. Human gene for physical performance.
Nature 393: 221–​222, 1998.
33. Mooses M and Hackney AC.Anthropometrics and body composition in East African runners: poten-
tial impact on performance. Int J Sports Physiol Perform 12: 422–​430, 2017.
34. Moran CN and Pitsiladis YP. Tour de France Champions born or made: where do we take the
genetics of performance? J Sport Sci 35: 1411–​1419, 2017.
35. Neel JV. The “thrifty genotype” in 1998. Nutr Rev 57: S2–​S9, 1999.
36. Olympic.org. Eero Mäntyranta. www.olympic.org/​eero-​mantyranta. [Accessed February 23, 2018.]
37. Oxford Dictionaries. Definition of Elite. In: en.oxforddictionariescom/​. Oxford: Oxford University
Press, 2018.
38. Pitsiladis YP, Tanaka M, Eynon N, Bouchard C, North KN, Williams AG, Collins M, Moran
CN, Britton SL, Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov, II, de Geus E, Alsayrafi
M, and Athlome Project Consortium.Athlome Project Consortium: a concerted effort to discover
genomic and other “omic” markers of athletic performance. Physiol Genomics 48: 183–​190, 2016.
39. Quarrie KL, Hopkins WG, Anthony MJ, and Gill ND. Positional demands of international rugby
union: evaluation of player actions and movements. J Sci Med Sport 16: 353–​359, 2013.
40. Rankinen T, Fuku N, Wolfarth B, Wang G, Sarzynski MA, Alexeev DG, Ahmetov, II,
Boulay MR, Cieszczyk P, Eynon N, Filipenko ML, Garton FC, Generozov EV, Govorun
VM, Houweling PJ, Kawahara T, Kostryukova ES, Kulemin NA, Larin AK, Maciejewska-​
Karlowska A, Miyachi M, Muniesa CA, Murakami H, Ospanova EA, Padmanabhan
S, Pavlenko AV, Pyankova ON, Santiago C, Sawczuk M, Scott RA, Uyba VV, Yvert T,
Perusse L, Ghosh S, Rauramaa R, North KN, Lucia A, Pitsiladis Y, and Bouchard C. No

370
371

Elite athletes as a model for research

evidence of a common DNA variant profile specific to world class endurance athletes. PLoS One
11: e0147330, 2016.
41. Reiner M, Niermann C, Jekauc D, and Woll A. Long-​term health benefits of physical activity –​a
systematic review of longitudinal studies. BMC Public Health 13: 813, 2013.
42. Ritti-​Dias RM, Avelar A, Salvador EP, and Cyrino ES. Influence of previous experience on resist-
ance training on reliability of one-​repetition maximum test. J Strength Cond Res 25: 1418–​1422, 2011.
43. Sakuma K, Aoi W, and Yamaguchi A. Current understanding of sarcopenia: possible candidates
modulating muscle mass. Pflugers Arch 467: 213–​229, 2015.
44. Saltin B, Henriksson J, Nygaard E, Andersen P, and Jansson E. Fiber types and meta-
bolic potentials of skeletal muscles in sedentary man and endurance runners. Ann N Y Acad Sci
301: 3–​29, 1977.
45. Samitz G, Egger M, and Zwahlen M. Domains of physical activity and all-​cause mortality: system-
atic review and dose-​response meta-​analysis of cohort studies. Int J Epidemiol 40: 1382–​1400, 2011.
46. Sarzynski MA, Loos RJ, Lucia A, Perusse L, Roth SM, Wolfarth B, Rankinen T, and
Bouchard C. Advances in exercise, fitness, and performance genomics in 2015. Med Sci Sports Exerc
48: 1906–​1916, 2016.
47. Schwellnus M, Soligard T, Alonso JM, Bahr R, Clarsen B, Dijkstra HP, Gabbett TJ, Gleeson
M, Hagglund M, Hutchinson MR, Janse Van Rensburg C, Meeusen R, Orchard JW, Pluim
BM, Raftery M, Budgett R, and Engebretsen L. How much is too much? (Part 2) International
Olympic Committee consensus statement on load in sport and risk of illness. Br J Sports Med
50: 1043–​1052, 2016.
48. Smart DJ, Hopkins WG, and Gill ND. Differences and changes in the physical characteristics of
professional and amateur rugby union players. J Strength Cond Res 27: 3033–​3044, 2013.
49. Storey A and Smith HK. Unique aspects of competitive weightlifting: performance, training and
physiology. Sports Med 42: 769–​790, 2012.
50. Swann C, Moran A, and Piggott D. Defining elite athletes: issues in the study of expert perform-
ance in sport psychology. Psychol Sport Exerc 16: 3–​14, 2015.
51. Thibault V, Guillaume M, Berthelot G, El Helou N, Schaal K, Quinquis L, Nassif H, Tafflet
M, Escolano S, Hermine O, and Toussaint J. Women and men in sport performance: the gender
gap has not evolved since 1983. J Sports Sci Med 9: 214–​223, 2010.
52. UK Sport. Talent transfer in full swing. www.uksport.gov.uk/​news/​2007/​02/​24/​talent-​transfer-​in-​
full-​swing. [Accesssed February 22, 2018.]
53. University LB. The Retired Athlete Health Research Project. www.leedsbeckett.ac.uk/​
ukrugbyhealth/​. [Accesssed February 23, 2018.]
54. Upadhyay J, Farr O, Perakakis N, Ghaly W, and Mantzoros C. Obesity as a Disease. Med Clin
North Am 102: 13–​33, 2018.
55. Webborn N, Williams A, McNamee M, Bouchard C, Pitsiladis Y, Ahmetov I, Ashley E,
Byrne N, Camporesi S, Collins M, Dijkstra P, Eynon N, Fuku N, Garton FC, Hoppe N,
Holm S, Kaye J, Klissouras V, Lucia A, Maase K, Moran C, North KN, Pigozzi F, and Wang
G. Direct-​ to-​
consumer genetic testing for predicting sports performance and talent identifica-
tion: consensus statement. Br J Sports Med 49: 1486–​1491, 2015.
56. Yang N, MacArthur DG, Gulbin JP, Hahn AG, Beggs AH, Easteal S, and North K. ACTN3
genotype is associated with human elite athletic performance. Am J Hum Genet 73: 627–​631, 2003.

371
372

26
TWIN AND FAMILY STUDIES
IN SPORT PERFORMANCE
José Maia and Peter T. Katzmarzyk

Introduction and aims

Descriptions of genealogies or the tracing of family lines across the history of mankind are
pervasive in many domains. Sometimes, these family histories are related to remarkable men or
women whose accomplishments are unparalleled. In a notable book about sport in the ancient
world, Kyle (2015) reports that Greek aristocrats “pushed the ideology of ascribed or perceived
arete, i.e., excellence or virtue, based on pedigree and family status, claiming that athletic excel-
lence came as a birthright in noble families with sporting traditions” (p. 82). For example,
Theogenes of Thasos, son of Timosthenes, was one of the most outstanding Greek athletes,
winning many victories in pankration and in boxing at the Olympian, Pythian, Nemean, and
Isthmian games. He was crowned in two Olympiads, in 480 and 476 BC, and impressed by his
amazing athletic accomplishments, there was a rumor that he wasn’t a mortal, but that he was
the son of a priest of Thasion Herakles and that the god impregnated his mother as narrated by
the Greek traveler and historian Pausanias.
The most famous family account in classical Greek and Roman mythology is about Hercules,
a crossbreed of god and man and a most outstanding hero. His family signature was unique, since
he was the son of Zeus, king of the Olympian Gods, and the mortal Alcmene.Through Zeus, he
was the grandson of Kronos and Rhea, and was also a great-​g randson of Uranus and Gaia. His
genetic material most certainly had a unique combination of his earthly mother’s mitochondrial
DNA to help him perform his incredible 12 labors. Alexander the Great, a mortal, also had an
exceptional family line (King Philip II of Macedon and Queen Olympias). Yet, his exposome
was unique as we know from history –​one of his tutors, Leonidas, taught him math, archery,
and horsemanship, whereas Aristotle, the famous Greek philosopher, taught him government,
politics, and the sciences.
Perhaps the first systematic inquiry into family lines of outstanding men and women was
that of Francis Galton –​Hereditary Genius –​published in 1869 where he tried to demonstrate
that genius and greatness runs in families, at least in well-​chosen and well-​bred British families.
He investigated 300 families comprising approximately 1000 eminent members, of whom 450
were illustrious. He presents two examples of sports family lines, one of oarsmen and another
of wrestlers. In his own words:

372
373

Twin and family studies in sport

I propose to supplement what I have written about brain by two short chapters on
muscle. No one doubts that muscle is hereditary in horses and dogs, but humankind
are so blind to facts and so governed by preconceptions, that I have heard it fre-
quently asserted that muscle is not hereditary in men. Oarsmen and wrestlers have
maintained that their heroes spring up capriciously, so I have thought it advisable to
make inquiries into the matter. The results I have obtained will beat down another
place of refuge for those who insist that each man is an independent creation, and
not a mere function, physically, morally, and intellectually, of ancestral qualities and
external influences.
(Galton, 1869, p. 305)

As far as we know, it was only in the Summer Olympics in Mexico (1968) and in Montreal
(1976) that genetic studies were carried out to identify putative genes associated with the
superior performance of “homo olimpicus” (Bouchard & Malina, 1984; de Garay et al., 1974).
Using the best available knowledge of the time, within the Mendelian tradition, about 17 single
gene systems (polymorphisms of antigens and enzymes of red cells) were investigated and data
obtained from homo olimpicus were compared to “normal” humans. These comparisons revealed
no relevant differences, even when separating athletes by sport. Of particular note is the book
prepared for the Montreal Olympic Games, a resumé of the finest available information of the
time concerning La préparation d’un champion: un essai à la preparation à la performance sportive
(Bouchard et al., 1973).Two chapters are of interest here: Qu’est qui fait un champion? (Bouchard
et al., 1973), and La contribution de l’hérédité au plan des determinants physique, physiologique et per-
ceptual de la performance sportive (Bouchard, 1973). The extent of today’s knowledge is vaster and
deeper (Ahmetov et al., 2016) when compared to that of sport geneticists in 1973. Nonetheless,
we are still struggling with the very same questions that exercised scientists of the last century
(Tucker & Collins, 2012), and some suggest that we may still be in the infancy of this extraor-
dinary quest (Pitsiladis et al., 2013). However, major efforts are currently under way, with great
expectations (Pitsiladis et al., 2016).
There is a saying, apparently reported by several sport scientists, that runs like this: if you
want to be a champion you had better choose your parents wisely. Of course no daughter/​son
chooses her/​his parents.Yet, its meaning is suggestive since belonging to a family of elite athletes
increases the odds of also being an elite athlete. Here are two examples separated by more than
two millennia: 1) Diagoras of Rhodes, descended from Damagetus, king of Ialysus, and, on
his mother’s side, from the Messenian hero, Aristomenes. From him, his family dynasty spans
through three generations of champions at Olympia, Isthmia and at Nemea games in Pankration
and boxing from 464 to 424 BCE (Kyle, 2015); his daughter, Kallipáteira, had a son Peisírrhodos
as well as a nephew, Euklēs, also winners in the Olympic Games. 2) Anastasia Liukin was born in
1989 in Moscow, and came to the US when she was 2 years old. Her parents,Valeri Liukin and
Anna Kotchneva, were Olympic gold medalists in artistic and rhythmic gymnastics, respectively.
Later, Anastasia was to become an exceptional gymnast. She won several world and Olympic
gymnastics medals in various apparatuses. During the Beijing 2008 Olympic Games she won
five medals, especially the highly cherished all-​round competition.
In a sense, this will be the motto for this chapter since it will be mostly devoted to summarize
available evidence from family lines of athletes. Its aims are to: 1) briefly review available twin
and nuclear family data on sport participation; 2) define, as precisely as possible, what is meant
by sport performance and/​or elite athleticism; 3) review available data on twins associated with
sport performance/​elite athleticism; 4) investigate family lines of outstanding athletes using his-
torical data as well as a dataset from elite athletes of Colombian origin; and 5) address the issue

373
374

J. Maia and P.T. Katzmarzyk

of intensive deliberate practice and its links to superior performance, together with the idea of
athletes’ exposome as well as genotype-​by-​exposome data on families of athletes.

Genetic factors and sports participation

When investigating the relevance of genetic factors in sport participation one finds a wide var-
iety of phenotype definitions (Barbosa et al., 2018; de Vilhena e Santos et al., 2012) that range
from very simple questions such as “Do you practice sports?”, to “How many hours per week
do you practice sport, in a noncompetitive setting, with your friends?”, or “Do you in your
leisure time practice any of the following sports …?” These phenotypes are then expressed dif-
ferently in quantitative terms, i.e., from binary data to sport indexes in arbitrary units of different
compositions.
As expected, the fraction of the total variance in sports practice that is due to genetic
differences among individuals, i.e., the heritability (h2), also varies substantially. Data on familial
resemblance in sport participation from family studies are sparse. Using sport participation
data from 2661 Portuguese families, we reported significant familial resemblance; however,
the spouse correlation (0.30) was higher than parent–​offspring correlations (0.18), and sibling
correlations (0.22), suggesting that genetic factors were not playing a major role (Maia et al.,
2014). On the other hand, in a review of available twin studies, the h2 estimates for sport par-
ticipation ranged from 0% to 85% (de Vilhena e Santos et al. 2012). With the exception of one
study in adolescents that reported h2 estimates of 0% in 13–​14-​and 15–​16-​year-​olds (Stubbe
et al., 2005), the h2 estimates ranged from 35% to 85% (de Vilhena e Santos et al. 2012). A study
of 304 Finnish adult twins reported an h2 estimate of 56% for sport participation, which falls
directly into this range (Mustelin et al., 2011).Thus, overall, there is good evidence of moderate
heritability for sport participation from twin studies.

Sport performance and elite athleticism

What exactly is meant by sport performance and elite athleticism? In formulating putative
answers to this question, the words of Aylsworth (2006) come to mind: “Morgan and his
students understood that accurate phenotype definition is an essential requirement for any sci-
entific inquiry into mechanisms of heredity” (p. 51).
Apart from the general idea that elite athleticism and/​or sport performance is a complex
phenotype governed by the additive and interactive effects of genes and the environment, seldom
do we find formal definitions. For example, Pérusse (2011) used family studies (HERITAGE
Family Study, the Quebéc Family Study, and the Canada Fitness Survey) to tackle the issue of
genetic factors in sport performance, but families in these studies were not elite sportsman/​
sportswomen; further, no formal definition of sport performance was provided. Guth and Roth
(2013), on the other hand, clearly distinguish physical performance from athletic performance,
saying that they are two different phenotypes, but apparently do not provide clear definitions.
Defining our phenotypes is clearly convoluted when considering individual sports or team
sports, since in the latter, technical, tactical, and strategic skills contribute to team perform-
ance. It has been further suggested that we should view sports teams as micro social systems
within dynamic systems/​complexity theory, with its manifold units of analysis –​the individual
player, small sets of players, or the team a whole (Araujo & Davids, 2016). For our purposes
we will use suggestions from Druzhevskaya et al. (2008) that classifies levels of athletes: sport
performance is that achieved by highly elite (winners of world championships, world cups,
or Olympic Games), elite (bronze and silver medalists in world championships, word cups,

374
375

Twin and family studies in sport

European championships, or Olympic Games) and subelite athletes (qualifiers or participants

in international championships of world class). Further, by elite athleticism we mean the out-
standing display of athletes’ attributes like strength, speed, flexibility, coordination, endurance,
and agility, for example, as seen in gymnasts, high jumpers, runners, basketball players, javelin
throwers, figure skaters, soccer, and team handball payers.
Our general definition of elite athleticism is not to be confused with, for example, the
extreme VO2max cut point (≥75 ml/​kg; range: 75.0–​92.9 ml/​kg) used to define cases in the
GENATHLETE study (Wolfarth et al., 2008). In the GENATHLETE study, these cases were
compared to sedentary controls to investigate several single nucleotide polymorphisms (SNPs) in
candidate genes in relation to this superior phenotype.To date, the results from GENATHLETE
have not provided strong evidence for differences in allele and genotype frequencies between
the elite endurance athletes and the sedentary controls (Rankinen et al., 2000; Rivera et al.,
1997; Wolfarth et al., 2000, 2007, 2008). Further research identified a promising panel of 45
markers in the GAMES consortium (GENATHLETE and Japanese endurance runners), and
tested them for replication in seven additional cohorts comprising 1520 endurance athletes
and 2760 controls (Rankinen et al., 2016). Unfortunately, the results did not identify a panel
of genomic variants common to the elite athlete groups, and the main conclusion was that
common genetic variants do not appear to be strong determinants of elite endurance athlete
status (Rankinen et al., 2016).

Twin studies
Olympic and international-​level twin athletes fascinate sport scientists for they also have the
potential to help unravel the influence of their similar genetic endowments and common
home environments on how they arrived at the summit of their sports’ careers. Here are some
examples: José and Ozzie Canseco in baseball, Horace and Harvey Grant in basketball, triplets
Leila, Liina, and Lily Luik in marathon running, Katherine and Michelle Plouffe in basketball,
Janne and Lieke Wevers in gymnastics, and Frank and Ronald de Boer in soccer.
Although debatable (Rende et al., 1990), Francis Galton was perhaps the first author to
suggest, in a landmark paper written in 1875 (Galton, 1875), using the twin method to address
the complex issue of disentangling the nature and nurture controversy in the expression of
human traits. Although old, and challenging, this issue is still present and pressing in behavioral
genetics (Plomin & Asbury, 2005) as well as in the sport sciences (Davids & Baker, 2007).
Most probably the first authors to use the twin method to address the issue of genetics in
the performance of homo sportivus were Grebe (1955, 1956, 1957) in Germany, and Gedda
(1955, 1960) in Italy.This last author presented in his 1960 paper a summary of elite Italian twin
athletes. Further, he coined the words family sportivation, as well as isosportivation (family members
practicing the same sport) and allosportivation (family members practicing different sports). The
main conclusion of this study was: “From the theoretical point of view we may assume that the
specific sports genotype, being composed of normal character factors, is transmitted as a dom-
inant. This is what I have been maintaining for some years, and Grebe holds the same opinion”
(p. 400).Yet, a review (Klissouras, 2001) on the nature and nurture of human performance did
not report on high-​elite or elite twin athletes’ putative dominant gene(s) vertically transmitted,
nor has the TERRY (Casini et al., 2002) or the RITA (Parisi et al., 2001) Italian twin athletes’
registers provided information about candidate gene(s) linked to elite sports performance.
In 2013, David Epstein (2013) wrote an amazing book with a thought-​provoking title: The
Sports Gene. One could naively think, along the same lines as in Grebe and Gedda, that some
major gene might be the Holy Grail behind the extraordinary performance of elite athletes.

375
376

J. Maia and P.T. Katzmarzyk

Exactly on the same lines as in the critique of Davids and Baker’s (2007) “single gene as a magic
bullet.” This is probably what happened with the angiotensin-​converting enzyme (ACE) gene
and the alpha-​actinin-​3 (ACTN3) gene and their implications for sports performance, and they
showed to be somewhat valid. For example, a meta-​analysis and systematic review (Ma et al.,
2013) concluded that there was solid evidence for the associations of the ACE II genotype and
endurance events as well as for ACTN3 R allele and power events.These conclusions have been
supported by Papadimitriou et al. (2016) in elite sprinters from ten countries, although their
conclusions were cautionary: “Despite sprint performance relying on many gene variants and
environment, the % sprint time variance explained by ACE and ACTN3 is substantial at the
elite level and might be the difference between a world record and only making the final” (p. 1).
Note that the amount of variance explained by ACTN3 and ACE genotypes was approximately
1–​1.5% which could be of paramount importance in winning a gold medal. Yet, it was also
stated that a substantial amount of performance variance was not explained and further research
with common and rare variants was still needed. Although Santiago et al. (2008) found a higher
frequency of the ACTN3 RR genotype in Spanish soccer players than in controls,Tringali et al.
(2014) were not able to replicate these results in high-​level Italian gymnasts even for the ACE
DD genotype. As a concluding note, we concur with Davids and Baker’s (2007) “single gene
as a magic bullet” in the sense that soccer players’ and gymnasts’ high-​level performance is far
more complex than meets the eye with single candidate genes, or that a potential genetic pro-
file of relevant polymorphisms to influence muscular strength that exclusively favors high-​level
performance is extremely rare to find (Hughes et al., 2011).

Family studies
Some geneticists are apparently favored by the “fire of inspiration” and write remarkable
accounts of their daunting quest. Three of these, all New York Times book award winners, were
written to illuminate the most astonishing accomplishments within the large field of genetics.
They all wrote about information, and the gene is but coded information, its copying, tran-
scribing, translation, and also passing it from one generation to the next. The first book was
written by Watson and Andrew (2003), narrating the remarkable history of the discovery of the
DNA structure.The title is very provocative: DNA: The Secret of Life. Also, Ridley (2000) wrote
an outstanding account of the genome in 23 chapters. The last in this trilogy was carved by
Mukherjee (2016). It is a monumental endeavor concerning the gene and its intimate history.
The firing seed of this work erupts from a serious health problem affecting some of his family
members, mostly his father’s relatives, and he asks poignant questions such as: Is the disease gen-
etic? Why are some of my family members spared from it? What triggers its emergence in some
relatives, i.e., what unveiled their predispositions? Who were carriers of the “defected gene(s)?”
Am I a carrier? Will these genes be passed on to my progeny? How will we manage the volley
between the hands of nature and nurture?
Although framed in a completely different context, sports excellence, also conceived as a
marker of a complex “dance” of genes (i.e., additive and epistatic effects), has been historically
hunted within family lines, trying to answer some of Muhkerjee’s questions and rooted inside
the tools of medical genetics, namely pedigree analysis. Perhaps the first serious attempts were
made by Grebe (1956) and Moser (1960) who examined a series of genealogical diagrams of
sportsmen and suggested that sports talent as well as sports abilities were dominant traits within
family lines governed by an unknown number of additive gene effects, as well as an intricate
set of environmental conditions that mingle with genes to produce outstanding displays of
sports performance. Although they suggested a dominant mode of transmission in these reports,

376
377

Twin and family studies in sport

Figure 26.1 A three-​generation pedigree of a successful Czech sports family.

one has to bear in mind that no formal segregation analyses were done because at the time no
sophisticated statistical techniques, algorithms, and software were available, as we have today
(see Sun 2002), to deal with the complexities of data from pedigrees of arbitrary size as well as
considering putative environmental covariates, i.e., important events in athletes’ sports life his-
tories that in all likelihood push them towards the road of excellence. In any case, Figure 26.1
is an example provided by Kovář (1980) of a pedigree of a high successful sports family mainly
with a gymnastics tone.
In 2010, one of the authors (JM), in conjunction with one of his Colombian students
(Luis Carlos Muñoz), started collecting information about Colombian families of athletes
from individual sports: track-​and-​field, weightlifting, cycling, wrestling, skating, karate, judo,
fencing, swimming, taekwondo, boxing, and diving. Our search started with a reference
athlete, i.e., a proband, who participated and/​or won a medal in at least one international
competition (i.e., an elite athlete). From an initial survey of 103 putative families, we only
had access to 78 which included 1577 relatives of whom 149 were high elite, elite, and
subelite. In order to better understand their family structure, all were divided according to
three conditions: 1) families with athletes whose parents, siblings, and progeny also were
athletes (N=7); 2) families with only siblings as athletes (N=30), and 3) families with no
parents and siblings as athletes (N=41). Yet, before we start presenting aspects of our data it
is necessary to inform the reader of the following: 1) today we have several techniques avail-
able that address the complex issue of familial aggregation (Keen & Elston, 2003; Mathew
et al., 2011; Naj et al., 2002) and we will use them to analyze some of the data; 2) although
sophisticated segregation methods of analysis to identify modes of inheritance in families of
arbitrary size are available (Sun, 2002), none has been used to study pedigrees of athletes so
far as we know, and we also will not use them; 3) finally, there are also techniques to model
the genotype–​phenotype relationships within pedigrees of arbitrary size (Elston et al., 1992),
but because we do not have any genetic information, no such models will be used. Even if
we did, using a genome-​wide association study approach, or family studies for the optimal
detection of rare variants, we would need thousands of subjects, or large-​size extended
pedigrees (Diego et al., 2015), which are far beyond the scope of this study. We also ask the

377
378

J. Maia and P.T. Katzmarzyk

Table 26.1 Parameter estimates from the Colombian sports families

Parameters Estimates Standard error P-​value

Heritability 0.57 0.06 <0.001

Residual family correlations
Full sibs 0.59 0.03 <0.001
Half-​sibs 0.14 0.01 <0.001
Parent–​offspring 0.29 0.03 <0.001
Environmental correlations
Full sibs 0.72 0.07 <0.001

Note: estimates adjusted for sex and race.

Table 26.2 Information summary of the seven most outstanding Colombian sports families

Family Level # family # (%) of Sports Achievement Age at onset of

name members in athletes practiced level sports practice
pedigree
≠ ~ ≠ ~ ≠ ~

Tovar Ante World class 16 7 (44%) x x x

Howard World class 15 2 (13%) x x x
Urrea World class 21 4 (19%) x x x
Lozano Pan American 13 2 (15%) x x x
Garcés Pan American 16 5 (31%) x x x
Lopez
Osorio Pan American 21 9 (43%) x x x
Trujillo Pan American 13 5 (39%) x x x

Note: # number; ≠ different; ~ similar.

reader to bear in mind the fact that sport scientists recognize sports performance as a highly
complex and multifactorial human trait most often expressed in various quantitative and
qualitative ways.Yet, for matters of convenience mainly when investigating a candidate gene
marking some aspect of this complexity, sports performance is often dichotomized using
some empirical threshold. Consistent with this, as well as with the classification suggested
by Druzhevskaya et al. (2008), we also treat sports performance as a discrete trait with three
categories (subelite, elite, and high elite) as mentioned above.
Our approach to the Colombian families of athletes will be mainly descriptive. To start
with, Table 26.1 contains information regarding the analysis of familial aggregation using
a model (Elston & Stewart, 1971) and an algorithm (Elston et al., 1992) implemented in
S.A.G.E. genetic epidemiology software (S.A.G.E 64, 2016). Sports performance shows a
substantial heritability (h2=0.57), and the residual familial correlations as well as the envir-
onmental correlations also reveal the importance of siblings’ other shared factors beyond
shared genes.
Further, a summary of the data from the available families in condition 1 are in Table 26.2.
The number of athletes, within the three-​generation pedigrees, belonging to the three sports
performance categories varies from 2 (13%) to 7 (44%), and their chosen sport is most often

378
379

Twin and family studies in sport

Figure 26.2 Pedigrees of Garcés Lopez (above) and Trujillo (below) sports families. Filled symbols
represent elite and subelite athletes.

equal. Additionally, their achievement level varies across the three levels, and the age at onset of
sports practice is also different.
In Figure 26.2 we show pedigrees of two representative families. The first one (Garcés
Lopez) emphasizes their cycling and cyclo-​ cross aggregation, whereas the second one
(Trujillo) mostly represents a passion for athletics, namely the dizygotic twin sisters’ elite and
subelite race walking history.
Further, from a pool of 30 three-​generation families with only elite and/​or subelite siblings’
athletes in the third generation, we chose two –​the Wilches Tumbia clustering in cycling
(above) and the Botero Coy (below) in skating (Figure 26.3).
From families in condition 3, i.e., families with sports “silence” in generations I and II as well
as in siblings, we chose two with unique signatures in their outstanding sports performance –​
Cecília Margarita Baena and Sandra Viviana Roa. Cecília Baena was world champion of in-​line
skating already at 13 years of age. At 16 years of age she won the 500-​meter and 1000-​meter
gold medals in the 2001 world championship. In her career she won six world titles, and several
other first class titles in Central America and Pan American championships.

379
380

J. Maia and P.T. Katzmarzyk

Figure 26.3 Pedigrees of elite and/​or subelite Colombian siblings’ sports families.

Sandra Viviana Roa is an exceptional wrestler having to her credit bronze and gold medals
in several Pan American and Central America championships. At the world level she was several
times in the top ten of her weight category.
To finish this very brief description on Colombian sports families, we chose two of them as
our case studies –​the Tovar Ante and the Urrutia families because they are unique examples.
The Tovar Ante family line has a unique signature given that they always have been involved in
shooting, pistol class, with the exception of Rafael Tovar Ante whose sports history is rooted
in motocross. The family history in pistol class started in the first generation with Bernardo
Tovar Paredes who was national champion in Colombia. Then, he transmitted, vertically, his
sports passion and his genes to his sons, especially to Bernardo Tovar Ante, who was an out-
standing athlete participating in Olympic Games as well as world championships. He is the family
member who conquered more medals. He then “contaminated” his wife with his passion since
environmental driven conditions were always in place that in all likelihood unveiled her innate
predispositions. She was also an elite athlete, and the same applies to their son and daughter, also a
unique combination of their genetic endowments and a highly favorable exposome (Figure 28.6).
The Urrutia family is an extreme example of “silence” in sports participation until the fourth
generation (Figure 26.7). The most outstanding athlete is Maria Isabel Urrutia. She started in

380
381

Figure 26.4 Family pedigree of Cecília Margarita Baena.

Figure 26.5 Family pedigree of Sandra Viviana Roa.

382

J. Maia and P.T. Katzmarzyk

Figure 26.6 Pedigree of the Tovar Ante family.

Figure 26.7 Pedigree of the Urrutia family.

track and field, and in 1978 was the youngest national champion of her age category in 100-​
meter sprint –​she was 13 years. She was also fond of discus and shot put, and won several
national and international championships. Later she participated in the Seoul Olympics and
reached 14th place in the overall ranking in both events. Then, at age 24 years, she also started
participating in weightlifting competitions, winning several championships, and a gold medal in
the Sidney 2000 Olympics (weight category <75 kg). Her brother Robinson is also an excellent
athlete in the 100-​meter sprint.

382
383

Twin and family studies in sport

We chose to close this section with three recent stories of successful sports family lines. The
first one is about Coco Vandeweghe, a top-​class tennis player (ranked 17 by the Women’s Tennis
Association), and the unique signature of her family. She is the daughter of Tauna Vandeweghe,
a US Olympic swimmer; her uncle and grandfather were elite US basketball players and her
grandmother was Miss America in 1952. The other two, with apparently similar signatures, are
about US athletes participating in the 2018 Peyong Chang Winter Olympics. Nancy Swider-​
Peltz Jr. participated in the Games with her brother, Jeffrey Swider-​Peltz Jr. Their mother is
Nancy Swider-​Peltz, a four-​time US Olympic speed skater who was the first US athlete to
compete in four Olympic Winter Games. Cochran-​Siegle was also in the 2018 Olympic Games.
His mother, Barbara Ann Cochran, won the Olympic gold in the slalom in 1972. His family line
also include skiing Olympians, like his aunts Lindy and Marilyn Cochran, uncle Bob Cochran,
and cousin Jimmy Cochran.

Deliberate practice, athletes’ exposome, and genetic-​by-​exposome

interactions
As far as we know, there is apparently no “long-​range” prospective study that investigates the life
course to excellence of any high-​elite, or elite athletes starting from their youth sport careers, or
even as far back as preconception via gametic imprinting (Brutsaert & Parra, 2009). The reason
is perceptibly simple –​we cannot know where such talented youth are because talent search
and identification enterprises have no clear-​cut definitions of what “talent” is and we can hardly
find an expeditious way of uniquely identifying putative candidates (Johnston et al., 2018; Maia,
1993; Sarmento et al., 2018). We simply do not exactly know what we are looking for because
of the complexities of individual development, heterogeneity in responses to sports training and
competition, as well as how the same individual will display their manifold attributes across her/​
his development. Although Guth and Roth (2013) stated that “the drive behind the prediction
of athletic performance with genetics is primarily aimed at the early identification of individ-
uals who will become exceptional athletes as adults” (p. 5), to date there is no record about
such findings by sports geneticists. Further, the person who first spotted Cristiano Ronaldo, the
world’s most famous soccer player, did not know about the uniqueness of the pearl he had in his
hands. He was a humble soccer coach working with children, but his gaze apparently ranged far.
It is a truism to say that elite athletes are born and then made. Bloom (1985) was probably
one of the first that described, in retrospective terms, a few invariant features shared by elite
athletes (tennis players and swimmers) during their roads towards outstanding performance: 1)
strong interest and emotional involvement in their chosen sport; 2) willingness to devote great
amounts of time and efforts towards excellence; 3) great competitiveness with their peers, and
firm determination to shine in their sport; and 4) high learning capacity and potential to reach
distinctiveness.
Papers by Ericsson et al. (1993, 2009) on intense deliberate practice, where the golden
number of 10,000 accumulated hours of training was coined, and its links to exceptional
achievements launched a serious debate in the sport science field. Their views were strongly
criticized, for example, by Tucker and Collins (2012) namely for inconsistency in their original
data, reference model, as well as their insistent call for the nurture aspect (deliberate practice)
and removal of the wealthy body of data on genetics of performance in the nature and nurture
equation. Additionally, systemic views linking complex synergies of genes and the manifold
expressions of family, peers, and training environments to best understand the march towards
excellence of any young athlete are now favored instead of the old motto nature versus nurture
(see, for example, Davids and Baker (2007); Brutsaert and Parra (2009), or Wai (2014); Plomin

383
384

J. Maia and P.T. Katzmarzyk

et al. (2014)). Yet, to date, there is apparently no longitudinal data linking all the exposome
conditions putative talented youth went through interacting with varied sets of genes to express
outstanding performance at various stages of their careers. These chronometric time effects
of the marriage of young athletes’ exposome and genome are simply not known. Even in the
exposome, the majority of the available data are retrospective and open to debate (Coutinho
et al., 2016). Recently, Rees et al. (2016) reviewed the literature on development of the world’s
best sporting talent centered on four domains: 1) the performer (birthdate, genetics, anthropo-
metric and physiological factors, psychological skills and motivational orientations, person-
ality traits), 2) the environment (birthplace, support from parents, siblings and coaches, athlete
support programs), 3) practice and training (volume of sport-​specific practice and training, early
specialization versus sampling and play), and 4) other potential factors, and concluded that “in
encouraging researchers, we would point to the relative dearth of prospective and multidiscip-
linary studies that could offer insight regarding the complex interactions that almost certainly
exist across domains” (p. 1050).

Conclusion
The world-​renowned biologist, E.O.Wilson, wrote the following in his book On Human Nature
(1978) that won the Pulitzer Prize in 1979:

Since each individual produced by the sexual process contains a unique set of genes,
very exceptional combinations of genes are unlikely to appear twice within the same
family. So if genius is to any extent hereditary, it winks on and off through the gene
pool in a way that would be difficult to measure or predict. Like Sisyphus rolling his
boulder up to the top of the hill only to have it tumble down again, the human gene
pool creates hereditary genius in many ways in many places only to have it come apart
in the next generation.

Marion Jones and Tim Montgomery, world-​renowned track and field athletes had a son. The
same happened, for example, to world-​ranked tennis players Steffi Graf and Andre Agassi (one
son and one daughter), as well as Roger and Mirka Federer (two twin pairs, one of girls and
one of boys). Will these children be future champions? Apparently, all odds favor them not only
in terms of sharing of genetic material, but also in their highly advantageous environmental
conditions, i.e., their exposome. Time will tell.

References
Ahmetov, I. I., Egorova, E. S., Gabdrakhmanova, L. J., & Fedotovskaya, O. N. (2016). Genes and athletic
performance: an update. Med Sport Sci, 61, 41–​54. doi: 10.1159/​000445240
Araujo, D. & Davids, K. (2016). Team synergies in sport: theory and measures. Front Psychol, 7, 1449.
doi: 10.3389/​fpsyg.2016.01449
Aylsworth, A. S. (2006). Defining disease phenotypes. In J. L. Haimes & M. A. Pericak-​Vance (Eds.), Genetic
Analysis of Complex Disease. (2nd ed., pp. 51–​90). Hoboken, NJ: John Wiley & Sons, Inc.
Barbosa, J., Gomes,T. N., Forjaz, C. L. M., & Maia, J. (2018). Correlational analysis of Physical activity levels
and sport participation. Research in twins and in families. RPEFE, 31, 505.
Bloom, B. S. (1985). Developing Talent in Young People. New York: Ballantine Publishing Books.
Bouchard, C. (1973). Qu’est-​ce qui fait un champion? In C. Bouchard, J. Brunelle & P. Godbout (Eds.), La
preparation d’un Champion. Un essai à la performance sportive. (pp. 75–​106). Québec: Éditions du Pélican.
Bouchard, C., Brunelle, J., & Godbout, P. (1973). La preparation d’un Champion. Un essai à la performance
sportive. Québec: Éditions du Pélican.

384
385

Twin and family studies in sport

Bouchard, C. & Malina, R. M. (1984). Genetics of Olympic athletes: a discussion of methods and issues.
In J. E. L. Carter (Ed.), Physical Structure of Olympic athletes. Part II Kinanthropometry of Olympic athletes
(pp. 28–​38). Basel: Karger.
Brutsaert,T. D. & Parra, E. J. (2009). Nature versus nurture in determining athletic ability. Med Sport Sci, 54,
11–​27. doi: 10.1159/​000235694
Casini, B., Pittaluga, M., & Parisi, P. (2002). Two Italian twin registers for research in human biology and
sport science. Twin Res, 5(5), 376–​381. doi: 10.1375/​136905202320906129
Coutinho, P., Mesquita, I., & Fonseca, A. M. (2016). Talent development in sport: a critical review of
pathways to expert performance. Int J Sports Sci Coach, 11(2), 279–​293.
Davids, K. & Baker, J. (2007). Genes, environment and sport performance: why the nature-​nurture dualism
is no longer relevant. Sports Med, 37(11), 961–​980.
de Garay, A. L., Levine, L., & Carter, J. E. L. (1974). Genetic and Anthropological Studies of Olympic Athletes.
New York: Academic Press.
de Vilhena e Santos, D. M., Katzmarzyk, P.T., Seabra, A. F., & Maia, J. A. (2012). Genetics of physical activity
and physical inactivity in humans. Behav Genet, 42(4), 559–​578. doi: 10.1007/​s10519-​012-​9534-​1
Diego, V. P., Kent, J. W., & Blangero, J. (2015). Familial studies: genetic inferences. In J. D. Wright (Ed.),
International Encyclopedia of the Social & Behavioral Sciences (2nd ed.) (pp. 715–​724). Oxford: Elsevier.
Druzhevskaya, A. M., Ahmetov, I. I., Astratenkova, I.V., & Rogozkin,V. A. (2008). Association of the ACTN3
R577X polymorphism with power athlete status in Russians. Eur J Appl Physiol, 103(6), 631–​634.
doi: 10.1007/​s00421-​008-​0763-​1
Elston, R. C., George,V.T., & Severtson, F. (1992).The Elston-​Stewart algorithm for continuous genotypes
and environmental factors. Hum Hered, 42(1), 16–​27. doi: 10.1159/​000154043
Elston, R. C. & Stewart, J. (1971). A general model for the genetic analysis of pedigree data. Hum Hered,
21(6), 523–​542. doi: 10.1159/​000152448
Epstein, D. (2013). The Sports Gene: Inside the Science of Extraordinary Athletic Performance.
New York: Penguim Book.
Ericsson, K. A., Krampe, R.T., & Tesch-​Römer, C. (1993).The role of deliberate practice in the acquisition
of expert performance. Psychol Rev, 100(3), 363–​406.
Ericsson, K. A., Nandagopal, K., & Roring, R. W. (2009). Toward a science of exceptional
achievement: attaining superior performance through deliberate practice. Ann N Y Acad Sci, 1172,
199–​217. doi: 10.1196/​annals.1393.001
Galton, F. (1869). Hereditary Genius: An Inquiry into its Laws and Consequences. London: Macmillan.
Galton, F. (1875). The history of twins, as a criterion of the relative powers of nature and nurture. Fraser’s
Magazine, 566–​576.
Gedda, L. (1955). La valutazione genética dell’atleta. Acta Genet Med Gemellol (Roma), 4, 249–​260.
Gedda, L. (1960). Sport and genetics. A study on twins (312) pairs. Acta Genet Med Gemellol (Roma), 9,
387–​405.
Grebe, H. (1955). Sport bei zwillingen. Acta Genet Med Gemellol (Roma), 4, 275–​295.
Grebe, H. (1956). Sportfamilien. Acta Genet Med Gemellol (Roma), 5, 318–​326.
Grebe, H. (1957). Genotypus und konstitution im sport. Sportmedizin, 10, 269–​274.
Guth, L. M. & Roth, S. M. (2013). Genetic influence on athletic performance. Curr Opin Pediatr, 25(6),
653–​658. doi: 10.1097/​MOP.0b013e3283659087
Hughes, D. C., Day, S. H., Ahmetov, I. I., & Williams, A. G. (2011). Genetics of muscle strength and
power: polygenic profile similarity limits skeletal muscle performance. J Sports Sci, 29(13), 1425–​1434.
doi: 10.1080/​02640414.2011.597773
Johnston, K., Wattie, N., Schorer, J., & Baker, J. (2018). Talent identification in sport: a systematic review.
Sports Med, 48(1), 97–​109. doi: 10.1007/​s40279-​017-​0803-​2
Keen, K. J. & Elston, R. C. (2003). Robust asymptotic sampling theory for correlations in pedigrees. Stat
Med, 22(20), 3229–​3247. doi: 10.1002/​sim.1559
Klissouras, C. (2001).The nature and nurture of human performance. Eur J Sport Sci 1(1), 1–​10. doi: 10.1080/​
17461390100071207
Kovář, R. (1980). Human Variation in Motor Abilities and its Genetic Analysis. Prague: Charles University Prague.
Kyle, D. G. (2015). Sport and Spectacle in the Ancient World. (2nd ed.). Chichester, West Sussex: Wiley
Blackwell.
Ma, F., Yang, Y., Li, X., Zhou, F., Gao, C., Li, M., & Gao, L. (2013). The association of sport performance
with ACE and ACTN3 genetic polymorphisms: a systematic review and meta-​analysis. PLoS One,
8(1), e54685. doi: 10.1371/​journal.pone.0054685

385
386

J. Maia and P.T. Katzmarzyk

Maia, J., Gomes, T. N., Tregouet, D. A., & Katzmarzyk, P. T. (2014). Familial resemblance of physical activity
levels in the Portuguese population. J Sci Med Sport, 17(4), 381–​386. doi: 10.1016/​j.jsams.2013.09.004
Maia, J. A. R. (1993). Anthropobiological Approach to Selection in Sport. A Multivariate Study of Biosocial Selection
Markers in Team Handball Players Aged 13 to 16 Years. (PhD), Faculty of Sport Sciences and Physical
Education, Porto, Portugal.
Mathew, G., Song,Y., & Elston, R. (2011). Interval estimation of familial correlations from pedigrees. Stat
Appl Genet Mol Biol, 10, Article 11.
Moser, H. (1960). Uber die vererbung der sportlichen fähigkeiten. (1st ed.). München.
Mukherjee, S. (2016). The Gene: An Intimate History. New York: Scribner.
Mustelin, L., Latvala, A., Pietilainen, K. H., Piirila, P., Sovijarvi, A. R., Kujala, U. M., Rissanen, A., & Kaprio,
J. (2011). Associations between sports participation, cardiorespiratory fitness, and adiposity in young
adult twins. J Appl Physiol (1985), 110(3), 681–​686. doi: 10.1152/​japplphysiol.00753.2010
Naj,A. C., Park,Y. S., & Beaty,T. H. (2002). Detecting familial aggregation. In R. C. Elston, J. M. Satagopan & S.
Sun (Eds.), Statistical Human Genetics: Methods and Protocols. (pp. 119–​150). New York: Humana Press.
Papadimitriou, I. D., Lucia, A., Pitsiladis, Y. P., Pushkarev, V. P., Dyatlov, D. A., Orekhov, E. F., Artioli, G.
G., Guilherme, J. P., Lancha, A. H., Jr., Gineviciene, V., Cieszczyk, P., Maciejewska-​Karlowska, A.,
Sawczuk, M., Muniesa, C. A., Kouvatsi, A., Massidda, M., Calo, C. M., Garton, F., Houweling, P. J.,
Wang, G., Austin, K., Druzhevskaya, A. M., Astratenkova, I. V., Ahmetov, I. I., Bishop, D. J., North, K.
N., & Eynon, N. (2016). ACTN3 R577X and ACE I/​D gene variants influence performance in elite
sprinters: a multi-​cohort study. BMC Genomics, 17, 285. doi: 10.1186/​s12864-​016-​2462-​3
Parisi, P., Casini, B., Di Salvo,V., Pigozzi, F., Pittaluga, M., Prinzi, G., & Klissouras,V. (2001). The Registry
of Italian Twin Athletes (RITA): Background, design, and procedures, and twin data analysis on
sport participation—​an application to twin swimmers. Eur J Sport Sci, 1(2), 1–​12. doi: 10.1080/​
17461390100071208
Pérusse, L. (2011). Role of genetics factors in sport performance: evidence from family studies. In C.
Bouchard and E. Hoffman (Eds.) Genetic and Molecular Aspects of Sport Performance (pp. 90–​100).
Chichester, West Sussex: Wiley-​Blackwell.
Pitsiladis, Y., Wang, G., Wolfarth, B., Scott, R., Fuku, N., Mikami, E., He, Z., Fiuza-​Luces, C., Eynon, N.,
& Lucia, A. (2013). Genomics of elite sporting performance: what little we know and necessary
advances. Br J Sports Med, 47(9), 550–​555.
Pitsiladis, Y. P., Tanaka, M., Eynon, N., Bouchard, C., North, K. N., Williams, A. G., Collins, M., Moran,
C. N., Britton, S. L., Fuku, N., Ashley, E. A., Klissouras, V., Lucia, A., Ahmetov, I. I., de Geus, E.,
& Alsayrafi, M. (2016). Athlome Project Consortium: a concerted effort to discover genomic and
other “omic” markers of athletic performance. Physiol Genomics, 48(3), 183–​190. doi: 10.1152/​
physiolgenomics.00105.2015
Plomin, R. & Asbury, K. (2005). Nature and nurture: genetic and environmental influences on behavior.
Ann Am Acad Pol Soc Sci, 600, 86–​98.
Plomin, R., Shakeshaft, N. G., McMillan, A., & Trzaskowski, M. (2014). Nature, nurture, and expertise.
Intelligence, 45, 46–​59. doi: 10.1016/​j.intell.2013.06.008
Rankinen, T., Fuku, N., Wolfarth, B., Wang, G., Sarzynski, M. A., Alexeev, D. G., Ahmetov, I. I., Boulay,
M. R., Cieszczyk, P., Eynon, N., Filipenko, M. L., Garton, F. C., Generozov, E. V., Govorun, V. M.,
Houweling, P. J., Kawahara,T., Kostryukova, E. S., Kulemin, N.A., Larin,A. K., Maciejewska-​Karlowska,
A., Miyachi, M., Muniesa, C. A., Murakami, H., Ospanova, E. A., Padmanabhan, S., Pavlenko, A. V.,
Pyankova, O. N., Santiago, C., Sawczuk, M., Scott, R. A., Uyba, V. V., Yvert, T., Perusse, L., Ghosh,
S., Rauramaa, R., North, K. N., Lucia, A., Pitsiladis, Y., & Bouchard, C. (2016). No evidence of a
common DNA variant profile specific to world class endurance athletes. PLoS One, 11(1), e0147330.
doi: 10.1371/​journal.pone.0147330
Rankinen, T., Wolfarth, B., Simoneau, J. A., Maier-​Lenz, D., Rauramaa, R., Rivera, M. A., Boulay, M. R.,
Chagnon,Y. C., Perusse, L., Keul, J., & Bouchard, C. (2000). No association between the angiotensin-​
converting enzyme ID polymorphism and elite endurance athlete status. J Appl Physiol (1985), 88(5),
1571–​1575. doi: 10.1152/​jappl.2000.88.5.1571
Rees, T., Hardy, L., Gullich, A., Abernethy, B., Cote, J., Woodman, T., Montgomery, H., Laing, S., & Warr,
C. (2016). The Great British Medalists Project: a review of current knowledge on the development
of the world’s best sporting talent. Sports Med, 46(8), 1041–​1058. doi: 10.1007/​s40279-​016-​0476-​2
Rende, R. D., Plomin, R., & Vandenberg, S. G. (1990). Who discovered the twin method? Behav Genet,
20(2), 277–​285.
Ridley, M. (2000). Genome: The autobiography of a species in 23 chapters. New York: Perennial.

386
387

Twin and family studies in sport

Rivera, M. A., Dionne, F.T.,Wolfarth, B., Chagnon, M., Simoneau, J. A., Perusse, L., Boulay, M. R., Gagnon,
J., Song, T. M., Keul, J., & Bouchard, C. (1997). Muscle-​specific creatine kinase gene polymorphisms
in elite endurance athletes and sedentary controls. Med Sci Sports Exerc, 29(11), 1444–​1447.
S.A.G.E 64. (2016). Statistical analysis for genetic epidemiology. http://​darwin.cwru.edu/​sage/​.
Santiago, C., Gonzalez-​Freire, M., Serratosa, L., Morate, F. J., Meyer, T., Gomez-​Gallego, F., & Lucia, A.
(2008). ACTN3 genotype in professional soccer players. Br J Sports Med, 42(1), 71–​73. doi: 10.1136/​
bjsm.2007.039172
Sarmento, H., Anguera, M. T., Pereira, A., & Araujo, D. (2018). Talent identification and development in
male football: a systematic review. Sports Med, 48(4), 907–​931. doi: 10.1007/​s40279-​017-​0851-​7
Stubbe, J. H., Boomsma, D. I., & De Geus, E. J. (2005). Sports participation during adolescence: a shift from
environmental to genetic factors. Med Sci Sports Exerc, 37(4), 563–​570.
Sun, X. (2002). Segregation analysis using a unified model. In R. C. Elston, J. M. Satagopan & S. Sun (Eds.),
Statistical Human Genetics: Methods and Protocols (pp. 211–​236). New York: Humana Press.
Tringali, C., Brivio, I., Stucchi, B., Silvestri, I., Scurati, R., Michielon, G., Alberti, G., & Venerando, B.
(2014). Prevalence of a characteristic gene profile in high-​level rhythmic gymnasts. J Sports Sci, 32(14),
1409–​1415. doi: 10.1080/​02640414.2014.893371
Tucker, R. & Collins, M. (2012). What makes champions? A review of the relative contribution of genes
and training to sporting success. Br J Sports Med, 46(8), 555–​561. doi: 10.1136/​bjsports-​2011–​090548
Wai, J. (2014). Experts are born, then made: Combining prospective and retrospective longitudinal data
shows that cognitive ability matters. Intelligence, 45, 74–​80. doi: 10.1016/​j.intell.2013.08.009
Watson, J. D. & Andrew, B. (2003). DNA: The Secret of Life. New York: Alfred A. Knopf.
Wilson, E. O. (1978). On Human Nature. Cambridge, MA: Harvard University Press.
Wolfarth, B., Rankinen,T., Muhlbauer, S., Ducke, M., Rauramaa, R., Boulay, M. R., Perusse, L., & Bouchard,
C. (2008). Endothelial nitric oxide synthase gene polymorphism and elite endurance athlete status: the
Genathlete study. Scand J Med Sci Sports, 18(4), 485–​490. doi: 10.1111/​j.1600-​0838.2007.00717.x
Wolfarth, B., Rankinen,T., Muhlbauer, S., Scherr, J., Boulay, M. R., Perusse, L., Rauramaa, R., & Bouchard,
C. (2007). Association between a beta2-​adrenergic receptor polymorphism and elite endurance per-
formance. Metabolism, 56(12), 1649–​1651. doi: 10.1016/​j.metabol.2007.07.006
Wolfarth, B., Rivera, M. A., Oppert, J. M., Boulay, M. R., Dionne, F.T., Chagnon, M., Gagnon, J., Chagnon,
Y., Perusse, L., Keul, J., & Bouchard, C. (2000). A polymorphism in the alpha2a-​adrenoceptor gene
and endurance athlete status. Med Sci Sports Exerc, 32(10), 1709–​1712.

387
388

27
SPORT CONCUSSION
GENETICS
Ryan T. Tierney and Jane K. McDevitt

Reading news headlines could lead someone to believe that playing contact sports such as foot-
ball will lead to a degenerative neurological condition in early adulthood. Indeed, participation
in sports such as high school football has declined as the number of news reports of concussion
in sports has increased (11). However, research indicates that although higher than the normal
population, only a low percentage of former athletes such as NFL players are reporting such
problems (20). Since most of these players were likely exposed to thousands of subconcussive
(and some concussive) impacts over a career, why are only some athletes exhibiting problems?
The answer may lie in intrinsic factors such as genetics in combination with their environment
(repetitive head impacts). Thus, risk of poor mental or cognitive health may not come from
playing contact sports alone, but playing them in addition to having a specific genetic predis-
position. In this chapter, we will explore genes and polymorphisms potentially associated with
sport concussion.

Concussion pathomechanics and pathophysiology

Traumatic brain injury (TBI) can be viewed along a continuum (Figure 27.1) with high
head-​impact acceleration events (e.g., high-​speed car accident) yielding severe TBI and low
head accelerations (e.g., soccer heading) eliciting seemingly no injury (i.e., subconcussion).
Concussion and subconcussion are on the low end of the continuum and are the most common
impact types encountered in sports.The level of injury is determined through clinical assessment
and, when appropriate, includes diagnostic imaging assessment. Subconcussion is a head impact
(or head acceleration event) that does not result in any outward signs (e.g., unsteadiness) or
subjective reporting of symptoms (e.g., headache) (4). A concussion is a head impact (or head
acceleration event) that does result in signs and symptoms (37). A concussion would not yield
any remarkable signs on diagnostic imaging (e.g., magnetic resonance imaging), which helps
differentiate concussion from a moderate or severe brain injury.
To understand the role genetics may play in predisposing someone to a concussion, it is
important to further understand concussion injury pathomechanics and pathophysiology.
A concussion is an injury to the brain caused by acceleration forces (15). These forces can
occur from something hitting the head, the head hitting something, or an indirect blow to the

388
389

Sport concussion genetics

Low Head High Head

Acceleration Acceleration

Subconcussion Concussion – Moderate TBI Severe TBI

Mild TBI

Figure 27.1 Traumatic brain injury pathomechanics spectrum.

body causing whiplash (3). The brain, which is floating in cerebrospinal fluid, moves rapidly
within the skull creating tensile, compressive, or shear forces on the brain tissue producing cell
deformation (i.e., pathomechanics) (45). The two main cell types in the brain are neurons, the
conducting cells, and glia, the supporting cells. Identifying structures that manage stress and
strain within these cells is important to finding a possible genetic predisposition to concussion
(Figure 27.2).
A number of neuronal structures can help manage mechanical stress including the cell mem-
brane, microtubules, and intermediate filaments. The cell membrane is stabilized by cholesterol
embedded, in part, by the function of apolipoprotein. Neuronal microtubules give cells form
and help transport cellular material, and are stabilized by the protein tau. Neuronal intermediate
filaments are composed of proteins such as neurofilament heavy (NEFH) and light (NEFL),
and form an internal cell scaffolding that helps create form and resist changes in cell shape. In
astrocytes, the predominant glial cell comprising a large percentage of brain volume, a main
component of intermediate filaments is glial fibrillary acidic protein (GFAP). These structures
help resist cell deformation in response to mechanical stress. If deformation is large enough (e.g.,
> 5% of resting length), this axonal stretch can lead to neurometabolic events and cell dysfunc-
tion (i.e., concussion pathophysiology) (13).
Concussion pathophysiology involves a neurometabolic cascade of cellular events that
occur post insult and trigger neuron dysfunction. These events have been described previ-
ously (16) and can include ion flux (e.g., Na+, Ca+) and glutamate release, cytoskeletal damage
to microtubules and neurofilaments, altered neurotransmission due to glutamate and calcium
channel disruption, inflammation, and potential cell death.Therefore, proteins involved in cyto-
skeleton stabilization (e.g., tau), cell membrane channel function, and glutamate regulation may
play key roles in attenuating concussion pathophysiology. The ability of the proteins mentioned
above to function properly to help withstand mechanical stress and restore proper neuronal
function is largely due to their structure. In the next sections, we will review some of the genes
that code for these proteins (see Tables 27.1 and 27.2), in particular distinguishing studies that
have focused on pathomechanical and pathphysiological outcomes.

Pathomechanical genetic associations

Apolipoprotein E
Apolipoprotein E (APOE) is one of the most studied genes associated with concussion and
manifests multiple genetic polymorphisms. The APOE gene is located on chromosome 19 and
produces the apolipoprotein E (ApoE) glycoprotein.This protein is involved in the redistribution
of lipids to maintain and restore the neuronal membrane after an injury (7). ApoE exists in three
isoforms, ApoE2, ApoE3, and ApoE4 encoded by three common alleles ε2, ε3, and ε4, respect-
ively, and result in changes in different single amino acid changes (34). The three allelic variants

389
390

Figure 27.2 Cellular structures involved in concussion pathomechanics and pathophysiology.

391

Table 27.1 Concussion pathomechanical genetic associations

Protein Gene Variant Findings

(abbreviation)

Apolipoprotein E Apolipoprotein E ε4 30 professional boxers (23–​76 years of age)

(APOE) underwent neurological examination using
the chronic brain injury scale. Authors found
the ε4 allele was associated with worse
cognition scores in high-​exposure boxers
(22)
ε4 318 collegiate athletes (~52% male) were
prospectively followed after a concussion.
Authors found after adjusting for sex,
weight, height, and team type that there
was no association with ε4 and concussion
risk (28)
ε4 195 collegiate athletes (163 football and 33
female soccer players) self-​reported the
number of previous concussions (72 with
concussion, 123 with no concussion). After
adjusting for age, school, and years of sport
experience, carrying the ε4 allele was not
associated with concussion history (49)
ε2, ε3, ε4, 196 collegiate athletes (163 football and 33
G(−219) female soccer players) self-​reported the
T-​promoter number of previous concussions (48 with
SNP concussion, 148 with no concussion).
Authors found carrying the T promoter
allele and X alleles were associated with
history of concussion (52)
Apolipoprotein Apolipoprotein G(−219)T-​ 196 collegiate athletes (163 football and 33
E E (APOE) promoter female soccer players) self-​reported the
SNP number of previous concussions (48 with
concussion, 148 with no concussion).
Authors found an increased risk of two
or more concussions when carrying the
minor T allele (52)
G(−219)T-​ 195 collegiate football and female soccer
promoter athletes self-​reported previous history
SNP of concussion (72 with concussion, 123
with no concussion). After adjusting for
age, school, and years of sport experience,
carrying the APOE promoter TT was
associated with increased concussion risk
(49)
(continued)
392

R.T. Tierney and J.K. McDevitt

Table 27.1 (Cont.)

Protein Gene Variant Findings

(abbreviation)
Tau Microtubule-​ Tau E6 195 collegiate football and female soccer
associated variants athletes self-​reported previous history of
protein tau concussion (72 with concussion, 123 with
(MAPT) no concussion). After adjusting for age,
school, and years of sport experience there
was no association between increased
concussion risk and the Ser53Pro tau
polymorphism (49)
Neurofilament Neurofilament rs165602, 48 athletes with a reported history of
heavy heavy glutamic concussion were matched by age, height,
polypeptide (NEFH) acid to an sport, and position to 48 athletes with no
alanine at history of concussion (~19 years of age).
amino acid Authors found no association between
sequence concussion history or risk and carrying
805, E805A the rare C allele (38)

Table 27.2 Concussion pathophysiological genetic associations

Protein Gene Variant Findings

(abbreviation)

Glutamate Glutamate GT(n)-​promoter 87 concussed athletes (~19 years of age; 74%

receptor receptor repeat male) were followed prospectively. Authors
inotropic inotropic polymorphism found an association with prolonged
NMDA type NMDA type recovery (greater than 60 days) from
subunit 2A subunit 2A concussion and carrying 25 or more GT
(GRIN2A) repeats (39)
Solute carrier Solute carrier rs74174284 40 concussed athletes (72% male; ~20 years
family 17 family 17 of age) were followed prospectively.
member member 7 Authors found an association with
7; vesicle (SLC17A7) prolonged recovery (greater than 20 days)
glutamate from concussion and carrying the G
transporters allele (33)
(VGLUT 1)

are distinguished by single nucleotide polymorphisms (SNPs) in exon 4 of the APOE gene. SNP
rs7412C>T is present in alleles ε3 and ε4. SNP rs429358T>C is present in allele ε4. The most
common allele is ε3 (65–​70% frequency), with two others being rarer variants (ε2, 5–​10%, and
ε4, 15–​20%;) (35). The rs7412C>T SNP that causes a change of an amino acid in ApoE2 effects
binding to the low-​density lipoprotein receptor, while the rs429358T>C SNP that produces
the ApoE4 polypeptide results in preferential binding of very low density lipoprotein particles
and reduced ApoE4 stability (35). Previous studies attributed these effects to altered interactions
between two functional domains within the ApoE polypeptide chain (35,54).

392
393

Sport concussion genetics

The effect of ApoE allelic variants on TBI outcome are likely modulated by mechanical
force, age, or time post TBI. Several studies have indicated an association between APOE geno-
type and moderate/​severe TBI poor outcome. Several studies have also demonstrated that the
rare APOE variants are associated with greater risk of concussion, greater severity of symptoms,
or poor outcomes following a concussion injury.
Recent meta-​analyses of studies investigating the role of ApoE4 on moderate and TBI indi-
cate an association with an increased risk of poor outcome in pediatric carriers of at least one
ApoE4 allele (29). Such an association presents differently depending on trauma severity. In
mild concussion, ApoE4 was not associated with post-​concussion outcomes in 58.3% of the
studies, while in more severe TBI, the role of ApoE4 was hazardous in 63.6% of the studies.
Another meta-​analysis of six studies including a total of 358 cases of pediatric TBIs revealed that
at 6 months, there was over two times higher odds of poor outcome following TBI in children
with at least one APOE ε4 allele, compared with children without an ε4 allele (26). In a meta-​
analysis of 12 studies on ApoE and functional outcome after TBI revealed an association with
increased risk of unfavorable long-​term functional outcome (≥6 months) (30). Due to the wide
spectrum of brain injury severity, it is difficult to ascertain if TBI is different from concussion
or repeated head impacts. Several studies are trying to bridge this gap; however, there are con-
troversial findings.
For example, carrying the ε4 allele has been associated with poor outcome following TBI.
Within a small sample of 30 boxers, including high-​exposure boxers (12 or more professional
matches; 23–​76 years of age), researchers examined the relationship between carrying the ε4
allele and their Chronic Brain Injury Scale scores. The Chronic Brain Injury Scale assesses
motor function (gait), cognitive deficits, and behavioral abnormalities, where 0 is normal and
above 4 is severely impaired. The authors found that carriers of the ε4 allele had worse Chronic
Brain Injury Scale scores (37% had normal scores). Within the same study, all boxers with
severe impairment were found to carry at least one ε4 allele, and had high exposure to boxing
(60% participated in 12 or more professional bouts) (22). Other studies, which examined self-​
reported concussion history and APOE genotype, reported no association with carrying the ε4
allele or other APOE variants were significantly associated with poor outcomes (28, 49, 50, 52).
In a large cohort of 318 collegiate athletes (~20 years of age; 51% male; eight different sports)
authors evaluated the association between APOE ε4 and concussion risk.The authors measured
time to first concussion and calculated athletic exposures. The mean number of athletic
exposures was approximately 100. Approximately 25% of the athletes carried the ε4 allele and
8% suffered a concussion during the course of the research study. The authors found no associ-
ation between carrying the APOE ε4 allele and sustaining a concussion (28). In a multicenter
study that included 195 college athletes (~19 years of age; male football, female soccer players)
authors assessed self-​reported concussion over an 8-​year time period obtained via a question-
naire. The athletes reported a total of 97 documented concussions. None of the athletes with
reported concussion carried the homozygous ε4 genotype; however, approximately 40% of the
concussed cohort carried the G(−219)T SNP in the APOE promoter region. Therefore the
authors demonstrated that homozygous carriers of the promoter G(−219)T SNP in the APOE
promoter region were at increased risk for having a history of concussion, but did not find any
difference between ε2, ε3, and ε4 allele carriers (49, 50).
In a similar multicenter study, the authors evaluated the self-​reported history of documented
concussions within a cohort of 196 collegiate athletes (163 football and 33 female soccer
players). Twenty-​five percent of the athletes suffered a concussion, of whom 32% carried the
ε4 allele, and 50% carried the promoter allele. The authors found a significant association was
reported between college athletes who carried all three alleles and history of concussion (i.e.,

393
394

R.T. Tierney and J.K. McDevitt

heterozygous ε2/​ε4 and carried the rare promoter allele G) (52). In the same study Tierney
et al. also found a significant association between homozygous carriers of G(−219)T SNP
and athletes reporting two or more concussions. With only limited evidence suggesting the
important role APOE allelic variants play in concussion incidence and recovery, conflicting
results do not allow us to make a formal confirmation of this hypothesis at this time. The
ultimate confirmation would require prospective, multicenter studies with sufficient power and
carefully selected inclusion/​exclusion criteria (18, 51). For example, a genome-​wide association
study would be beneficial to identify if APOE is a significant genetic marker for brain injury.
For this type of research design, thousands of athletes’ genomes with a concussion history would
be compared to thousands of those with no concussion history.
One of the reasons APOE is targeted as a genetic marker is due to the fact that head injury
is reported to trigger amyloid B protein deposition, which is also a key factor in the pro-
gression of Alzheimer’s disease (AD) (44). Similar to TBI, AD displays as a cognitive decline.
A recent review found this cognitive decline seems to be accelerated in patients carrying
the APOE ε4 allele (31). This was attributed to the APOE4 protein being not as efficient
in delivering cholesterol to maintain synaptic transmission. Additionally, the authors of the
review found that APOE4 protein promotes proinflammatory processes that further exacer-
bate the disease progression. Similar to genetic associations to brain injury, not everyone
carrying the rare allele/​genotype has poor outcomes. Approximately 34–​65% of patients with
AD carry the APOE ε4 allele, while it is present in about 20–​30% of the nonaffected adult
population (43).

Tau
Tau is a microtubule-​associated protein that stabilizes microtubules. The human tau gene
(MAPT) is located on chromosome 17 and contains 16 exons (coding regions of the gene).Two
different tau gene haplotypes have been identified (H1 and H2), consisting of eight common
SNPs. H1 is the most common, and it is overexpressed in disorders such as progressive neuro-
logical disorders. In the central nervous system, alternative splicing of exons 2, 3, and 10 result
in the appearance of six tau isoforms (2).
There is a paucity of research examining tau genotype and concussion. Terrell et al.
(details of study reported above) (49, 50) reported that tau variants in exon 6 (Ser53Pro
(TauSer), Hist47Tyr (TauHis)) were not associated with acute concussion. In spite of this,
tau does seem to be an important protein to focus on for genetic predisposition to brain
injury due to its association with chronic traumatic encephalopathy (CTE) and AD. CTE is
a neurodegeneration characterized by the abnormal accumulation of hyperphosphorylated
tau protein within the brain; however, CTE can only be definitively diagnosed postmortem
(40). CTE is also thought to be associated with subconcussive blows to the head (14).
Therefore, it is plausible that tau genetic polymorphisms may be key in determining the
risk of CTE or AD in relation to contact sport participation. Currently, there are no genetic
studies involving tau, contact sport participation, and CTE. In a postmortem study, authors
examined numerous genes in 17 military personnel and football athletes with a history of
repetitive brain injury and found that MAP2 is upregulated in brains with late-​stage CTE
(42). Similar to MAPT, MAP2 codes for microtubule-​associated proteins involved in stabil-
izing the dendrite shape during neurodevelopment. The authors did not evaluate MAPT.
Currently, many of the studies examining tau focus on cerebral spinal fluid and plasma tau
protein concentrations following acute brain injury (56).

394
395

Sport concussion genetics

Neurofilament
The neuronal cytoskeleton is composed of 50% neurofilaments including light (NEFL), medium
(NEFM) and heavy neurofilament (NEFH) (17, 53). These neurofilaments are composed of
proteins that determine the quality of the cytoskeleton. The NEFH gene resides on chromo-
some 22, and several variants within this gene have been associated with neurodegenerative
conditions, such as amyotrophic lateral sclerosis (1).
Only the NEFH has been examined to date in association with concussion outcomes. The
authors of one study sought to determine an association between a SNP within a gene coding
for a neuronal structural protein (i.e., neurofilament heavy) and previous occurrence and severity
of concussions in college athletes (38). Using a case–​control study design, 48 athletes with a
reported history of concussion were matched by age, height, sport, and position to 48 athletes
with no history of concussion (~19 years of age). The authors reported that 24% of the athletic
cohort carried the NEFH rare allele, but there was no significant association between carrying
the rare allele and concussion history or severity. Although the rare allele alters the NEFH pro-
tein sequence, this SNP does not seem to influence an athlete’s susceptibility to concussion (38).

Pathophysiological genetic associations

Glutamate receptor ionotropic NMDA receptor 2A (GRIN2A)
The primary function of the N-​methyl-​D-​aspartate (NMDA) receptor is to act as synaptic
connectivity between two neurons as well as to trigger postsynaptic potentials and dendritic
spikes, where action potentials are formed. Mechanical stress activates NMDA receptors via
overstimulation by the increased glutamate concentration within the synaptic cleft, and causes
excitotoxicity (injury to the nerve due to excessive stimulation by glutamate). Ionotropic
NMDA receptors are recognized as the major source of glutamate excitotoxicity dependent on
the influx of Ca2+ when glutamate binds to NMDA receptors. NMDA receptors are composed
of four subunits forming a ligand-​gated cation (e.g., Ca2+) channel in which the NR1–​NR2A
heterodimer is the functional unit. The main NR2 subunits are NR2A and NR2B encoded
by the GRIN2A and GRIN2B genes, respectively. GRIN2A is located on chromosome 16 and
contains multiple SNPs in the coding, intronic, and promoter regions, and has been implicated
in several cognitive brain diseases including dementia, AD, depression, and schizophrenia (21).
Functional involvement of NMDA receptors in the concussion stress response is supported by
several lines of evidence (25, 41). An important role of polypeptide components of the NMDA
receptor was demonstrated in experiments with genetically manipulated mice. Less brain
ischemia was detected in NR2A or NR2A/​NR2B knockout mice, after they were subjected to
focal cerebral ischemia. The lack of NR2A is likely to alleviate glutamate excitotoxicity due to
the decreased amount of blood volume, which could be explained by decreased NMDA channel
activity (41). The reduced functionality of the NMDA receptors results in less Ca2+ entering
the cell (25). Taken together, the above data support the premise that variability in NMDA
receptor expression could be a risk factor for concussion outcome. Therefore, variants of the
genes coding for the components of the NMDA receptor complex are attractive candidates for
association with concussion incidence or recovery.
One important type of genetic variability is the number of tandem repeats (i.e., GT) in the
promoter region of the GRIN2A gene. GRIN2A expression is modulated by (GT)n VNTR
in the promoter region. The (GT)n VNTR in the promoter region had been earlier associated
with altered expression level of GRIN2A (23, 25). The length of (GT)n repeat modulates

395
396

R.T. Tierney and J.K. McDevitt

GRIN2A expression level, with longer alleles (≥25 repeats) associated with lower transcription
of GRIN2A mRNA. In a study of 87 athletes (~19 years of age; 74% male) suffering with a
concussion, homozygous carriers of the longer alleles (>25 repeats) were found to be six times
more likely to recover in 60 or more days, compared with homozygous carriers of the shorter
(<25 repeats) alleles (39). With athletes categorized to one of two groups (prolonged recovery
vs. normal recovery), significant variation between the frequencies of longer alleles and shorter
alleles was detected, where the carriers of longer alleles were two times more likely to be
in the prolonged recovery group than those carrying shorter alleles. Moreover, homozygous
carriers of longer alleles demonstrated a significant association with prolonged recovery when
compared with homozygous carriers of shorter alleles (39).

Vesicle glutamate transporters

Uptake of glutamate at the synapse following a mechanical insult (e.g., concussion) is important
for the restoration of normal neuron function.The synaptic uptake of glutamate is facilitated by
vesicular transporters (i.e., vesicle glutamate transporter (VGLUT)-​1,VGLUT2, and VGLUT3)
encoded by the solute carrier subfamily of genes located on chromosomes 19 (SLC17A7), 11
(SLC17A6), and 12 (SLC17A8), correspondingly.VGLUT1 performs synaptic uptake of gluta-
mate and deposits it into neurosecretory vesicles and is expressed in cerebrum, cerebellum, and
hippocampus. Downregulation of VGLUT1 vesicular transport production was shown to cause
severe changes in the neurological phenotype of experimental animals (12). Knockout studies
demonstrated reduction of vesicle pool size accompanied by residually high concentrations of
glutamate within the synaptic cleft (55). Another piece of evidence indicating that VGLUT
activity could modulate synaptic efficacy came from the clinical studies of VGLUT expression
in schizophrenic patients. VGLUT expression in the hippocampus and the dorsolateral pre-
frontal cortex was reduced in the brains of schizophrenic patients (9).
If the expression level of a protein responsible for reducing the amount of glutamate (e.g.,
VGLUT1) within the synaptic cleft is altered due to genetic variation, this may affect the
severity of the concussive injury, and the recovery time. A study in 40 athletes (~20 years of age;
73% male) demonstrated that carriers of the rs74174284:G allele in the SLC17A7 promoter
were five times more likely to exceed 20 days to recover from a concussion (33). These results
are in parallel with the hypothesis that low expression of SLC17A7 encoding VGLUT1 prob-
ably reduces glutamate transport in the carriers of G allele (55).

Future directions
Sport concussion is its own entity within TBI science. Concussion is not severe TBI or a
neurodegenerative disease. As we learn more about how these injuries and diseases overlap, it
provides an opportunity to explore new genetic associations. Table 29.3 provides a summary of
some genes (and proteins) studied in other conditions or that have a hypothetical role in the
pathomechanics and/​or pathophysiology of sport concussion. Genetic associations studied in
other neurological conditions, in addition to our ever-​g rowing understanding of concussion,
are the key to enlightening the science of sport concussion genetics.
This type of genetics research can be challenging for many reasons. For example, there are
multiple outcomes with variable responses that could have genetic associations such as the
acute and long-​term response to repetitive subconcussion, acute concussion, or severe TBI.
Further complicating matters is that each of these injuries are based on clinical decision-​making
and management. For example, a researcher studying concussion response (e.g., return to play

396
397

Table 27.3 Potential concussion genetic associations

Protein Gene (abbreviation) Variant Findings

Glial fibrillary Glial fibrillary acidic 10 different A literature review revealed isoforms α, β, γ, δ,
acidic protein protein (GFAP) isoforms and ε were associated with a greater risk of
within Alexander’s disease (46)
exons 6, 7
and the 5ʹ
UTR
S100 calcium S100 calcium rs9722 794 participants (396 ischemic stroke patients
binding binding B and 398 controls; ~58 years of age; ~61%
protein B (S100B) male). Authors found carriers of the
A allele had an increased S100B serum,
which increased risk for ischemic stroke
(32)
Calcium Calcium rs121908225 Within 3 participants with delayed severe
voltage-g​ ated voltage-​gated edema and 152 nonaffected family
channel channel subunit members and controls, authors found that
subunit alpha-​1A carriers of the T allele had increased brain
alpha-​1A (CACNA1A) swelling and coma after TBI (27)
In a review of literature, authors found an
increased risk of hemiplegic migraine and
cerebral edema after minor head injury in
carriers of the S218L mutation in exon 5
(48)
Calcium Calcium rs704326 1q31 and 1q32 provide genes that code for
voltage-g​ ated voltage-​gated proteins such as ATPase, and proteins
channel channel subunit integral for active transport of sodium ions.
subunit alpha-​1E This SNP within exon 43 lies within this
alpha-​1E (CACNA1E) region of the gene that is associated with
migraines (10)
Glutamate Glutamate rs228411 Using the FBAT program, the authors found
receptor receptor the C allele carriers had greater symptoms
ionotropic ionotropic scored and were associated with attention
NMDA type NMDA type deficit disorder (8)
subunit 2B subunit 2B
(GRIN2B)
Glutamate Glutamate Tandem Associated with autism and intellectual
ionotropic ionotropic repeats in disability (5, 6, 19)
receptor receptor AMPA exons 4, 5,
AMPA type type subunit 1–​ and 12
subunit 1–​4 4 (GRIA1–​4)
(continued )
398

R.T. Tierney and J.K. McDevitt

Table 27.3 (Cont.)

Protein Gene (abbreviation) Variant Findings

Brain-​derived Brain-​derived rs6562 113 participants (75 with mild TBI and 38
neurotrophic neurotrophic healthy; ~33 years of age; 61% male) were
factor factor (BDNF) assessed using five different cognitive
assessments. The authors found that the
Met allele was associated with slower
processing speed in the entire group, but
not within the mild TBI group specifically.
Other SNPS in linkage disequilibrium may
increase risk of slower processing speeds
after mild traumatic brain injury (36)

time) should consider factors such as: When was injury identified? Who identified the injury?
What assessments were perfomed? Was rehabilitation implemented? What was the return to play
protocol? and When was it initiated? Due to the number of factors affecting outcomes, large
multicentered prospective studies involving a large number of particpants are needed.
The largest study of sport concussion ever conducted began in 2015 by the Care Consortium
(www.careconsotium.net). The Care Consortium is a joint effort by the National Collegiate
Athletic Association (NCAA) and the US Department of Defense involving 30 NCAA
institutions as data collection sites. The consortium aims to conduct prospective, longitudinal
research to study the natural history of concussion and conduct advanced studies integrating
biomechanics, neuroimaging, and neurobologic and genetic markers of injury. Over the initial
3-​year period, more than 39,000 baseline and 2800 postinjury assessments have been performed.
Additionally, researchers are looking to examine athletes over the long term, making this one
of the most important research endeavors ever conducted. Projects such as this are the key to
advancing the science of sport concussion genetics.

References
1. Al-​Chalabi A, Andersen PM, Nilsson P, Chioza B, Andersson JL, Russ C, Shaw CE, Powell
JF, and Leigh PN. Deletions of the heavy neurofilament subunit tail in amyotrophic lateral sclerosis.
Human Molecular Genetics 8: 157–​164, 1999.
2. Andreadis A, Brown WM, and Kosik KS. Structure and novel exons of the human tau gene.
Biochemistry 31: 10626–​10633, 1992.
3. Aubry M, Cantu R, Dvorak J, Graf-​Baumann T, Johnston K, Kelly J, Lovell M, McCrory P,
Meeuwisse W, and Schamasch P. Summary and agreement statement of the first international con-
ference on concussion in sport,Vienna. British Journal of Sports Medicine 36: 6–​10, 2002.
4. Bailes JE, Petraglia AL, Omalu BI, Nauman E, Talavage T. Role of subconcussion in repetitive
mild traumatic brain injury. Journal of Neurosurgery 119: 1235–​1245, 2013.
5. Bonnet C, Leheup B, Béri M, Philippe C, Grégoire M-​J, and Jonveaux P. Aberrant GRIA3
transcripts with multi-​exon duplications in a family with X-​linked mental retardation. American Journal
of Medical Genetics Part A 149A: 1280–​1289, 2009.
6. Chiyonobu T, Hayashi S, Kobayashi K, Morimoto M, Miyanomae Y, Nishimura A,
Nishimoto A, Ito C, Imoto I, Sugimoto T, Jia Z, Inazawa J, and Toda T. Partial tandem
duplication of GRIA3 in a male with mental retardation. American Journal of Medical Genetics Part A
143A: 1448–​1455, 2007.

398
399

Sport concussion genetics

7. Dardiotis E, Fountas KN, Dardioti M, Xiromerisiou G, Kapsalaki E, Tasiou A, and

Hadjigeorgiou GM. Genetic association studies in patients with traumatic brain injury. Neurosurgical
Focus 28: 1–​12, 2010.
8. Dorval KM, Wigg KG, Crosbie J, Tannock R, Kennedy JL, Ickowicz A, Pathare T, Malone
M, Schachar R, and Barr CL. Association of the glutamate receptor subunit gene GRIN2B with
attention-​deficit/​hyperactivity disorder. Genes, Brain and Behavior 6: 444–​452, 2007.
9. Eastwood SL and Harrison PJ. Decreased expression of vesicular glutamate transporter 1 and
complexin II mRNAs in schizophrenia: further evidence for a synaptic pathology affecting glutamate
neurons. Schizophrenia Research 73: 159–​172, 2005.
10. Fernandez F, Curtain RP, Colson NJ, Ovcaric M, MacMillan J, and Griffiths LR. Association
analysis of chromosome 1 migraine candidate genes. BMC Medical Genetics 8: 57, 2007.
11. Feudtner C and Miles SH. Traumatic brain injury news reports and participation in high school
tackle football. JAMA Pediatrics 172: 492–​494, 2018
12. Fremeau RT, Kam K, Qureshi T, Johnson J, Copenhagen DR, Storm-​Mathisen J, Chaudhry FA,
Nicoll RA, and Edwards RH.Vesicular glutamate transporters 1 and 2 target to functionally distinct
synaptic release sites. Science 304: 1815–​1819, 2004.
13. Gaetz, M. The neurophysiology of brain injury. Clinical Neurophysiology 115: 4–​18, 2004.
14. Gavett BE, Stern RA, Cantu RC, Nowinski CJ, and McKee AC. Mild traumatic brain injury: a
risk factor for neurodegeneration. Alzheimer’s Research and Therapy 2: 18, 2010.
15. Gennarelli TA, Segawa H, Wald U, Czernicki Z, Marsh K, and Thompson C. Physiological
response to angular acceleration of the head. In: Head injury: Basic and Clinical Aspects, edited by
Grossman RG, Gildenberg PL. New York: P. L. Ravens Press, 1982, p. 129–​140.
16. Giza CC and Hovda DA. The new neurometabolic cascade of concussion. Neurosurgery
75: S24–​S33, 2014.
17. Godsel LM, Hobbs RP, and Green KJ. Intermediate filament assembly and dynamics of disease.
Cell 18: 28–​37, 2007.
18. Gordon KE. Apolipoprotein E genotyping and concussion: time to fish or cut bait. Clinical Journal
of Sports Medicine 20: 405–​406, 2010.
19. Guilmatre A, Dubourg C, Mosca AL, Legallic S, Goldenberg A, Drouin-​Garraud V, Layet
V, Rosier A, Briault S, Bonnet-​Brilhault F, Laumonnier F, Odent S, Le Vacon G, Joly-​Helas
G, David V, Bendavid C, Pinoit JM, Henry C, Impallomeni C, Germano E, Tortorella G, Di
Rosa G, Barthelemy C, Andres C, Faivre L, Frébourg T, Saugier Veber P, and Campion D.
Recurrent rearrangements in synaptic and neurodevelopmental genes and shared biologic pathways
in schizophrenia, autism, and mental retardation. Archives in General Psychiatry 66: 947–​956, 2009.
20. Guskiewicz KM, Marshall SW, Bailes J, McCrea M, Cantu RC, Randolph C, and Jordan BD.
Association between recurrent concussion and late-​life cognitive impairment in retired professional
football players. Neurosurgery 57: 719–​726, 2005.
21. Huang YJ, Lin CH, Lane HY, and Tsai GE. NMDA neurotransmission dysfunction in behavioral
and psychological symptoms of Alzheimer’s disease. Current Neuropharmacology 10: 272–​285, 2012.
22. Jordan BD, Relkin NR, Ravdin LD, Jacobs AR, Bennett A, and Gandy S. Apolipoprotein E
epsilon4 associated with chronic traumatic brain injury in boxing. Journal of the American Medical
Association 278: 136–​140, 1997.
23. Inoue H, Yamasue H, Tochigi M, Suga M, Iwayama Y, Abe O, Yamada H, Rogers MA, Aoki
S, Kato T, Sasaki T, Yoshikawa T, and Kasai K. Functional (GT)n polymorphisms in promoter
region of N-​methly-​D-​aspartate receptor 2A subunit (GRIN2A) gene affect hippocampal and amyg-
dala volumes. Genes, Brain and Behavior 9: 269–​275, 2010.
24. Itokawa M, Yamada K, Yoshitsugu K, Toyota T, Suga T, Ohba A, Watanabe A, Hattori E,
Shimizu H, Kumakura T, Ebihara M, Meerabux JM, Toru M, and Yoshikawa T. A micro-
satellite repeat in the promoter of the N-​methyl-​D-​Aspartate receptor 2A subunit (GRIN2A)
gene suppresses transcriptional activity and correlates with chronic outcome in schizophrenia.
Pharmacogenetics 13: 271–​278, 2003.
25. Iwayama-​Shigeno Y, Yamada K, Itokawa M, Toyota T, Meerabux J. Extended analyses support
the association of a functional (GT)n polymorphism in the GRIN2A promoter with Japanese schizo-
phrenia. Neuroscience Letters 378: 102–​105, 2005.
26. Kassam I, Gagnon F, and Cusimano MD. Association of the APOE-​epsilon4 allele with out-
come of traumatic brain injury in children and youth: a meta-​analysis and meta-​regression. Journal of
Neurology, Neurosurgery, and Psychiatry 87: 433–​440, 2016.

399
400

R.T. Tierney and J.K. McDevitt

27. Kors EE, Terwindt GM, Vermeulen FL, Fitzsimons RB, Jardine PE, Heywood P, Love S, van
den Maagdenberg AM, Haan J, Frants RR, and Ferrari MD. Delayed cerebral edema and fatal
coma after minor head trauma. role of the CACNA1A calcium channel subunit gene and relationship
with familial hemiplegic migraine. Annals of Neurology 49: 753–​760, 2001.
28. Kristman VL, Tator CH, Kreiger N, Richards D, Mainwaring L, Jaglal S, Tomlinson G, and
Comper P. Does the apolipoprotein epsilon 4 allele predispose varsity athletes to concussion? A pro-
spective cohort study. Clinical Journal of Sports Medicine 18: 322–​328, 2008.
29. Lawrence DW, Comper P, Hutchison MG, and Sharma B. The role of apolipoprotein E epsilon
(epsilon)-​4 allele on outcome following traumatic brain injury: a systematic review. Brain Injury
29: 1018–​1031, 2015.
30. Li L, Bao Y, He S, Wang G, Guan Y, Ma D, Wu R, Wang P, Huang X, Tao S, Liu Q, Wang Y,
and Yang J. The association between apolipoprotein E and functional outcome after traumatic brain
injury: a meta-​analysis. Medicine (Baltimore) 94: e2028, 2015.
31. Liu C, Kanekiyo T, Xu H and Bu1 G. Apolipoprotein E and Alzheimer disease: risk, mechanisms,
and therapy. Nature Reviews Neurology 9: 106–​118, 2013.
32. Lu Y, Wang R, Huang H, Qin H, Liu C, Xiang Y, Wang C, Luo H, Wang J, Lan Y, and Wei
Y. Association of S100B polymorphisms and serum S100B with risk of ischemic stroke in a Chinese
population. Scientific Reports 8: 971, 2018.
33. Madura SA, McDevitt JK, Tierney RT, Mansell JL, Hayes DJ, Gaughan JP, and Krynetskiy E.
Genetic variation in SLC17A7 promoter associated with response to sport-​related concussions. Brain
Injury 30: 908–​913, 2016.
34. Mahley RW. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology.
Science 240: 622–​630, 1988.
35. Mahley RW. Apolipoprotein E: From cardiovascular disease to neurodegenerative disorders. Journal
of Molecular Medicine (Berlin) 94: 739–​746, 2016.
36. McAllister TW, Tyler AL, Flashman LA, Rhodes H, McDonald BC, Saykin AJ, Tosteson
TD, Tsongalis GJ, and Moore JH. Polymorphisms in the brain-​ derived neurotrophic factor
gene influence memory and processing speed one month after brain injury. Journal of Neurotrauma
29: 1111–​1118, 2012.
37. McCrory P, Meeuwisse W, Dvorak J, Aubry M, Bailes J, Broglio S, Cantu RC, Cassidy
D, Echemendia RJ, Castellani RJ, Davis GA, Ellenbogen R, Emery C, Engebretsen L,
Feddermann-​Demont N, Giza CC, Guskiewicz KM, Herring S, Iverson GL, Johnston KM,
Kissick J, Kutcher J, Leddy JJ, Maddocks D, Makdissi M, Manley GT, McCrea M, Meehan
WP, Nagahiro S, Patricios J, Putukian M, Schneider KJ, Sills A, Tator CH, Turner M, and Vos
PE. Consensus statement on concussion in sport-​The 5th international conference on concussion in
sport held in Berlin, October 2016. British Journal of Sports Medicine 51: 838–​847, 2017.
38. McDevitt JK, Tierney RT, Mansell JL, Driban JB, Higgins M, Toone N, Mishra A, and
Krynetskiy E. Neuronal structural protein polymorphism and concussion in college athletes. Brain
Injury 25: 1108–​1113, 2011.
39. McDevitt J, Tierney RT, Phillips J, Gaughan JP, Torg JS, and Krynetskiy E.Association between
GRIN2A promoter polymorphism and recovery from concussion. Brain Injury 29: 1674–​1681, 2015.
40. McKee AC, Cairns NJ, Dickson DW, Folkerth RD, Keene CD, Litvan I, Perl DP, Stein TD,
Vonsattel J, Stewart W, Tripodis Y, Crary JF, Bieniek KF, Dams-​O’Connor K, Alvarez
VE, Gordon WA, and TBI/​CTE group. The first NINDS/​NIBIB consensus meeting to define
neuropathological criteria for the diagnosis of chronic traumatic encephalopathy. Acta Neuropathology
131: 75–​86, 2016.
41. Morikawa E, Mori H, Kiyama Y, Mishina M, Asano T, and Kirino T. Attenuation of focal
ischemic brain injury in mice deficient in the epsilon1 (NR2A) subunit of NMDA receptor. Journal
of Neuroscience 18: 9727–​9732, 1998.
42. Mufson EJ, He B, Ginsberg SD, Carper BA, Bieler GS, Crawford F, E. Alvarez VE, Huber
BR, Stein TD, McKee AC, and Perez SE. Gene profiling of nucleus basalis tau containing neurons
in chronic traumatic encephalopathy: a Chronic Effects of Neurotrauma Consortium Study. Journal
of Neurotrauma 35: 1260–​1271, 2018.
43. Myers RH, Schaefer EJ, Wilson PW, D’Agostino R, Ordovas JM, Espino A, Au R, White RF,
Knoefel JE, Cobb JL, McNulty KA, Beiser A, and Wolf PA. Apolipoprotein E epsilon4 association
with dementia in a population-​based study: The Framingham study. Neurology 46:673–​677, 1996.

400
401

Sport concussion genetics

44. Nathoo N, Chetty R, van Dellen JR, and Barnett GH. Genetic vulnerability following traumatic
brain injury: the role of apolipoprotein E. Molecular Pathology. 56: 132–​136, 2003.
45. Ommaya AK and Gennarelli TA. Cerebral concussion and traumatic unconsciousness: correlation
of experimental and clinical observations on blunt head injuries. Brain 97: 633–​654, 1974.
46. Quinlan RA, Brenner M, Goldman JE, and Messing A. GFAP and its role in Alexander disease.
Experimental Cellular Research 313: 2077–​2087, 2007.
47. Single FN, Rozov A, Burnashev N, Zimmermann F, Hanley DF, Forrest D, Curran T,
Jensen V, Hvalby O, Sprengel R, and Seeburg PH. Dysfunctions in mice by NMDA receptor
point mutations NR1(N598Q) and NR1(N598R). Journal of Neuroscience 20: 2558–​2566, 2000.
48. Stam AH, Luijckx GJ, Poll-​Thé BT, Ginjaar IB, Frants RR, Haan J, Ferrari MD, Terwindt
GM, and van den Maagdenberg AM. Early seizures and cerebral oedema after trivial head trauma
associated with the CACNA1A S218L mutation. Journal of Neurology, Neurosurgery, and Psychiatry
80: 1125–​1129, 2009.
49. Terrell TR, Bostick RM, Abramson R, Xie D, Barfield W, Cantu R, Stanek M, and Ewing
T. APOE, APOE promoter, and Tau genotypes and risk for concussion in college athletes. Clinical
Journal of Sports Medicine 18: 10–​17, 2008.
50. Terrell TR, Bostick RM, Barth J, McKeag D, Cantu RC, Sloane R, Galloway L, Erlanger
D, Valentine V, and Bielak K. Genetic polymorphisms, concussion risk, and post concus-
sion neurocognitive deficits in college and high school athletes. British Journal of Sports Medicine
47: e1, 2013.
51. Terrell TR, Bostick R, Barth J, Sloane R, Cantu RC, Bennett E, Galloway L, Laskowitz D,
Erlanger D, McKeag D, Valentine V, and Nichols G. Multicenter cohort study on association of
genotypes with prospective sports concussion: methods, lessons learned, and recommendations. The
Journal of Sports Medicine and Physical Fitness 57: 77–​89, 2017.
52. Tierney RT, Mansell JL, Higgins M, McDevitt JK, Toone N, Gaughan JP, Mishra A, and
Krynetskiy E. Apolipoprotein E genotype and concussion in college athletes. Clinical Journal of Sports
Medicine 20: 464–​468, 2010.
53. Wagner OI, Rammensee S, Korde N, Wen Q, Leterrier JF, and Janmey PA. Softness, strength
and self-​repair in intermediate filament networks. Experimental Cell Research 313: 2228–​2235, 2007.
54. Wintjens R, Bozon D, Belabbas K, MBou F, Girardet JP, Tounian P, Jolly M, Boccara F,
Cohen A, Karsenty A, Dubern B, Carel JC, Azar-​Kolakez A, Feillet F, Labarthe F, Gorsky
AM, Horovitz A, Tamarindi C, Kieffer P, Lienhardt A, Lascols O, Di Filippo M, and Dufernez
F. Global molecular analysis and APOE mutations in a cohort of autosomal dominant hypercholes-
terolemia patients in France. The Journal of Lipid Research 57: 482–​491, 2016.
55. Wojcik SM, Rhee JS, Herzog E, Sigler A, Jahn R, Takamori S, Brose N, and Rosenmund C.
An essential role for vesicular glutamate transporter 1 (VGLUT1) in postnatal development and con-
trol of quantal size. Proceedings of the National Academy of Science USA 101: 7158–​7163, 2004.
56. Zetterberg H, Morris HR, Hardy J, and Blennow K. Update on fluid biomarkers for concussion.
Concussion 1: 3, 2016.

401
402

28
SYSTEMS GENETIC FACTORS
UNDERLYING SOFT
TISSUE INJURY
Masouda Rahim, Alison V. September, and Malcolm Collins

Introduction
Although there are numerous benefits to participating in regular physical activity (7, 59), the
risk of acute and chronic musculoskeletal soft tissue injuries has risen (44), with approximately
30–​50% of all sporting injuries affecting tendons (23). Over the last few decades, the incidence
of these sport-​associated injuries has escalated in both elite and recreational athletes due to
increased participation, and intensity and duration of sporting activities (44).The exact etiology
underlying soft tissue injuries remains to be elucidated though they are recognized as complex
phenotypes with both intrinsic and extrinsic factors contributing to injury risk (4).
Genetics, in particular, is increasingly recognized as a key intrinsic risk factor predisposing
individuals; with over 80 polymorphisms implicated in the profiles of several musculoskeletal
soft tissue injuries, such as anterior cruciate ligament (ACL) ruptures, tennis elbow, tendinopathy
of the posterior tibialis tendon, Achilles tendinopathy, and rotator cuff injuries, to date (53, 65,
72). These genetic loci encode an array of proteins essential for maintaining the structural and
functional integrity of musculoskeletal soft tissues and were primarily uncovered through case–​
control genetic association studies following a candidate gene approach (19).With the advance-
ment of genomics techniques in recent years, there has been a transition towards employing a
hypothesis-​free approach, using genome-​wide association studies (GWAS) and next-​generation
sequencing (NGS) technologies. These methods have the added advantage of potentially iden-
tifying novel candidates and providing researchers with new insight into the underlying patho-
physiological processes (32).
This chapter will review our current understanding and examine the latest findings on the
genetic contribution to musculoskeletal soft tissue injuries. Additionally, the clinical significance
of this research area will be discussed and future directions proposed.

Genetic risk factors underlying musculoskeletal soft tissue injuries

Candidate gene studies
Through the application of molecular genetics techniques, aided by molecular and cellular
biology techniques, researchers have begun exploring and highlighting functional biological

402
403

Factors underlying soft tissue injury

pathways and molecular mechanisms contributing to the pathogenesis of sporting injuries. The
genes implicated thus far include structural proteins, such as the collagens and fibrillins, as
well as regulators of the extracellular matrix, such as proteoglycans, matrix metalloproteinases
(MMPs), growth factors, cytokines and caspases (Tables 28.1–​28.3). Collectively, these proteins
function to maintain the integrity of the extracellular matrix (ECM) allowing the tissue to
respond and ideally appropriately adapt to loading. Keeping in mind the intricacies of the
ECM remodeling pathway, it is not surprising that alongside the independent genetic associ-
ations, several gene–​gene interactions have also been reported implicating proteins functioning
in common biological pathways (57, 63, 67). For the purpose of this review, several of the main
findings will be presented together with the proposed biological mechanisms.
Structurally, tendons and ligaments are collagenous tissues predominantly composed of
heterotrimeric type I collagen fibrils. The COL1A1 gene encoding the α1(I) chain of type
I collagen contains a functional Sp1 binding site polymorphism (rs1800012 G/​T) within its first
intron.The rare T allele was reported to have increased binding affinity for the Sp1 transcription
factor leading to increased production of α1(I) chains and by implication increased type I col-
lagen homotrimers in the tissue (40). The rs1800012 polymorphism has been associated with
risk of cruciate ligament ruptures (25, 50, 68), shoulder dislocations (25), and acute soft tissue
ruptures in a combined analysis (15); with all except one of the studies (68) reporting an under-​
representation of the TT genotype in the cases compared to the controls (Table 28.1). However,
the contrasting results may be attributable to differences in the mechanism of ACL injuries
in skiers (8). A later study by Ficek et al. (2013) reported the association of a second func-
tional single nucleotide polymorphism (SNP) (rs1107946 G/​T) within the promoter region of
COL1A1, which is in linkage disequilibrium with the Sp1 polymorphism (17). Together, the
two form a functional haplotype that regulates COL1A1 transcription (24) and the combin-
ation of the two polymorphisms was associated with a reduced risk of ACL rupture (17).
Type V collagen is a minor fibrillar collagen implicated in regulating the collagen fibril diam-
eter and its lateral growth during fibrillogenesis, with the α1(V) chain encoded by the COL5A1
gene. A number of polymorphisms (rs13946 T/​C, rs14776422 C/​T, rs5574880 G/​A, rs12722
T/​C, rs3196378 C/​A, rs71746744 −/​AGGG, rs16399 −/​ATCT, rs1134170 A/​T) within the 3ʹ-​
untranslated region (UTR) of COL5A1 have been investigated to date (Table 28.1). Specifically,
several independent and haplotype associations were reported with Achilles tendinopathy (10,
43, 66), Achilles tendon ruptures (10), tennis elbow (2), ACL ruptures in females (38, 46, 51), and
carpal tunnel syndrome –​an occupational injury in which tendon involvement was previously
proposed (11).The 3ʹ-​UTR of genes usually contains important regulatory elements and there-
fore it is not surprising that several of these polymorphisms were reported to affect COL5A1
mRNA secondary structure (1). Laguette et al. (2011) identified two major allelic forms of the
COL5A1 3ʹ-​UTR with functional consequences. Specifically, the “T functional form,” gen-
erally identified in tendinopathic patients, demonstrated increased mRNA stability compared
to the “C functional form” (35). Moreover, the microRNA, Hsa-​miR-​608 (MIR608) binds to
the 3ʹ-​UTR and regulates COL5A1 mRNA stability and the CC genotype of the functional
rs4919510 (C/​G) polymorphism was associated with increased Achilles tendinopathy risk (1).
In light of these findings, a functional hypothesis for the COL5A1 3ʹ-​UTR was proposed
wherein altered COL5A1 mRNA stability may alter type V collagen production with poten-
tial implications on collagen fibril diameter and, as a result, the biomechanical properties of
the tissue (14). In addition to these findings, genetic associations have also been noted for the
COL3A1, COL11A1, COL11A2, COL12A1, and COL27A1 genes (Table 28.1).
The structural integrity of tendons and ligaments is tightly regulated by myriad proteoglycans,
MMPs, caspases, cytokines, growth factors and signaling factors. Collectively these regulatory

403
 newgenrtpdf
404
Table 28.1 Genetic variants within the collagen genes and the associated musculoskeletal soft tissue injuries in which they were implicated

Gene and chromosomal Encoded protein Protein function Polymorphism Location Soft tissue injury Reference
location

COL1A1 (17q21) α1 (I) collagen chain Major fibrillar collagen rs1107946 (G/​T) Promoter ACLa 17
rs1800012 (G/​T) Intron 1 ACLb, shoulder 15, 17, 25, 50, 68
dislocations, acute
soft tissue ruptures
COL3A1 (2q31) α1 (III) collagen chain Major fibrillar collagen rs1800255 (G/​A) Exon 30 ACL 46, 69
COL5A1 (9q34) α1 (V) collagen chain Minor fibrillar collagen rs13946 (C/​T) 3ʹ-​UTR LE 2
rs12722 (T/​C) ACL, AT, LE 2, 10, 38, 43, 46,
51, 66
rs3196378 (C/​A) ATPb 10, 66
rs71746744 (−/​AGGG) AT 1
rs16399 (ATCT/​−)
rs1134170 (A/​T)
COL11A1 (1p21) α1 (XI) collagen chain Minor fibrillar collagen rs3753841 (T/​C) Exon 52 ATa 22
rs1676486 (C/​T) Exon 62 ATa 22
COL11A2 (6p21) α2 (XI) collagen chain rs1799907 (T/​A) Intron 6 ATa 22
COL12A1 (6q12–​q13) α1 (XII) collagen chain FACIT collagen rs970547 (A/​G) Exon 65 ACL 46, 52
COL27A1 (9q32) α1 (XXVII) collagen Minor fibrillar collagen rs946053 (G/​T) Intron 41 ATa 64
chain

Associations denoted with an a indicate the variant was associated as part of an inferred haplotype or an inferred allele combination. Associations denoted with a b
indicate the variant was independently associated with the musculoskeletal soft tissue injury, as well as part of an inferred haplotype or an inferred allele combination.
ACL=anterior cruciate ligament ruptures; AT=Achilles tendinopathy; ATP=Achilles tendon pathology; FACIT=fibril-​associated collagens with interrupted triple
helices; LE=lateral epicondylitis (tennis elbow); PTT=tendinopathy of the posterior tibialis tendon.
405

Factors underlying soft tissue injury

elements function within the broader matrix remodeling pathway and are activated in response
to mechanical loading to allow the tissue to remodel, heal, and adapt. It is hypothesized that
dysregulation of this pathway may negatively impact the tissue thereby increasing the likelihood
of injury.
The ECM glycoprotein tenascin-​C (TNC) is abundantly expressed in tissues experiencing
high tensile and compressive loads and plays a role in regulating cell–​matrix interactions (33).
Mokone et al. (2005) reported a GT dinucleotide repeat polymorphism in intron 17 of the
gene modulated the risk of Achilles tendon injuries (42) (Table 28.2). Following on from this,
Saunders et al. (2013) identified the G-​C-​A inferred haplotype of two TNC SNPs (rs13321
G/​C and rs2104772 T/​A) and the neighboring COL27A1 rs946053 (G/​T) SNP to increase
risk of Achilles tendinopathy in two independent study groups (64). More recently, Kluger and
colleagues (2016) found the rs13321 and rs2104772 polymorphisms were also associated with
degenerative rotator cuff tendinopathy (RCT) (33). In total, the authors identified 15 SNPs,
within a region of the TNC gene, which were associated with susceptibility to RCT, of which
six polymorphisms remained significantly associated after Bonferroni correction: rs1138545 (C/​
T), rs3789870 (C/​T), rs10759753 (T/​C), rs72758637 (G/​C), rs7021589 (A/​G), and rs7035322
(G/​T) (33). Subsequent analysis of the same variants identified the novel C-​A-​G (rs1138545 C/​
T, rs2104772 A/​T, rs10759752 A/​G) haplotype to be associated with a failure to heal after RCT
(34). These results are therefore highlighting a genomic interval within the TNC gene which
requires a more comprehensive examination.
Proteoglycans are structurally and functionally integral to tendon and ligament tissues by
providing resistance to compressive forces, regulating fibrillogenesis, and contributing to matrix
remodeling through various interactions with ECM components (76). Up until now, the genes
encoding aggrecan (ACAN), biglycan (BGN), decorin (DCN), and lumican (LUM) have only
been investigated with susceptibility to ACL ruptures (41) (Table 28.2). Independent associ-
ations were observed for the ACAN rs1516797 (T/​G) (13, 41) and DCN rs516115 (T/​G) SNPs
with the rs1516797 T allele and the rs516115 GG genotype decreasing susceptibility to ACL
ruptures. Interestingly, the association with DCN was limited to female participants and several
other studies have also reported sex-​specific associations (46, 51, 54). In addition, the authors
also highlighted several genomic regions for further interrogation across the ACAN, BGN, and
the LUM-​DCN genes, through inferred haplotype analysis (41).
The MMPs are a family of endopeptidases catalyzing a broad spectrum of matrix
components. Functional polymorphisms within the MMP1 and MMP8 promoters were
implicated in tendinopathy of the posterior tibialis tendon (PTT) (21) (Table 28.2). Specifically,
the rs1799750 (1G/​2G) 2G and rs1144393 (A/​G) G alleles, as well as a haplotype of the two
SNPs, conferred an increased risk of PTT tendinopathy (6, 21). The rs1799750 2G allele was
also associated with a risk of rotator cuff tears (3). Additionally, the MMP8 rs11225395 (C/​T)
TT genotype and T allele were significantly associated with an increased risk of developing
PTT tendinopathy (20).
Several variants within the MMP3 gene (rs679620 A/​G, rs591058 T/​C and rs650108 A/​
G) have been implicated in injury phenotypes. Raleigh et al. (2009) reported the rs679620
GG, rs591058 CC, and rs650108 AA genotypes were associated with an increased risk of
Achilles tendinopathy while the inferred A-​T-​G haplotype, comprising the alternate alleles,
was associated with decreased risk (58). The functional rs3025058 (5A/​6A) polymorphism has
also been implicated in Achilles tendinopathy (18), rotator cuff tears (3), and ACL ruptures
(39). Additionally, an inferred haplotype spanning the MMP10–​MMP1–​MMP3–​MMP12
gene cluster, localized to chromosome 11q22, was associated with a decreased risk of ACL
ruptures (49). There is currently limited research on the role of the ADAMs (a disintegrin

405
 newgenrtpdf
406
Table 28.2 Genetic variants within genes encoding regulatory components of the extracellular matrix and the associated musculoskeletal soft tissue injuries in which
they were implicated

Gene and Encoded protein Protein function Polymorphism Location Soft tissue injury Reference
chromosomal
location

TNC (9q33) Tenascin-​C glycoprotein Regulates cell–​matrix interactions G-​T tandem repeat Intron 17 AT 42
rs1330363 (G/​A) Intron 15 ATa 64
rs2104772 (T/​A) Exon 17
rs13321 (G/​C) Exon 24
rs1138545 (G/​A) Exon 10 RCTb 33
rs3789870 (C/​T) Intron 10
rs7021589 (A/​G) Intron 17
rs10759753 (T/​C) Intron 19
rs72758637 (G/​C) Intron 19
rs7035322 (G/​T) Intron 28
FBN2 Fibrillin-​2 Component of connective tissue rs331079 (G/​T) Intron 7 ACL, AT 27
(5q23-​q31) microfibrils
MIR608 MicroRNA 608 Regulation of COL5A1 mRNA rs4919510 (C/​G) 3ʹ-​UTR ATP 1, 10
(10q24) stability
ACAN (15q26) Aggrecan Major proteoglycan of articular rs2351491 (C/​T) Exon 11 ACLa 41
cartilage –​contributes to load-​ rs1042631 (C/​T)
bearing properties of cartilage rs1516797 (T/​G) Intron 12 ACLb
BGN (Xq28) Biglycan Involved in collagen rs1126499 (C/​T) Exon 4 ACLa 41
fibrillogenesis rs1042103 (G/​A) Exon 8
DCN (12q21) Decorin rs13312816 (C/​T) Intron 1 ACLa 41
rs516115 (T/​G) Intron 3 ACLb
LUM (12q21) Lumican rs2268578 (T/​C) Intron 3 ACLa 41
 newgenrtpdf
407
MMP1 (11q22) Matrix metalloproteinase 1 Degradation of collagenous rs1799750 (1G/​2G) Promoter ACLa, PTTb, RCT 3, 6, 21, 49
and noncollagenous ECM rs1144393 (A/​G) PTTb 6
MMP3 (11q22) Matrix metalloproteinase 3 components rs3025058 (5A/​6A) Promoter ACL, ATb 18, 39
rs679620 (A/​G) Exon 2 ACLa, ATPb 18, 28, 49, 58
rs591058 (T/​C) Intron 4 ATPb 18, 58
rs650108 (A/​G) Intron 8
MMP8 (11q22) Matrix metalloproteinase 8 rs11225395 (C/​T) Promoter PTT 20
MMP10 (11q22) Matrix metalloproteinase 10 rs486055 (C/​T) Exon 1 ACLa 49
MMP12 (11q22) Matrix metalloproteinase 12 rs2276109 (A/​G) Promoter ACLb 49
TIMP2 (17q25) Metalloproteinase inhibitor 2 Essential regulator of ECM rs4789932 (C/​T) Promoter ATP 26, 28
turnover and remodeling

Associations denoted with an a indicate the variant was associated as part of an inferred haplotype or an inferred allele combination. Associations denoted with a b
indicate the variant was independently associated with the musculoskeletal soft tissue injury, as well as part of an inferred haplotype or an inferred allele combination.
ACL=anterior cruciate ligament ruptures; AT=Achilles tendinopathy; ATP=Achilles tendon pathology; ECM=extracellular matrix; PTT=tendinopathy of the
posterior tibialis tendon; RCT=rotator cuff tendinopathy.
408

M. Rahim et al.

and metalloproteinase), ADAMTS (ADAM with thrombospondin motifs), and TIMPs (tissue
inhibitors of metalloproteinases) in injury risk (26, 28).
September et al. (2011) investigated the interleukin genes (interleukin-​1β (IL-​1B), interleukin-​
1 receptor antagonist (IL-​1RN) and interleukin-​6 (IL-​6)) with risk of Achilles tendinopathy
(Table 28.3). Although no independent associations were noted, the variants were observed
to collectively modulate the risk of Achilles tendinopathy in combination with the previously
associated COL5A1 rs12722 SNP (67). Similarly, a subsequent study collectively implicated
polymorphisms within the IL-​1B, IL-​6, IL-​6R, and COL5A1 genes with risk of ACL ruptures
(57). These two inferred haplotypes are therefore very similar and only differ at the A2 allele
implicated for the IL-​1RN rs2234663 VNTR locus. Although both studies only investigated
relatively small sample sizes, the similarities between the associations are highlighting a poten-
tially important biological interacting network underpinning both acute and chronic muscu-
loskeletal soft tissue injuries which needs exploration (57). Several independent and haplotype
associations have been reported for the VEGFA and KDR genes (54–​56, 61) functioning in the
angiogenesis pathway, the CASP8 gene involved in apoptosis (10, 45, 57), and the GDF5 gene
involved in regulating cell growth and differentiation (48) (Table 28.3). Additional research is
necessary to understand the full spectrum of signaling molecules and the manner in which they
may potentially impact ECM remodeling.

Hypothesis-​free approach
Traditionally, genetics research in the field of musculoskeletal soft tissue injuries has followed
a candidate gene approach (19). However, it is necessary to consider the limitations of such
an approach (37). For example, the majority of these studies are conducted in relatively small
cohorts and therefore, the reported associations are not always reproducible or able to detect
genetic contributions of small effect sizes (odds ratio <2.0) (37). It should, however, be noted that
in many of these studies careful attention was given to include well-​defined cases and controls.
As a result, these studies have been able to add to the body of knowledge in understanding risk
and are advantageous in locating genomic intervals of interest, especially when the effect sizes
have been relatively large, allowing subsequent research to follow a more directed approach.
A novel approach to explore tendinopathy, employed by Saunders et al. (2016), was the
application of bioinformatics tools to data-​mine existing databases for interactions and bio-
logical relationships to increase the probability of identifying novel and unobvious candidate
genes. The authors used the BioOntological Relationship Graph database (BORG) to inte-
grate multiple sources of genomic and biomedical knowledge, including human, mouse, and rat
orthologs, and following further prioritization were able to identify four strong candidate genes
for investigation (62).
More recently, studies are now opting for an unbiased, hypothesis-​free approach employing
GWAS or NGS methods such as whole-​exome and whole-​genome sequencing (WES/​WGS).
Although this research is still in its infancy and the detection and interpretation of these loci
may not be as straightforward, these studies can lead to new discoveries on the genetic and
biological contributions to complex phenotypes (73). Moreover, the findings may deliver new
insight into previously reported associations. However, the main disadvantage to these studies,
at present, is the lack of large, well-​defined cohorts for investigation of soft tissue injuries (74).
Consequently, public databases are often utilized and due to the inclusion of these heteroge-
neous cohorts, particularly for the controls, these studies can still produce negative results.
The first GWAS investigating susceptibility to soft tissue injuries was undertaken in a
canine ACL rupture model which identified 99 loci as risk variants for ACL ruptures (5). The

408
 newgenrtpdf
409
Table 28.3 Genetic variants within genes encoding cell signaling molecules and the associated musculoskeletal soft tissue injuries in which they were implicated

Gene and Encoded protein Protein function Polymorphism Location Soft tissue Reference
chromosomal location injury

IL-​1B (2q14) Interleukin-​1β Role in the inflammatory pathway and rs16944 (T/​C) Promoter ACLb, ATa, 57, 67
ECM degradation rs1143627 (C/​T)
IL-​1RN (2q14) Interleukin-​1 receptor Antagonist for IL-​1α and IL-​1β rs2234663 VNTR Intron 2 ACLa, ATa 57, 67
antagonist
IL-​6 (1q21) Interleukin-​6 Role in apoptosis and the rs1800795 (G/​C) Promoter ACLa, ATa 57, 67
inflammatory pathway
IL-​6R (1q21) Interleukin-​6 receptor Receptor for IL-​6 rs2228145 (A/​C) Exon 9 ACLa 57
VEGFA (6p21) VEGF-​A isoform Essential regulator of angiogenesis rs699947 (C/​A) Promoter ACLb, ATPb 54, 56
rs1570360 (G/​A) Promoter
rs2010963 (G/​C) 3ʹ-​UTR
KDR (4q11–​4q12) Kinase insert-​domain receptor Receptor for VEGF and mediates rs2071559 (G/​A) Promoter ACLb, ATa 54–​56, 61
VEGF signaling rs2305948 (G/​A) Exon 7 ATb
rs1870377 (T/​A) Exon 11 ACLb, ATa
GDF5 (20q11) Growth differentiation factor Regulates cell growth and rs143383 (T/​C) 5ʹ-​UTR ATP 48
5 differentiation
CASP8 (2q33–​q34) Caspase 8 Initiator caspase rs3834129 (ins/​del) Promoter ACLb, ATb 10, 45, 57
rs1045485 (G/​C) Exon 9

Associations denoted with an a indicate the variant was associated as part of an inferred haplotype or an inferred allele combination. Associations denoted with a b
indicate the variant was independently associated with the musculoskeletal soft tissue injury, as well as part of an inferred haplotype or an inferred allele combination.
ACL: anterior cruciate ligament ruptures; AT: Achilles tendinopathy; ATP: Achilles tendon pathology; ECM: extracellular matrix;VEGF: vascular endothelial growth
factor.
410

M. Rahim et al.

highlighted variants implicated biological networks regulating aggrecan signaling, cellular pro-
liferation, and membrane transport proteins involved in functions such as signal transduction
and pH regulation for further investigation (5). Subsequently, Kim et al. (2017) conducted a
genome-​wide screen for ACL ruptures, Achilles tendinopathy, and Achilles tendon ruptures in
a human model (31). In contrast to the canine model, no significant associations were observed
though this may be attributable to a poor study design. For instance, care was not taken to
ensure well-​defined cases and controls. Specifically, the GWAS cohort included patients irre-
spective of physical activity level whereas the candidate gene studies typically include physically
active individuals with no history of previous tendon or ligament injury. Three independent
studies investigating shoulder dislocations identified a total of five genetic markers for risk using
a GWAS approach (29, 60, 70). A genome-​wide screen for ankle injuries, including sprains,
strains, and other joint derangements, reported two loci as contributors (30). These studies also
share the same limitations due to study design.
NGS technologies have revolutionized genomic research for complex phenotypes but, to
date, only one paper has been published employing NGS technologies to explore suscepti-
bility to musculoskeletal soft tissue injuries. Caso et al. (2016) used WES analysis on twin sib-
ling males surgically diagnosed with ACL ruptures and their nonaffected progenitors (12). The
authors reported a set of 11 new variants shared by the family members which were associated
with noncontact ACL ruptures and which they hypothesized may contribute to homeostatic
imbalance of the ECM (12). As evidenced from the literature, considerable research is still
needed in this sphere to decode the genetic architecture of musculoskeletal soft tissue injuries.
Collectively, the results of the candidate gene, bioinformatics, GWAS, and WES studies add to
the evidence recognizing musculoskeletal soft tissue injuries as complex, polygenic conditions.
Additional studies are necessary to understand and validate the biological significance of these
loci. Nevertheless, these findings are an important first step in unravelling the genetic contribu-
tion to soft tissue injury risk.
Researchers also need to consider applying a broader approach incorporating genomics,
transcriptomics, epigenomics, and proteomics to provide a more complete understanding of
injury predisposition. Li et al. (2017) identified 39 differentially expressed microRNAs in ACL
tissues from patients with osteoarthritis (36). The dysregulated microRNAs were predicted
to interact with target genes involved in several essential biological processes including car-
tilage development and remodeling, collagen biosynthesis and degradation, and ECM homeo-
stasis. Trancriptome-​wide gene expression analysis of ACL tissue from time of injury identified
numerous differentially expressed transcripts representing important biological processes (9).
Studies such as these provide insight into the healing capacity of ruptured ACLs and may be par-
ticularly relevant to the development of effective ACL repair and reconstruction techniques (9).

Clinical relevance
The precise mechanisms culminating in injury are yet to be elucidated and therefore any pre-
dictive or diagnostic, as well as injury-​r isk tests available at present are largely premature (16).
Despite this, there are a growing number of direct-​to-​consumer genetic tests available on
the market without offering the appropriate counseling or clarification on the results (75).
Moreover, these tests do not take into consideration the multifactorial nature of sporting
injuries and are therefore more than likely to interpret the results incorrectly (16, 65). In fact,
the primary objective in elucidating the genetic contribution to musculoskeletal soft tissue
injuries is to identify the biological pathways and molecular mechanisms contributing to patho-
genesis warranting subsequent in-​depth exploration. Only upon validation of these genetic and

410
411

Factors underlying soft tissue injury

molecular markers in larger samples within independent studies will it be meaningful to design
possible genetic tests examining an individual’s injury susceptibility. In particular, insight into
how the fine balance between matrix synthesis and degradation is regulated and dysregulated
is required to inform new therapeutic interventions and identify potential biological targets.
Although predictive genetic testing is not possible for these injuries, a genetic test included
in a holistic risk assessment tool may have future applications in the design of prehabilitation
strategies for genetically predisposed individuals or in analyzing how an individual is likely
to respond to load. Moreover, one’s genetic makeup may aid in the prescription of appro-
priate treatment strategies after injury and to inform decision-​making on operative versus
nonoperative injury management.

Future work
The main limitations to the current evidence are that: 1) the majority of studies have investigated
relatively small samples, 2) the findings have not been replicated in independent populations,
and 3) the studies were primarily conducted in Caucasian populations and therefore the results
are not necessarily representative of populations across the world. As a result, additional research
in large, independent sample sets is ultimately required to confirm the reported genetic asso-
ciations. The establishment of international consortia is therefore essential to achieving this
goal by pooling knowledge and resources to overcome the current barriers (47, 74). In doing
so, researchers will be also able to undertake GWAS and NGS approaches in large, carefully
phenotyped cohorts of cases and controls which may aid in the discovery of rare variants with
large effect sizes on modulating injury risk, and assist in refining the genomic regions to which
these variants cluster.
At present, there is limited functional evidence to support the current genetic associations
(24, 35). The results from genetic studies cannot be understood in isolation. Rather these
results need to pave the direction of future research, with the proposed markers interrogated
using a multidisciplinary approach, including omics technologies, RNA sequencing analysis,
differential gene expression, and protein studies to establish their role in injury susceptibility
(12, 65). In addition, the clinical relevance of these biological markers also needs investiga-
tion (9, 14, 33, 71).

Conclusion
Today, genetics is increasingly recognized as a key risk factor predisposing individuals to an
increased risk of ligament and tendon injuries.The genes implicated thus far encode an array of
proteins essential to maintaining tissue homeostasis including collagens, noncollagenous struc-
tural proteins, and regulators of the ECM. Musculoskeletal soft tissue injuries are likely a result
of disturbance to biological networks and not isolated proteins and genes and therefore the
interactions between proteins functioning in common biological pathways are intriguing and
should also be explored (5). In addition, recent research from bioinformatics analysis, GWAS
and WES analysis are proposing novel candidates and biological pathways for examination.
Research in this domain will continue to provide insight into the genetic, molecular, and cel-
lular mechanisms that may potentially serve as therapeutic targets. Thus, it is critical that large
data sets are collected and international consortia are established to effectively pool resources
and a multidisciplinary approach utilized to understand the biological significance of these gen-
etic loci in contributing to musculoskeletal soft tissue injury risk susceptibility.

411
412

M. Rahim et al.

References
1. Abrahams Y, Laguette M-​J, Prince S, Collins M. Polymorphisms within the COL5A1 3’-​UTR
that alters mRNA structure and the MIR608 gene are associated with Achilles tendinopathy. Ann
Hum Genet 77: 204–​214, 2013.
2. Altinisik J, Meric G, Erduran M, Ates O, Ulusal AE, Akseki D. The BstUI and DpnII variants
of the COL5A1 gene are associated with tennis elbow. Am J Sports Med 43: 1784–​1789, 2015.
3. Assunção JH, Godoy-​Santos AL, dos Santos MCLG, Malavolta EA, Gracitelli MEC, Ferreira
Neto AA. Matrix metalloproteases 1 and 3 promoter gene polymorphism is associated with rotator
cuff tear. Clin Orthop Relat Res 475: 1904–​1910, 2017.
4. Bahr R, Krosshaug T. Understanding injury mechanisms: a key component of preventing injuries
in sport. Br J Sports Med 39: 324–​329, 2005.
5. Baker LA, Kirkpatrick B, Rosa GJM, Gianola D, Valente B, Sumner JP, Baltzer W, Hao
Z, Binversie EE, Volstad N, Piazza A, Sample SJ, Muir P. Genome-​wide association ana-
lysis in dogs implicates 99 loci as risk variants for anterior cruciate ligament rupture. PLoS One
12: e0173810, 2017.
6. Baroneza JE, Godoy-​Santos A, Ferreira Massa B, Boçon de Araujo Munhoz F, Diniz
Fernandes T, Leme Godoy Dos Santos MC. MMP-​1 promoter genotype and haplotype asso-
ciation with posterior tibial tendinopathy. Gene 547: 334–​337, 2014.
7. Behrens G, Fischer B, Kohler S, Park Y, Hollenbeck AR, Leitzmann MF. Healthy lifestyle
behaviors and decreased risk of mortality in a large prospective study of U.S. women and men. Eur J
Epidemiol 28: 361–​372, 2013.
8. Bere T, Flørenes TW, Krosshaug T, Koga H, Nordsletten L, Irving C, Muller E, Reid RC,
Senner V, Bahr R. Mechanisms of anterior cruciate ligament injury in World Cup alpine skiing: a
systematic video analysis of 20 cases. Am J Sports Med 39: 1421–​1429, 2011.
9. Brophy RH, Tycksen ED, Sandell LJ, Rai MF. Changes in transcriptome-​wide gene expression
of anterior cruciate ligament tears based on time from injury. Am J Sports Med 44: 2064–​2075, 2016.
10. Brown KL, Seale KB, El Khoury LY, Posthumus M, Ribbans WJ, Raleigh SM, Collins M,
September AV. Polymorphisms within the COL5A1 gene and regulators of the extracellular
matrix modify the risk of Achilles tendon pathology in a British case-​control study. J Sports Sci
35: 1475–​1483, 2017.
11. Burger M, de Wet H, Collins M. The COL5A1 gene is associated with increased risk of carpal
tunnel syndrome. Clin Rheumatol 34: 767–​774, 2015.
12. Caso E, Maestro A, Sabiers CC, Godino M, Caracuel Z, Pons J, Gonzalez FJ, Bautista
R, Claros MG, Caso-​Onzain J, Viejo-​Allende E, Giannoudis P V, Alvarez S, Maietta P,
Guerado E. Whole-​exome sequencing analysis in twin sibling males with an anterior cruciate liga-
ment rupture. Injury 47: S41–​S50, 2016.
13. Cięszczyk P, Willard K, Gronek P, Zmijewski P, Trybek G, Gronek J, Weber-​Rajek M,
Stastny P, Petr M, Lulińska-​Kuklik E, Ficek K, Kemeryte-​Riaubiene E, Maculewicz E,
September AV. Are genes encoding proteoglycans really associated with the risk of anterior cru-
ciate ligament rupture? Biol Sport 2: 97–​103, 2017.
14. Collins M, Posthumus M. Type V collagen genotype and exercise-​related phenotype relationships.
Exerc Sport Sci Rev 39: 191–​198, 2011.
15. Collins M, Posthumus M, Schwellnus MP. The COL1A1 gene and acute soft tissue ruptures. Br
J Sports Med 44: 1063–​1064, 2010.
16. Collins M, September AV, Posthumus M. Biological variation in musculoskeletal injuries: current
knowledge, future research and practical implications. Br J Sports Med 49: 1497–​1503, 2015.
17. Ficek K, Cieszczyk P, Kaczmarczyk M, Maciejewska-​Karłowska A, Sawczuk M, Cholewinski
J, Leonska-​Duniec A, Stepien-​Slodkowska M, Zarebska A, Stepto NK, Bishop DJ, Eynon
N. Gene variants within the COL1A1 gene are associated with reduced anterior cruciate ligament
injury in professional soccer players. J Sci Med Sport 16: 396–​400, 2013.
18. Gibbon A, Hobbs H, van der Merwe W, Raleigh SM, Cook J, Handley CJ, Posthumus M,
Collins M, September AV. The MMP3 gene in musculoskeletal soft tissue injury risk profiling: a
study in two independent sample groups. J Sports Sci 35: 655–​662, 2017.
19. Gibson WT. Genetic association studies for complex traits: relevance for the sports medicine prac-
titioner. Br J Sports Med 43: 314–​316, 2009.

412
413

Factors underlying soft tissue injury

20. Godoy-​Santos, A, Ortiz RT, Junior RM, Fernandes TD, Santos MCLG. MMP-​8 poly-
morphism is genetic marker to tendinopathy primary posterior tibial tendon. Scand J Med Sci Sport
24: 220–​223, 2014.
21. Godoy-​Santos A, Cunha M V, Ortiz RT, Fernandes TD, Mattar R, dos Santos MCLG.
MMP-​1 promoter polymorphism is associated with primary tendinopathy of the posterior tibial
tendon. J Orthop Res 31: 1103–​1107, 2013.
22. Hay M, Patricios J, Collins R, Branfield A, Cook J, Handley CJ, September AV, Posthumus
M, Collins M. Association of type XI collagen genes with chronic Achilles tendinopathy in inde-
pendent populations from South Africa and Australia. Br J Sports Med 47: 569–​574, 2013.
23. Järvinen TAH, Kannus P, Maffulli N, Khan KM. Achilles tendon disorders: etiology and epidemi-
ology. Foot Ankle Clin 10: 255–​266, 2005.
24. Jin H, van’t Hof RJ, Albagha OME, Ralston SH. Promoter and intron 1 polymorphisms of
COL1A1 interact to regulate transcription and susceptibility to osteoporosis. Hum Mol Genet
18: 2729–​2738, 2009.
25. Khoschnau S, Melhus H, Jacobson A, Rahme H, Bengtsson H, Ribom E, Grundberg E,
Mallmin H, Michaëlsson K. Type I collagen alpha1 Sp1 polymorphism and the risk of cruciate
ligament ruptures or shoulder dislocations. Am J Sports Med 36: 2432–​2436, 2008.
26. El Khoury L, Posthumus M, Collins M, Handley CJ, Cook J, Raleigh SM. Polymorphic vari-
ation within the ADAMTS2, ADAMTS14, ADAMTS5, ADAM12 and TIMP2 genes and the risk of
Achilles tendon pathology: a genetic association study. J Sci Med Sport 16: 493–​498, 2013.
27. El Khoury L, Posthumus M, Collins M, van der Merwe W, Handley C, Cook J, Raleigh S.
ELN and FBN2 gene variants as risk factors for two sports-​related musculoskeletal injuries. Int J
Sports Med 36: 333–​337, 2014.
28. El Khoury L, Ribbans WJ, Raleigh SM. MMP3 and TIMP2 gene variants as predisposing factors
for Achilles tendon pathologies: attempted replication study in a British case-​control cohort. Meta
Gene 9: 52–​55, 2016.
29. Kim S, Kleimeyer J, Ahmed M, Avins A, Fredericson M, Dragoo J, Ioannidis J. A genetic
marker associated with shoulder dislocation. Int J Sports Med 38: 508–​514, 2017.
30. Kim SK, Kleimeyer JP, Ahmed MA, Avins AL, Fredericson M, Dragoo JL, Ioannidis JPA.
Two genetic loci associated with ankle injury. PLoS One 12: e0185355, 2017.
31. Kim SK, Roos TR, Roos AK, Kleimeyer JP, Ahmed MA, Goodlin GT, Fredericson M,
Ioannidis JPA, Avins AL, Dragoo JL. Genome-​wide association screens for Achilles tendon and
ACL tears and tendinopathy. PLoS One 12: e0170422, 2017.
32. Kitsios GD, Zintzaras E. Genome-​wide association studies: hypothesis-​”free” or “engaged”? Transl
Res 154: 161–​164, 2009.
33. Kluger R, Burgstaller J, Vogl C, Brem G, Skultety M, Mueller S. Candidate gene approach
identifies six SNPs in tenascin-​C (TNC) associated with degenerative rotator cuff tears. J Orthop Res
35: 894–​901, 2017.
34. Kluger R, Huber KR, Seely PG, Berger CE, Frommlet F. Novel tenascin-​C haplotype modifies
the risk for a failure to heal after rotator cuff repair. Am J Sports Med 45: 2955–​2964, 2017.
35. Laguette M-​J, Abrahams Y, Prince S, Collins M. Sequence variants within the 3’-​UTR of the
COL5A1 gene alters mRNA stability: implications for musculoskeletal soft tissue injuries. Matrix Biol
30: 338–​345, 2011.
36. Li B, Bai L, Shen P, Sun Y, Chen Z, Wen Y. Identification of differentially expressed micrornas in
knee anterior cruciate ligament tissues surgically removed from patients with osteoarthritis. Int J Mol
Med 40: 1105–​1113, 2017.
37. Little J, Higgins JPT, Ioannidis JP a., Moher D, Gagnon F, von Elm E, Khoury MJ, Cohen
B, Davey-​Smith G, Grimshaw J, Scheet P, Gwinn M, Williamson RE, Zou GY, Hutchings
K, Johnson CY, Tait V, Wiens M, Golding J, van Duijn C, McLaughlin J, Paterson A, Wells
G, Fortier I, Freedman M, Zecevic M, King R, Infante-​Rivard C, Stewart A, Birkett N.
STrengthening the REporting of Genetic Association studies (STREGA) –​an extension of the
STROBE statement. Eur J Clin Invest 39: 247–​266, 2009.
38. Lulińska-​Kuklik E, Rahim M, Domańska-​Senderowska D, Ficek K, Michałowska-​
Sawczyn M, Moska W, Kaczmarczyk M, Brzeziański M, Brzeziańska-​Lasota E, Cieszczyk
P, September AV. Interactions between COL5A1 gene and risk of the anterior cruciate ligament
­rupture. J Hum Kinet 62: 65–​71, 2017.

413
414

M. Rahim et al.

39. Malila S, Yuktanandana P, Saowaprut S, Jiamjarasrangsi W, Honsawek S. Association between

matrix metalloproteinase-​3 polymorphism and anterior cruciate ligament ruptures. Genet Mol Res
10: 4158–​4165, 2011.
40. Mann V, Hobson EE, Li B, Stewart TL, Grant SFA, Robins SP, Aspden RM, Ralston SH.
A COL1A1 Sp1 binding site polymorphism predisposes to osteoporotic fracture by affecting bone
density and quality. J Clin Invest 107: 899–​907, 2001.
41. Mannion S, Mtintsilana A, Posthumus M, van der Merwe W, Hobbs H, Collins M, September
AV. Genes encoding proteoglycans are associated with the risk of anterior cruciate ligament ruptures.
Br J Sports Med 48: 1640–​1646, 2014.
42. Mokone GG, Gajjar M, September A V, Schwellnus MP, Greenberg J, Noakes TD, Collins
M. The guanine-​thymine dinucleotide repeat polymorphism within the tenascin-​C gene is associated
with Achilles tendon injuries. Am J Sports Med 33: 1016–​1021, 2005.
43. Mokone GG, Schwellnus MP, Noakes TD, Collins M. The COL5A1 gene and Achilles tendon
pathology. Scand J Med Sci Sports 16: 19–​26, 2006.
44. Morrow JR, Defina LF, Leonard D, Trudelle-​Jackson E, Custodio MA. Meeting physical activity
guidelines and musculoskeletal injury: the WIN study. Med Sci Sports Exerc 44: 1986–​1992, 2012.
45. Nell E-​M, van der Merwe L, Cook J, Handley CJ, Collins M, September AV. The apoptosis
pathway and the genetic predisposition to Achilles tendinopathy. J Orthop Res 30: 1719–​1724, 2012.
46. O’Connell K, Knight H, Ficek K, Leonska-​Duniec A, Maciejewska-​Karlowska A, Sawczuk
M, Stepien-​Slodkowska M, O’Cuinneagain D, van der Merwe W, Posthumus M, Cieszczyk
P, Collins M. Interactions between collagen gene variants and risk of anterior cruciate ligament
rupture. Eur J Sport Sci 15: 341–​350, 2015.
47. Pitsiladis YP, Tanaka M, Eynon N, Bouchard C, North KN, Williams AG, Collins M, Moran
CN, Britton SL, Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov II, de Geus E, Alsayrafi
M. Athlome Project Consortium: a concerted effort to discover genomic and other “omic” markers
of athletic performance. Physiol Genomics 48: 183–​190, 2016.
48. Posthumus M, Collins M, Cook J, Handley CJ, Ribbans WJ, Smith RKW, Schwellnus
MP, Raleigh SM. Components of the transforming growth factor-​beta family and the patho-
genesis of human Achilles tendon pathology –​a genetic association study. Rheumatology (Oxford)
49: 2090–​2097, 2010.
49. Posthumus M, Collins M, van der Merwe L, O’Cuinneagain D, van der Merwe W, Ribbans
WJ, Schwellnus MP, Raleigh SM. Matrix metalloproteinase genes on chromosome 11q22 and the
risk of anterior cruciate ligament (ACL) rupture. Scand J Med Sci Sport 22: 523–​533, 2012.
50. Posthumus M, September AV, Keegan M, O’Cuinneagain D, Van der Merwe W, Schwellnus
MP, Collins M. Genetic risk factors for anterior cruciate ligament ruptures: COL1A1 gene variant.
Br J Sports Med 43: 352–​356, 2009.
51. Posthumus M, September AV, O’Cuinneagain D, van der Merwe W, Schwellnus MP, Collins
M. The COL5A1 gene is associated with increased risk of anterior cruciate ligament ruptures in
female participants. Am J Sports Med 37: 2234–​2240, 2009.
52. Posthumus M, September AV, O’Cuinneagain D, van der Merwe W, Schwellnus MP,
Collins M. The association between the COL12A1 gene and anterior cruciate ligament ruptures. Br
J Sports Med 44: 1160–​1165, 2010.
53. Rahim M, Collins M, September AV. Genes and musculoskeletal soft-​tissue injuries. In: Medicine
and Sport Science, edited by Posthumus M, Collins M. Basel: Karger, p. 68–​91, 2016.
54. Rahim M, Gibbon A, Hobbs H, van der Merwe W, Posthumus M, Collins M, September
AV. The association of genes involved in the angiogenesis-​associated signaling pathway with risk of
anterior cruciate ligament rupture. J Orthop Res 32: 1612–​1618, 2014.
55. Rahim M, Hobbs H, van der Merwe W, Posthumus M, Collins M, September AV. Investigation
of angiogenesis genes with anterior cruciate ligament rupture risk in a South African population. J
Sports Sci 36: 551–​557, 2018.
56. Rahim M, El Khoury LY, Raleigh SM, Ribbans WJ, Posthumus M, Collins M, September
AV. Human genetic variation, sport and exercise medicine, and Achilles tendinopathy: role for
angiogenesis-​associated genes. Omi A J Integr Biol 20: 520–​527, 2016.
57. Rahim M, Mannion S, Klug B, Hobbs H, van der Merwe W, Posthumus M, Collins M,
September AV. Modulators of the extracellular matrix and risk of anterior cruciate ligament
ruptures. J Sci Med Sport 20: 152–​158, 2017.

414
415

Factors underlying soft tissue injury

58. Raleigh SM, van der Merwe L, Ribbans WJ, Smith RKW, Schwellnus MP, Collins M.
Variants within the MMP3 gene are associated with Achilles tendinopathy: possible interaction with
the COL5A1 gene. Br J Sports Med 43: 514–​520, 2009.
59. Reiner M, Niermann C, Jekauc D, Woll A. Long-​term health benefits of physical activity –​a sys-
tematic review of longitudinal studies. BMC Public Health 13: 813, 2013.
60. Roos TR, Roos AK, Avins AL, Ahmed MA, Kleimeyer JP, Fredericson M, Ioannidis JPA,
Dragoo JL, Kim SK. Genome-​wide association study identifies a locus associated with rotator cuff
injury. PLoS One 12: e0189317, 2017.
61. Salles JI, Duarte MEL, Guimarães JM, Lopes LR, Vilarinho Cardoso J, Aguiar DP, Machado
Neto JO, Machado DE, Perini JA. Vascular endothelial growth factor receptor-​2 polymorphisms
have protective effect against the development of tendinopathy in volleyball athletes. PLoS One
11: e0167717, 2016.
62. Saunders CJ, Jalali Sefid Dashti M, Gamieldien J. Semantic interrogation of a multi know-
ledge domain ontological model of tendinopathy identifies four strong candidate risk genes. Sci Rep
6: 19820, 2016.
63. Saunders CJ, van der Merwe L, Cook J, Handley CJ, Collins M, September AV. Extracellular
matrix proteins interact with cell-​signaling pathways in modifying risk of Achilles tendinopathy. J
Orthop Res 33: 898–​903, 2015.
64. Saunders CJ, van der Merwe L, Posthumus M, Cook J, Handley CJ, Collins M, September
AV. Investigation of variants within the COL27A1 and TNC genes and Achilles tendinopathy in two
populations. J Orthop Res 31: 632–​637, 2013.
65. September AV, Rahim M, Collins M. Towards an understanding of the genetics of
tendinopathy. In: Advances in Experimental Medicine and Biology, edited by Ackermann PW, Hart DA.
New York: Springer, p. 109–​116, 2016.
66. September AV, Cook J, Handley CJ, van der Merwe L, Schwellnus MP, Collins M. Variants
within the COL5A1 gene are associated with Achilles tendinopathy in two populations. Br J Sports
Med 43: 357–​365, 2009.
67. September AV, Nell E-​M, O’Connell K, Cook J, Handley CJ, van der Merwe L, Schwellnus M,
Collins M. A pathway-​based approach investigating the genes encoding interleukin-​1β, interleukin-​
6 and the interleukin-​1 receptor antagonist provides new insight into the genetic susceptibility of
Achilles tendinopathy. Br J Sports Med 45: 1040–​1047, 2011.
68. Stȩpień-​Słodkowska M, Ficek K, Eider J, Leońska-​Duniec A, MacIejewska-​Karłowska A,
Sawczuk M, Zarȩbska A, Jastrzȩbski Z, Grenda A, Kotarska K, Ciȩszczyk P. The +1245G/​T
polymorphisms in the collagen type I alpha 1 (COL1A1) gene in polish skiers with anterior cruciate
ligament injury. Biol Sport 30: 57–​60, 2013.
69. Stępień-​Słodkowska M, Ficek K, Maciejewska-​Karłowska A, Sawczuk M, Ziętek P, Król P,
Zmijewski P, Pokrywka A, Cięszczyk P. Overrepresentation of the COL3A1 AA genotype in
Polish skiers with anterior cruciate ligament injury. Biol Sport 32: 143–​147, 2014.
70. Tashjian RZ, Granger EK, Farnham JM, Cannon-​Albright LA, Teerlink CC. Genome-​wide
association study for rotator cuff tears identifies two significant single-​nucleotide polymorphisms. J
Shoulder Elb Surg 25: 174–​179, 2016.
71. Tashjian RZ, Granger EK, Zhang Y, Teerlink CC, Cannon-​Albright LA. Identification of a
genetic variant associated with rotator cuff repair healing. J Shoulder Elb Surg 25: 865–​872, 2016.
72. Vaughn NH, Stepanyan H, Gallo RA, Dhawan A. Genetic factors in tendon injury: a systematic
review of the literature. Orthop J Sport Med 5: 2325967117724416, 2017.
73. Visscher PM, Brown MA, McCarthy MI, Yang J. Five years of GWAS discovery. Am J Hum Genet
90: 7–​24, 2012.
74. Wang G, Tanaka M, Eynon N, North KN, Williams AG, Collins M, Moran CN, Britton SL,
Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov II, de Geus E, Alsayrafi M, Pitsiladis
YP. The future of genomic research in athletic performance and adaptation to training. In: Medicine
and Sport Science, edited by Posthumus M, Collins M. Basel: Karger, p. 55–​67, 2016.
75. Williams AG, Wackerhage H, Day SH. Genetic testing for sports performance, responses to
training and injury risk: practical and ethical considerations. In: Medicine and Sport Science, edited by
Posthumus M, Collins M. Basel: Karger, p. 105–​119, 2016.
76. Yanagishita M. Function of proteoglycans in the extracellular matrix. Acta Pathol Jpn
43: 283–​293, 1993.

415
416

29
SEX AND
PERFORMANCE: NATURE
VERSUS NURTURE
Mindy Millard-​Stafford and Matthew T. Wittbrodt

Ever since the first Olympiad in the modern era (1896), sport is recognized to play a relevant
societal role to promote education, health, intercultural dialogue, and the development of indi-
viduals –​regardless of sex, race, age, gender orientation, and socioeconomic background (5).
Conversely, sport performance is also affected by societal influences as one of many factors
that ultimately coalesce and contribute to athletic success. These factors could generally be
categorized as either predominantly inherent (i.e., nature) and/​or environmental (i.e., nurture).
It is well accepted that elite athletes are endowed with exceptional genetic potential but require
the appropriate deliberate sport-​specific practice/​training and environment to develop these
talents (14, 55). The relative contribution of nature to sport performance is likely above 50%
but less than 100% (14). Figure 29.1 presents a general theoretical model that considers many
of these factors influencing sport performance without basis for the relative weighting of each.
Regarding those of primarily genetic origin, sex is clearly one fundamental genetic factor that
dictates performance, particularly in sports objectively measured by time, distance, or weight.
The term “sex” differences relative to the discussion in this chapter is based upon the bio-
logical genetic difference between men and women rather than the term “gender” differences.
A far wider concept than biological sex, “gender” implies life conditions, cultural and societal
expectations about femininity/​masculinity, and fundamentally one’s own sense of self. As will
be discussed elsewhere (Chapter 34), sex is recognized as not a true binary classification due to
rare conditions of sexual development (30), which present increasingly complex challenges in
the contemporary world of sport competition.
Considered as the “weaker sex,” creating an equal and level competitive playing field has
been the basis for classifying most sports by sex. In fact, there are few sports today (eques-
trian, sailing) where men directly compete against women in the same classification. There
are separate divisions (and length of matches) at Wimbledon, and distinct competitions held
for bicycle stage races and FIFA World Cups. EA Sports only started to provide female soccer
players for girls (and boys) to select from when playing video games in 2015. The “fairer” sex
has been segregated from most sport competitions for a long time. Is this basis solely due to
biological and/​or cultural advantages for men? Historical traditions? A quote from the famous
marathoner, Grete Waitz (world record holder 1979–​1983) epitomizes this concept: “As long as
women are women, I don’t think they will surpass men.” Some of these complex issues affecting

416
417

Sex and performance: Nature vs. nurture

PHYSIOLOGICAL NEUROMUSCULAR PSYCHOLOGICAL

NATURE NURTURE

Sex SPORT Interventions:

Age PERFORMANCE Coaching
Physical /Anatomical Training
Characteristics Nutrition
Performance Genes Societal Influences:
Cultural Acceptance
Motivation/Rewards

Figure 29.1 A theoretical model for contributing factors impacting sport performance.

our understanding of “sex differences” in sport performance must also be considered when
discussing the tension between the biological versus environmental factors influencing the sex
difference in sports performance.

Historical perspectives
Since the modern era of the Olympic Games, women have been playing “catch-​up” to men in
terms of societal acceptance for female participation and the available opportunities for sport
competition. The first Olympiad (1896) was open only to men. However, at the 1900 Paris
Olympics, 22 women participated out of a total 997 athletes, representing approximately 2% of
all athletes. Women competed alongside their male counterparts in sailing, croquet, and eques-
trian. Only two sports were segregated by sex: tennis and golf. Over time, more Olympic sports
were gradually added: archery, swimming, and athletics (track and field) by 1928. Yet, in the
inaugural Olympic 800-​meter track event for women, several runners “reportedly” collapsed
during competition although the veracity of this report is debated (41). Officials subsequently
deemed the distance too great a call on feminine strength with grave concerns physical exhaus-
tion would “harm a women’s reproductive organs.” The 800 meters for women did not return
to the Olympic Games program for another 30 years! Ironically, this same rationale by male
officials in the International Olympic Committee continued for sports such as the women’s ski
jump until Sochi 2014.
Unfortunately, the persistent notion labeling certain athletic events as “inappropriate” for
women during most of the 20th century prevailed (e.g., women were considered “too frail” to
compete in demanding events lasting more than 1–​2 minutes). The 1500-​meter run was not a
permanent event to the Olympic Games until 1972, 76 years later than was available for men.
In the US, women were banned from official competition in the Boston marathon until 1972.
Prior to 1972, race officials would physically remove women from the course if entered under
an assumed “male” name.The women’s marathon has an even shorter Olympic history than the
1500 meters, as it was not included until the 1984 Los Angeles Games. Ironically, in that first
Olympic marathon for women (won by American Joan Benoit Samuelson), another top woman
Gabriela Andersen-​Scheiss also collapsed in front of TV cameras on the final track leg due to

417
418

M. Millard-Stafford and M.T. Wittbrodt

heat stroke. Fortunately, sport governing bodies did not repeat the mistake of eliminating the
women’s marathon based upon presumed “risk.”
In the United States, a federal law –​Title IX legislation –​passed in 1972 requiring equal
funding for federally sponsored educational programs, including sports.To date, the influence of
this policy continues to spur girls’ participation in sports (34).Yet, not until 40 years later, in the
2012 London Summer Olympic Games, would every country include a female athlete as part
of their delegation (comprising 44% of all competitors). However, coaching staffs and the upper
echelons of sports administrative bodies have not caught up (46). Prior to 2004, no female coach
had ever been a part of the US Olympic swimming staff. Furthermore, less than 14% of Division
I National Collegiate Athletic Association (NCAA) women’s swimming programs have a female
head coach as of 2015. Moreover, the International Olympic Committee has stated the goal of
women holding 20% of all administrative and decision-​making positions within national organ-
izing committees and international federations.
Thus, across many sport disciplines, women had few role models in sport and started com-
peting chronologically later than their male counterparts. The hammer throw, pole vault, and
3000-​meter steeplechase were the last athletics events introduced into the Olympics, while
women still do not compete in the 50 km racewalk (only recently recognized by the IAAF
and held for the first time in the London 2017 World Athletics Championships). Even in the
2016 Rio Olympics, the 1500 meters freestyle, the longest pool event in Olympic swimming
for men, was not a competitive distance for women (only up to 800 meters). However, this
inequality is finally slated to be rectified with the addition of the 1500 meters freestyle for
women in Tokyo 2020 (and 800 meters for men), nearly a century since swimming has been in
the Olympic Games!
In summary, over the past 40 years in the US, the environment or nurture factor for women
to pursue and compete in sport has clearly changed to level the playing field. Equal access to
training and coaching has widened (e.g., in many sports such as swimming, boys and girls even
train together up past the age group through college ranks). When a culture is acceptable for
girls to play a particular sport starting in their youth, girls are more likely to be encouraged
to pursue their interest through parents, friends, and environment to develop their skills. This
factor increases young girls’ motivation and confidence in their ability to pursue sport at the
highest level. The global impact of one country’s federal legislation, however, remains difficult
to measure. In terms of sport performance, opportunities for sport participation clearly have
an important influence on the motivation for women to devote their lives to sport, since the
Olympic Games results in financial incentives for winning medals and other “rewards” both
intrinsic and extrinsic in nature often follow (5).

Other nurture influences on the performance gap

Age is also an inherent factor contributing to sport performance (Figure 29.1). In some women’s
sports such as figure skating or gymnastics, a delay in maturation has biomechanical advantages
for younger-​aged female competitors that are not necessarily similar for men in the same sport
(3). However, not long ago, assumptions regarding the “peak” age for sprint/​power or team
sports was believed to be in the early 20s, especially for women, due solely to the fact that
women did not continue to compete beyond high school. For example in 1968, Debbie Meyer,
at 16 years of age, became the first female swimmer to win three individual gold medals in one
Olympics (200-​, 400-​, and 800-​meter freestyles), a feat not repeated until Rio 2016 by 19-​year-​
old Katie Ledecky (already appearing in her second Olympiad). Debbie Meyer seemed to be
at the “peak” of her career, but did not defend her titles and “retired” in 1972. In that era, strict

418
419

Sex and performance: Nature vs. nurture

amateurism rules prevailed and few college swimming programs, let alone athletic scholarships,
were available to continue the rigorous training required to be successful in the sport.
In terms of contributing factors that might underlie the sex differences, motivation for
athletes to stay in the sport are reflected by an increased age of elite athletes over time. An ana-
lysis of the change in US Olympic swimming and track competitors’ average age is depicted in
Figure 29.2. Male and female swimmers (combined) increased in age significantly from 1972
(18.5 ± 2.7 years) to 1992 (21.9 ± 3.0 years) and became progressively higher in 2012 (23.6
± 3.9 years). After 2000, the age for female Olympians (21 years) remained similar but males
increased (from 21 to 24.5 years) over this same period of time (Figure 29.2A).These data indi-
cate elite US male swimmers are consistently older than females but that over the past 40+ years,
age has increased for both men (to >24 years) and women (>21 years). In the track running
events (Figure 29.2B), there is no clear sex difference in the age of US Olympians although the
increased age over time remains consistent to that observed in swimming.Track competitors (up
through the marathon) also appear to be older overall compared to swimmers. Recent reports
comparing the top five places in key marathon competitions over the years indicate men and
women physiologically peak at a similar age (29–​30 years) in running performance (24). This
differs somewhat from the 2015 report (1) that 20 years old is the peak age for international
swimming in both sprint and distance events, with “little sex difference” in estimates of peak
age. However, there are increasingly more “outlier” older athletes defying published trends by
competing longer in their sport due to professional opportunities available and returning after
childbirth (e.g., 41-​year-​old 2008 Olympic silver medalist Dara Torres). Thus, a current change
in this nurture variable is reflected in the older ages of contemporary elite athletes, which
have advanced past the 1972 statistic when 17 years old was the average age for an elite female
swimmer.

When does the performance gap appear?

National youth surveys indicate girls are overall less physically active than boys, particularly
striking around adolescence (53). It would also seem logical that since girls tend to mature
earlier than boys, there would be little physical differences in sport performance prior to sexual
maturation (with biological differences gradually more pronounced due to increasing testos-
terone in boys). If a “gap” in sports performance is related primarily to biology (i.e., hormonal
changes throughout growth and maturation), then this should occur progressively in sports
where it is common for boys and girls to train together and have access to similar coaching and
league competition. In swimming, boys and girls train together in the same pool practices from
youth through adulthood with the same access to coaching often through the college ranks,
although they compete separately. In Figure 29.3, the sex difference is plotted based upon US
national age-​g roup records. For each age group of interest, performance time(s) were recorded
by sex classification across each event. The percentage sex difference in performance time was
then calculated using a pairwise comparison by rank using the following equation:

Female time − Male time

% Difference = ×100
Male time

Each dot represents the difference between matched pairs of boys and girls from 1st to 16th place
across history in USA Swimming (as of 2015), thus reducing selection bias. Any points above
zero represent a case where similarly ranked boys’ times were faster than the corresponding
girls’ times. As early as 9–​10 years old (and likely prior to puberty for boys), in more cases than

419
420

Figure 29.2 Mean (± standard deviation) age for US Olympic team members in swimming (top) and
running events in track (bottom) over time.
421

Sex and performance: Nature vs. nurture

Figure 29.3 Performance gap in elite top-​16 all-​time US age group, senior national and world record
swimmers. Differences were significant across each age group (except for 15–​16 vs. 17–​18 years and
senior national vs. world record) with progressively increasing performance gaps between boys and girls.

not, the boys seem to be somewhat faster (~1.1%), which progressively increases with each
successive 2-​year age group category. Moving to 11–​12-​year-​olds, only in the longer distances
(1500 meters) are girls sometimes faster and with each successive age group thereafter, the
performance gap widens significantly between boys and girls with the greatest discrepancy
observed in the sprints (50 meters). On average, across all events at the highest level of senior
national competition (no age limitations), the performance gap averages remarkably close to
approximately 10%. It is impossible to know the maturity of the respective boys and girls within
each chronological age bracket in this cross-​sectional data set, but typically early maturing boys
dominate the national top times at early ages (9, 10) but often those individuals do not remain
in the national ranking lists in later years. In contrast, national record holders for girls in 11–​
12 years age groups often remain in the nationally ranked lists in later age groups. Finally, unlike
the relatively similar sex difference observed across the spectrum of Olympic running events
(49), the sex difference in sprints (50 meters) was nearly double that of endurance pool events
(400 meters and above), and this inverse relationship between sex difference and swim distance
is observed as early as 11–​12 years and remains throughout growth and maturity.

Sports participation and opportunity as a nurture factor

Using data on elite athletes’ performances to understand the sex difference may be challenging
when attempting to tease out the influence of environmental or “nurture” factors. In older
master athletes, Hunter and Stevens (23) indicate greater than one-​third of the sex difference
is based on differential participation. The potential number of competitors (i.e., depth of field)
impacts record performances worldwide (9); thus, if opportunities are not equal for men and
women, the utility of using performances from isolated races or databases may be limited by
fewer females competing.

421
422

M. Millard-Stafford and M.T. Wittbrodt

Most sport governing bodies track their membership by sex and age group. In USA
Swimming, participation numbers for girls equal and, in many age groups, exceed that of boys
(56). In 2013 and 2014, male and female swimmers comprised 43% and 57% of year-​round
USA Swimming athlete membership, respectively (147,180 vs. 193,388 athletes). The only age
group (from <8 to >19 years) in which there were more males than females was over 19 years
old (4785 females, 5961 males representing 1.5% and 1.8% of all swimmers, respectively); other-
wise, more girls participate year round in swimming than boys beginning at 8 years of age.Youth
boys have a slightly wider array of sports to select from (e.g., football), and therefore, youth
swimming for boys may be less popular compared to other team sports.
Educational institutions in the US that require federal funding (high schools and univer-
sities) are also required to provide participation data. However, in contrast to club participa-
tion, NCAA statistics (35) indicate that since the early 1980s in Division I swimming (which
can offer athletic scholarships), the number of intercollegiate women athletes (and teams) has
increased from 3038 (161 teams) to 5393 (195 teams) while men swimmers have dropped from
4109 (181 teams) to 3839 (134 teams). This greater female participation rate in swimming is
likely attributed not only to Title IX legislation which increased opportunities for girls, but also
the long-​standing historical “social acceptability” of swimming and somewhat fewer popular
sport options for girls to participate in as a youth. The reason for slightly greater American men
currently remaining in USA Swimming (club and collegiate) beyond 19 years old is unclear, but
may reflect greater professional opportunities for men and/​or other challenges for women to
devote sufficient time towards training in their child-​bearing years. Therefore, the parity in par-
ticipation rates across most of the age groups for US swimmers suggests that the sex difference
in performing this sport is not likely due to a smaller pipeline of female athletes, although other
sociological factors cannot be ruled out.

Quantifying the performance gap

Why discuss the aforementioned contextual historical events and sociological factors? In order
to fully appreciate the “science” of sex differences, the social environment must be considered
when examining research that compares phenotypes and performances of men and women.
Two different approaches used to “quantify” the sex difference have been to either examine
physiological responses (i.e., aerobic capacity, muscular strength) in groups of men and women
or calculate the performance difference or “gap” (e.g., in world records) from similarly trained
elite athletes representing the extreme tail of the population distribution. With a plethora of
publications over the past two decades, scientists have weighed in on both sides of the argu-
ment: either the performance gap is predetermined and will remain (19, 25, 43, 44, 53) or that
it will eventually close (4, 10, 12, 25, 29, 42, 51, 52, 60) and perhaps remain in only select events.
The first study to examine this question, “Will women soon outrun men?”, was published
in the highly respected journal, Nature, in 1992 (60). Brian Whipp and Susan Ward plotted the
historical progression of mean running velocity in world records over time for the 200 meters
up through the marathon. Since the slope for women’s improvement was far steeper over time
(due to the relatively “late start” in distance running competitions and compressed time scale),
the model predicted that female marathon performance would intersect with men’s in 1998
at a time under 2 hours and 2 minutes. Based on their statistical models for all running events,
this suggested the performance gap between men and women would eventually disappear,
prompting much media interest. However, those knowledgeable about distance running under-
stood this was an incorrect assumption based on the rapid positive societal change rather than an
emerging biological advantage for women.To illustrate this miscalculation of the “opportunities

422
423

Sex and performance: Nature vs. nurture

bias,” as of 2017, the marathon world record was 2:02:57 for men but 2:15:25 for women (set
by Paula Radcliffe in 2003).
As sport opportunities for women have improved over time, the performance gap appears
to have stabilized in the last couple of decades, particularly in objectively measured sports (in
centimeters, seconds, or kilograms) where the human body is the motor. Recent analyses of
world record performances indicate a stability in the performance gap (43, 61) across a number
of sports and event distances (53) since the early 1980s. However, there have been circumstances
(e.g., anaerobic/​power events after more controlled drug testing was implemented to screen for
anabolic steroids) in which the gap has actually widened (44). In 1998, the magnitude of the
performance gap was roughly 11% slower for women in running events from the 1500 meters
through the marathon distance (49). Sparling et al. (49) stated the biological differences between
men and women provide men with an advantage in distance running and recent data have not
altered this conclusion regarding these competitive running distances (18, 22, 25), although
a lower depth of field (i.e., fewer elite competitors) and participation are still acknowledged
as factors (23). Thus, one major limitation in using world records to examine sex differences
include the fact that participation and opportunity may still not be universal around the globe
(5) and records are set under different conditions and competition venues.
The impact of Title IX on the performance gap in the US was recently assessed (33). As
observed in Figure 29.4, computing male–​female differences in time by rank (1st through 8th
place at the Olympic Trials) by distance and event, the performance gap closed by approxi-
mately 2% and 5% in Olympic trial swimming (top) and running (bottom), respectively, from
1972 to 1980 (with no meaningful change thereafter over nearly 40 years). Performance gaps
of 13% in elite mid-​distance running and 8% in swimming (of similar ~4-​minute duration)
have remained steady for four decades, the 5% differential between sports indicative of load
carriage disadvantages of higher female body fatness in running. Conversely, sprint swimming
exhibits a greater sex difference than sprint running suggesting height and anthropometric
advantages unique to swim block starts compared to overcoming height disadvantages in track
block starts (57).
Based on world records, what is the range overall for the performance gap? Sandbakk et al.
(43) have recently summarized the performance gap of 8–​12% between the sexes as stable with
two notable exceptions: >12% for “upper body power” events and <8% in “ultra”-​endurance
swimming –​although this point remains somewhat controversial if controlling for representa-
tive population sampling (58).Vogt et al. (58) reported performance gaps of 6.6% in 25 km open
water events during the 2008 and 2012 Olympic Games and 7.8% for world championships.
These values contrast with other selected open water ultra-​distance races (42) reporting sex
differences varying from 14.4% to as low as 3.7%. Moreover, some (29) reported women were
actually 12–​14% faster than men in a 46 km event held in <20°C water. In contrast, a stable
sex difference of approximately 11.5% has been reported in a 26 km Lake Zurich, Switzerland,
swim event (11). An “advantage” for women during prolonged open water swimming (no turns
or block starts) is postulated to be the enhanced buoyancy of additional body fat. However,
“ultra”-​endurance open water swimming has only recently been added to elite competition
(Olympics, world championships); thus, the lack of an equivalent depth of field in selected
events likely also contributes to these differential findings (9, 45, 46). Moreover, any potential
female metabolic advantage for fat utilization during ultra-​endurance running (events longer
than marathon distance) has limited supporting evidence (43), particularly when coupled with
drastically lower recreational participation rates (20%) in female ultra-​marathoners (28). In add-
ition to sampling bias, the variable duration of ultra-​running events (6 hours up to 10 days)
complicates the female–​male performance comparison.

423
424

Figure 29.4 Performance gap in US Olympic trials (1972–​2016) for top eight finalists by event. Scatter
plot for male/​female computed comparisons for (A) swimming and (B) running events with mean ±
standard deviation denoted by the white line and grey shaded region. The performance gap was higher in
1972 and 1976 compared to all other subsequent years (1980–​2016).
Source: Modified from Millard-​Stafford et al. (33).
425

Sex and performance: Nature vs. nurture

Sex differences due to biology

Why are men faster and stronger than women? The most plausible answer is a biological one
based on XX and XY chromosomes, and specifically the testis-​determining factor or sex-​
determining region (SRY) on the Y chromosome that dictates development of male sexual
characteristics (27). SRY induces a cascade of events involving other genes (e.g., SOX9) to
promote development of the testes and suppress development of the ovaries (2, 27). The phys-
ical characteristics under the influence of testosterone are fairly well documented (31) and
appear to provide biological advantages in sport ranging from musculoskeletal morphology
to neurocognitive influences on behavior. Only selected key physiological characteristics are
presented in this section, based upon the prevailing evidence on sport performance when com-
paring representative cross-​sections of the population (which in itself is a scientific challenge
when examining the published literature on sex differences). Rarely are elite male and female
athletes (of similar world ranking) brought into the same laboratory for assessment of their
physical characteristics and physiological responses during exercise. Historically, such descriptive
studies have been cross-​sectional, occurring at different historical time points and laboratories
(37, 38), often relying upon different techniques and methods for measurement.
Female hormones have important influences on physiology beyond reproduction but bio-
medical studies still lag in this regard (50). Although the effects of fluctuating levels of female
sex steroid hormones (e.g., estrogens, progestins) during the menstrual cycle on women’s sport
performance have been a topic of interest, firm conclusions have yet to be reached (7, 13).
Mechanisms linked to altered substrate (e.g., fat) metabolism (36), thermoregulatory (6), car-
diorespiratory, and psychological changes have all been implicated more or less to some degree
as to whether these interact to enhance or detract from performance. However, reaching
a consensus is unlikely due to an overall lack of studies and inherent individual variability
among the women examined (7), in addition to the fact that few studies are performed on
elite athletes.
Body composition is clearly a differentiating factor between the sexes, specifically when
compared in the leanest elite athletes such as world-​class marathoners (15, 40). Although
assessment methods change over time and have inherent known error and variability, women
(due to estrogen effects) have greater essential (sex-​specific) fat compared to men. Even an elite
female distance runner has a body fatness 8–​10% greater compared to her male counterpart.
This excess fat acts as added load to carry during weight-​bearing exercise since it does not
contribute to propulsive force. Thus, in weight-​bearing sports (e.g., distance running, jumping),
the sex difference may be more pronounced due to the penalty of added load carriage (e.g.,
sex-​specific fat) relative to muscle mass. Excess load carriage explained roughly one-​third of
the endurance running performance gap in both experimental models (8) and in population
comparisons of elite athlete performances for running versus swimming (33). Moreover, greater
relative fat mass was a stronger predictor of the sex difference in running performance compared
to either running economy or cardiorespiratory capacity (48).
Distance running performance is related to maximal aerobic capacity (V̇ O2max), which is
greater in elite male versus female marathoners (37, 39).This sex difference in V̇ O2max has been
attributed to greater male relative heart/​lung size, muscle mass, and hemoglobin concentration
(47) in addition to body composition differences. However, at an elite level, the highest sustain-
able work rate and running economy are better predictors of performance (26) than V̇ O2max;
yet, sex-​based differences in running economy are not observed (16). Therefore, the current
world record in the women’s marathon (2:15.25 by Paula Radcliffe in 2003) has been deemed
“equivalent” to a 2-​hour marathon time projected in the future for men (22).

425
426

M. Millard-Stafford and M.T. Wittbrodt

Muscle cross-​sectional area is generally regarded to be directly proportional to muscular

strength and on average, men have greater overall muscle mass than women although this varies
by sport (32). However, does testosterone enhance the anabolic response to increase muscle
mass following training? Lower testosterone concentrations in women presumably limits protein
synthesis and muscle hypertrophy, although this direct relationship is not consistently observed
in experimental studies examining the acute resistance training-​induced responses in men and
women (59). Although peak strength is greater for men than women (particularly in the upper
body), women may have greater ability to sustain the same relative maximal force (sustained or
intermittent isometric contractions) compared to men (17, 20), although this is not universal
across all tasks. Dr. Sandra Hunter has summarized (21) this body of scientific literature, acknow-
ledging challenges such as limited comparable biomedical data on women and future research is
needed before definitive sex differences in physiology or response to training can be determined.

Summary
Although sport opportunities for women have improved, the performance gap between men
and women has stabilized within recent decades. Over this period, the average age of elite
running and swimming athletes has increased, suggesting greater opportunities to pursue sport
as a professional. Despite these changes in the sociocultural landscape, significant performance
differences in comparable elite male and female athletes remain. These performance gaps are
not uniform across different sports or within the sprint-​endurance domain in a single sport.
Although the performance gap narrows as swimming distance increases from the pool sprints
to open water endurance events, this is not consistent with weight-​bearing events such as
running, likely due to the differential impact of load carriage (e.g., added sex-​specific fat mass in
women). Other biological factors also contribute, such as lower maximal aerobic capacity, peak
muscular forces, and anthropometrical differences.The performance gap between boys and girls
occurs early (prior to adolescence) and progresses with maturity (and presumably testosterone
influences) past the teenage years. Current evidence from world record databases suggests
women will not close the performance gap with men in sports where the human body acts as
the motor. Recent stability in historical trends of elite athletes suggest the current perform-
ance gap is due primarily to underlying biological/​genetic differences rather than sociocultural
disparities in sports participation, but challenges are inherent by relying upon a cross-​sectional
approach to answer the question: Are sex differences due to nature versus nurture?

References
1. Allen SV and Hopkins WG. Age of peak competitive performance of elite athletes: a systematic
review. Sports Medicine 45: 1431–​1441, 2015.
2. Arboleda VA and Vilain E. The evolution of the search for novel genes in mammalian sex determin-
ation: from mice to men. Molecular Genetics and Metabolism 104: 67–​71, 2011.
3. Baxter-​Jones ADG, Thompson AM, and Malina RM. Growth and maturation in elite young
female athletes. Sports Medicine and Arthroscopy Review 10: 42–​49, 2002.
4. Beneke R, Leithauser RM, and Doppelmayr M. Women will do it in the long run. British Journal
of Sports Medicine 39: 410, 2005.
5. Capranica L, Piacentini MF, Halson S, Myburgh KH, Ogasawara E, and Millard-​Stafford
M. The gender gap in sport performance: equity influences equality. International Journal of Sports
Physiology and Performance 8: 99–​103, 2013.
6. Charkoudian N and Stachenfeld N. Sex hormone effects on autonomic mechanisms of thermo-
regulation in humans. Autonomic Neuroscience: Basic and Clinical 196: 75–​80, 2016.
7. Constantini NW, Dubnov G, and Lebrun CM. The menstrual cycle and sport performance. Clinics
in Sports Medicine 24: e51–​82, xiii–​xiv, 2005.

426
427

Sex and performance: Nature vs. nurture

8. Cureton KJ and Sparling PB. Distance running performance and metabolic responses to running
in men and women with excess weight experimentally equated. Medicine and Science in Sports and
Exercise 12: 288–​294, 1980.
9. Deaner RO. Physiology does not explain all sex differences in running performance. Medicine and
Science in Sports and Exercise 45: 146–​147, 2013.
10. Deaner RO, Carter RE, Joyner MJ, and Hunter SK. Men are more likely than women to slow in
the marathon. Medicine and Science in Sports and Exercise 47: 607–​616, 2015.
11. Eichenberger E, Knechtle B, Knechtle P, Rust CA, Rosemann T, Lepers R, and Senn O. Sex
difference in open-​water ultra-​swim performance in the longest freshwater lake swim in Europe.
Journal of Strength and Conditioning Research 27: 1362–​1369, 2013.
12. Eichenberger E, Knechtle B, Rust CA, Knechtle P, Lepers R, and Rosemann T. No gender
difference in peak performance in ultra-​endurance swimming performance –​analysis of the “Zurich
12-​h Swim” from 1996 to 2010. Chinese Journal of Physiology 55: 346–​351, 2012.
13. Friden C, Hirschberg AL, and Saartok T. Muscle strength and endurance do not significantly
vary across 3 phases of the menstrual cycle in moderately active premenopausal women. Clinical
Journal of Sport Medicine 13: 238–​241, 2003.
14. Georgiades E, Klissouras V, Baulch J, Wang G, and Pitsiladis Y. Why nature prevails over nur-
ture in the making of the elite athlete. BMC Genomics 18: 835, 2017.
15. Graves JE, Pollock ML, and Sparling PB. Body composition of elite female distance runners.
International Journal of Sports Medicine 8 (Suppl 2): 96–​102, 1987.
16. Helgerud J. Maximal oxygen uptake, anaerobic threshold and running economy in women and men
with similar performances level in marathons. European Journal of Applied Physiology and Occupuational
Physiology 68: 155–​161, 1994.
17. Hicks AL, Kent-​Braun J, and Ditor DS. Sex differences in human skeletal muscle fatigue. Exercise
and Sport Science Reviews 29: 109–​112, 2001.
18. Hoffman MD. Performance trends in 161-​km ultramarathons. International Journal of Sports Medicine
31: 31–​37, 2010.
19. Holden C. An everlasting gender gap? Science 305: 639–​640, 2004.
20. Hunter SK.The relevance of sex differences in performance fatigability. Medicine and Science in Sports
and Exercise 48: 2247–​2256, 2016.
21. Hunter SK. Sex differences in human fatigability: mechanisms and insight to physiological responses.
Acta Physiologica 210: 768–​789, 2014.
22. Hunter SK, Joyner MJ, and Jones AM. The two-​hour marathon: what’s the equivalent for women?
Journal of Applied Physiology 118: 1321–​1323, 2015.
23. Hunter SK and Stevens AA. Sex differences in marathon running with advanced age: physiology
or participation? Medicine and Science in Sports and Exercise 45: 148–​156, 2013.
24. Hunter SK, Stevens AA, Magennis K, Skelton KW, and Fauth M. Is there a sex difference in the
age of elite marathon runners? Medicine and Science in Sports and Exercise 43: 656–​664, 2011.
25. Joyner MJ. Physiological limits to endurance exercise performance: influence of sex. The Journal of
Physiology 595: 2949–​2954, 2017.
26. Joyner MJ, Ruiz JR, and Lucia A. The two-​hour marathon: who and when? Journal of Applied
Physiology 110: 275–​277, 2011.
27. Kashimada K and Koopman P. Sry: the master switch in mammalian sex determination.
Development (Cambridge, England) 137: 3921–​3930, 2010.
28. Knechtle B. Ultramarathon runners: nature or nurture? International Journal of Sports Physiology and
Performance 7: 310–​312, 2012.
29. Knechtle B, Rosemann T, Lepers R, and Rust CA. Women outperform men in ultradistance
swimming: the Manhattan Island Marathon Swim from 1983 to 2013. International Journal of Sports
Physiology and Performance 9: 913–​924, 2014.
30. Kousta E, Papathanasiou A, and Skordis N. Sex determination and disorders of sex development
according to the revised nomenclature and classification in 46,XX individuals. Hormones (Athens,
Greece) 9: 218–​131, 2010.
31. Lamb DR. Androgens and exercise. Medicine and Science in Sports 7: 1–​5, 1975.
32. Maughan RJ, Watson JS, and Weir J. Relationships between muscle strength and muscle cross-​
sectional area in male sprinters and endurance runners. European Journal of Applied Physiology and
Occupational Physiology 50: 309–​318, 1983.

427
428

M. Millard-Stafford and M.T. Wittbrodt

33. Millard-​Stafford M, Swanson AE, and Wittbrodt MT. Nature vs. nurture: have performance
gaps between men and women reached an asymptote? International Journal of Sports Physiology and
Performance: 13: 1–​19, 2018.
34. National Federation of State High School Associations (NFHS). High school participa-
tion increases for 28th consecutive year. NFHSnewscom www.nfhs.org/​articles/​high-​school-​sports-​
participation-​increases-​for-​28th-​straight-​year-​nears-​8-​million-​mark/​, 2017.
35. National Collegiate Athletic Association. NCAA sports sponsorship and participation rates
report 1981–​82 –​ 2014–​15. www.ncaa.org, 2016.
36. Oosthuyse T and Bosch AN.The effect of the menstrual cycle on exercise metabolism: implications
for exercise performance in eumenorrhoeic women. Sports Medicine 40: 207–​227, 2010.
37. Pate RR, Sparling PB, Wilson GE, Cureton KJ, and Miller BJ. Cardiorespiratory and metabolic
responses to submaximal and maximal exercise in elite women distance runners. International Journal
of Sports Medicine 8 (Suppl 2): 91–​95, 1987.
38. Pollock ML. Characteristics of elite class distance runners-​overview. Annals of the New York Academy
of Sciences 301: 278–​282, 1977.
39. Pollock ML. Submaximal and maximal working capacity of elite distance runners. Part
I: Cardiorespiratory aspects. Annals of the New York Academy of Sciences 301: 310–​322, 1977.
40. Pollock ML, Gettman LR, Jackson A, Ayres J, Ward A, and Linnerud AC. Body composition of
elite class distance runners. Annals of the New York Academy of Sciences 301: 361–​370, 1977.
41. Robinson R. “Eleven Wretched Women”: what really happened in the first Olympic women’s 800
m. Runner’s World, May 14, 2012.
42. Rust CA, Knechtle B, Rosemann T, and Lepers R. Women reduced the sex difference in open-​
water ultra-​distance swimming La Traversee Internationale du Lac St-​ Jean, 1955–​2012. Applied
Physiology Nutrition and Metabolism 39: 270–​273, 2014.
43. Sandbakk O, Solli GS, and Holmberg HC. Sex differences in world-​record performance: the
influence of sport discipline and competition duration. International Journal of Sports Physiology and
Performance 13: 2–​8, 2018.
44. Seiler S, De Koning JJ, and Foster C. The fall and rise of the gender difference in elite anaerobic
performance 1952–​2006. Medicine and Science in Sports and Exercise 39: 534–​540, 2007.
45. Senefeld J, Smith C, and Hunter SK. Sex differences in participation, performance, and age of
ultramarathon runners. International Journal of Sports Physiology and Performance 11: 635–​642, 2016.
46. Smith M and Wrynn A. Women in the 2012 Olympic and Paralympic Games: An Analysis of Participation
and Leadership Opportunities. Ann Arbor, MI: SHARP Center for Women and Girls, 2013.
47. Sparling PB. A meta-​analysis of studies comparing maximal oxygen uptake in men and women.
Research Quarterly for Exercise and Sport 51: 542–​552, 1980.
48. Sparling PB and Cureton KJ. Biological determinants of the sex difference in 12-​min run per-
formance. Medicine and Science in Sports and Exercise 15: 218–​223, 1983.
49. Sparling PB, O’Donnell EM, and Snow TK. The gender difference in distance running perform-
ance has plateaued: an analysis of world rankings from 1980 to 1996. Medicine and Science in Sports and
Exercise 30: 1725–​1729, 1998.
50. Stachenfeld NS and Taylor HS. Challenges and methodology for testing young healthy women in
physiological studies. American Journal of Physiology Endocrinology and Metabolism 306: E849–​E853, 2014.
51. Tanaka H. The battle of the sexes in sports. Lancet 360: 92, 2002.
52. Tatem AJ, Guerra CA, Atkinson PM, and Hay SI. Athletics: momentous sprint at the 2156
Olympics? Nature 431: 525, 2004.
53. Thibault V, Guillaume M, Berthelot G, Helou NE, Schaal K, Quinquis L, Nassif H, Tafflet
M, Escolano S, Hermine O, and Toussaint JF.Women and men in sport performance: the gender
gap has not evolved since 1983. Journal of Sports Science & Medicine 9: 214–​223, 2010.
54. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, and McDowell M. Physical
activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise
40: 181–​188, 2008.
55. Tucker R and Collins M. What makes champions? A review of the relative contribution of genes
and training to sporting success. British Journal of Sports Medicine 46: 555–​561, 2012.
56. USA Swimming. Statistics 2014. www.usaswimming.org/​_​Rainbow/​Documents/​a31bc239-​b31f-​
4834-​87bf-​accb09e8a834/​Statistics-​2014.pdf. 2014.
57. Uth N. Anthropometric comparison of world-​class sprinters and normal populations. Journal of Sports
Science & Medicine 4: 608–​616, 2005.

428
429

Sex and performance: Nature vs. nurture

58. Vogt P, Rüst CA, Rosemann T, Lepers R, and Knechtle B. Analysis of 10 km swimming per-
formance of elite male and female open-​water swimmers. SpringerPlus 2: 603, 2013.
59. West DW, Burd NA, Churchward-​Venne TA, Camera DM, Mitchell CJ, Baker SK, Hawley
JA, Coffey VG, and Phillips SM. Sex-​based comparisons of myofibrillar protein synthesis after
resistance exercise in the fed state. Journal of Applied Physiology 112: 1805–​1813, 2012.
60. Whipp BJ and Ward SA. Will women soon outrun men? Nature 355: 25, 1992.
61. Wild S, Rust CA, Rosemann T, and Knechtle B. Changes in sex difference in swimming speed
in finalists at FINA World Championships and the Olympic Games from 1992 to 2013. BMC Sports
Science, Medicine and Rehabilitation 6: 25, 2014.

429
430
431

SECTION 6

The ethics of systems genetics

in exercise and sport

In Section 6 we come to the end of our endeavor to provide both breadth and depth in our ana-
lysis of the systems genetics of sport and exercise. Unlike previous chapters, with their in-​depth
examination of research studies and tables and figures of genes and outcomes, the authors in
Section 6 cause us to pause and reflect on how this knowledge is being used, and is anticipated
to be used, by athletes, coaches, and other stakeholders.
We start with a discussion of the controversial topic of race and sport performance, exam-
ining whether genetic advantage in sport is based on geographic ancestry. Are East Africans
genetically endowed to be better distance runners, while West Africans better sprinters? Would
Jamaican bobsledders beat German bobsledders if only for more opportunities to pursue this
winter sport and perfect their technique in the tropics? Or is race simply a proxy for cultural
differences and environmental and economic opportunities that mirror race differences in other
aspects of our lives? Drs. Nauright and Wiggins in Chapter 30 focus on these themes and
­provide unique insights into these questions.
Tying back to the preceding sections, Chapter 31, by Drs. Andrew Venezia and Stephen Roth,
examines the explosion of genetic testing for a variety of traits now available to anyone with
an Internet connection and a credit card. Genetic testing technology is robust and becoming
relatively inexpensive, allowing companies to develop tests and sell them online to a variety
of consumers. Whether the science backing the results or conclusions of those tests is high-​
quality or inconclusive is another question altogether. With genetic testing comes a variety of
interesting and challenging ethical questions around talent identification and performance pre-
diction, especially appropriate to consider for children.
Beyond genetic testing is the realm of gene therapy, the direct manipulation of genetic
sequences to alter our biology. Gene therapy has had mixed success in the realm of disease
therapy, but progress is being made. The potential for athletes to misuse gene therapy for per-
formance enhancement, otherwise known as gene doping, has been anticipated for over a
decade, with the World Anti-​Doping Agency holding its first scientific workshop on gene
doping in 2002. How should we view gene therapy in sport? Should it be completely banned
or used only for injury treatment in rare cases? Or, should we embrace this new technology and
the performance enhancements that are sure to be generated with its careful use? Drs. Verner
Møller and Rasmus Bysted Møller of Aarhus University in Denmark outline their views on the

431
432

The ethics of systems genetics in exercise and sport

ethics of gene doping in Chapter 32, while Dr. Andy Miah of the University of Salford in the
UK counters with a transhumanist perspective in Chapter 33.
While gene therapy in sport and gene doping remained mostly hypothetical concerns as
this book was coming to conclusion, an area of very real controversy is the subject of the final
chapter in this section. Chapter 34, contributed by Dr. Jaime Schultz of Pennsylvania State
University in the US, examines the ethics of sex and gender testing in sport competition.
As outlined in Section 5, we know that sex differences result in very real differences in per-
formance in many sports, resulting in sex-​specific leagues and competitions. But how is sex
determined for the purpose of distinguishing the members of those competitions? And is sex
truly a binary of male and female that fits neatly into the structure of our sport leagues? This
final chapter shows the difficult history of sex and gender testing and outlines the challenges of
present-​day approaches to ensuring “fair” competition between the sexes.
The chapters in Section 6 are meant to open readers’ eyes to the consequences of this growing
knowledge of systems genetics in sport. While researchers may have the best of intentions, and
in fact may not be focused on sport at all in their conceptions of their work, knowledge of the
genetic underpinnings of traits relevant to sport performance will be used, misused, and abused
by athletes, much as we have seen with drug-​related doping for many decades. As leaders in our
respective fields, we must be aware of the indirect implications of our work and not shy away
from helping to inform and educate the public about the consequences of both genetic testing
and gene therapy in the world of sport.

432
433

30
RACE AND SPORTS
PERFORMANCE
John Nauright and David K. Wiggins

From the 19th century and the emergence of modern sports competitions, athletes and scientists
sought ways to enhance and improve performance relative to other athletes, countries, and
races. While much of performance science was initially applied to race horses, increasingly
understanding of the moving body, its capabilities and its limits engaged scientists internation-
ally (12).While modern science was being applied to the body, anthropologists and ethnologists
were attempting to understand differences between races and cultures. Ideas about temperament
were applied to cultures and sports. For example, many English observers believed that the sport
of cricket did not translate well outside of upper levels of society in the English-​speaking world
because the French, other Europeans, and native peoples did not have the patience or tempera-
ment to master the psychological aspects of the sport. Additionally, urbanization and the spread
of disease, which led to sanitation movements and sewage systems for cities, raised alarm bells
for European and North American observers by 1900 as they became more and more obsessed
with the fear that the “white race” was in decline. After the South African (or Boer) War of
1899–​1902, numerous commissions in Britain examined the state of working class British men
and determined, in publications such as the Report of the Inter-​departmental Committee on Physical
Deterioration (1904), that the average health of British men had declined as a result of urbaniza-
tion. Following on from this, defeats of British Isles national and regional rugby teams by New
Zealand (1905), South Africa (1906, 1912), and Australia (1908), whose troops had been deemed
to be more healthy than English ones, provided further “evidence” that it was only in colonial
settings (Australia, New Zealand, Canada, the American West in the US) where a rugged and
rural lifestyle was thought to abound where the white race was potentially improving (20, 21).
The emergence of black athletes particularly in boxing (Peter Jackson, Jack Johnson),
American football (Paul Robeson and others), cycling (Marshall “Major” Taylor), and athletics
(Sol Butler) between 1890 and 1920 added to fears that white athletes might not be as able
to compete in the long run. The world championship victory by Jack Johnson over white
Canadian Tommy Burns in Sydney, Australia in 1908, coupled with Johnson’s interaction with
white women, raised racial concerns to extreme levels (27). The response was to harden the
color line and segregation became the norm in American baseball and basketball, though a small
number of black athletes were able to compete through to the 1930s in American football and
in Olympic sports.

433
434

J. Nauright and D.K. Wiggins

We should remember the emergence of black athletic success in the US occurred against
hardening racial lines as promoted in popular culture through Thomas Dixon’s novel The
Clansman (1905) which was later turned into the movie Birth of a Nation (1915). Dixon’s book
was an argument against desegregation and he argued black men would turn towards savage
behavior if allowed too much freedom (24).
The first attempt at a scientific examination of physical abilities of races appeared in Joseph-​
Marie Degerano’s The Observation of Savage Peoples which was published in 1800. The use of
dynamometers during the 19th century was common and used in several widely publicized
studies to substantiate claims that English or French men were the strongest and Australian
Aborigines the weakest among groups of peoples. Many of these early experiments were shown
to lack scientific validity and merely served to legitimate the ideologies of hierarchy of races
which always placed northern European or North American white men at the top of the scales
of civilization as well as in physical capacity. Athletic challenges by black athletes, however,
placed a strain on this dominant ideology and so analyses began to link athleticism with savagery
and physicality became increasingly juxtaposed against mental capabilities (12).
In 1936, W. Montague Cobb, a prominent black medical doctor and physical anthropolo-
gist from Howard University, Washington DC, published the well-​known essay “Race and
Runners” in The Journal of Health and Physical Education that attempted to refute the claims by
popular writers and others both in and outside of sport that black people were endowed with
innate anatomical and physiological gifts that accounted for their great athletic performances.
Based on anthropometric measurements he had taken from track and field star Jesse Owens and
other black athletes, Cobb made clear his view that it was erroneous to argue that black people
possessed a musculature and genes unique to their race that allowed them to shatter athletic
records and achieve unparalleled success on America’s playing fields. Instead, noted Cobb, the
success of black individuals in sport, albeit in only a select number of sports, resulted from a
combination of great physical talent and opportunities as well as hard work, desire, dedication,
and a host of other character traits so admired in American culture (6, 34).
Cobb’s refutation of the concept of biological determinism did not put an end to the debate
regarding the impact of race on sport performance that had been waged since the latter part
of the 19th century and continues, in one form or another, to this very day. It was merely one
of the most explicit and logical denunciations of the persistent belief that race predisposed
black people to superior athletic performances made by a prestigious scientist with a national
profile by virtue not only of his academic work but his civil rights activism and leadership in
national organizations. Importantly, it was also a denunciation of the notion of biological deter-
minism made during a period of time that the athletic accomplishments of black athletes were
becoming far more visible at both the national and international levels of competition. Notable
examples of this fact were the increasing number of black athletes distinguishing themselves in
sport at prestigious predominantly white universities in the US and in both the 1932 Olympic
Games in Los Angeles and 1936 Olympic Games in Berlin (11, 34).
The public debate regarding race and sport performance lost some of its steam during World
War II as athletic programs at all levels of competition lost some of their best performers to the
war effort and with the cancellation or alteration of games and leagues and organizations. Once
the war to defend Franklin Roosevelt’s “Four Freedoms” was complete, the debate would heat
up as ordinary fans, popular writers, coaches, academicians, and others from various walks of
life would offer their views as to what accounted for the great performances of black athletes.
Not unexpectedly, the debate always seemed to gain momentum during the summer Olympic
Games as black athletes, particularly in track and field, displayed their speed and jumping abil-
ities against other countries’ best athletes in front of a worldwide audience. This was certainly

434
435

Race and sports performance

the case regarding the 1948 Olympic Games in London, 1952 Olympic Games in Helsinki, and
1956 Olympic Games in Melbourne, as the outstanding accomplishments of such great black
athletes from the US as Harrison Dillard, Mal Whitfield, Milt Campbell, Barney Ewell, and
William Steele would attest (11, 18).
Two essays published in 1952 point to the decidedly different views that would continue
to exist regarding the debate over race and sports performance. G.P. Meade, manager of the
Colonial Sugars Company in Gramercy, Louisiana, and a man who had a fascination with
records and sports accomplishments, proffered in his essay “The Negro in Track Athletics” in
the Scientific Monthly that why African Americans had provided such a large proportion of
extremely high class performances could not be accounted for by innate natural differences,
“but probably because of sociologic reasons” (18). British medical doctor Adolph Abrahams,
on the other hand, took a decidedly different position in his article “Race and Athletics” in
the Eugenics Review. He came down on the side of innate physiological differences to account
for the outstanding performances of black athletes. “I think there is some special quality in his
[African American] muscles” wrote Abrahams. “A more rapid contractibility or reduced vis-
cosity. Alternatively, there may be a superior co-​ordination related to his nervous system” (18).
The pattern of the debate during the 1950s continued unabated during the following decade
as every Summer Olympic Games brought more individuals into the discussion. If there is one
distinguishing feature of the debate during the 1960s it is the increasing attention given to the
topic in newspapers and popular journals, an indication perhaps more than anything else of the
ever-​expanding visibility of the Olympic Games resulting from television coverage and other
media forms. The decade began with British medical doctor James M. Tanner publishing the
results of the anthropometric measurements he completed on 137 athletes from the Rome
Olympics of 1960. Included in his text, The Physique of the Olympic Athlete,Tanner conceded that
economic and social factors undoubtedly played a part in the number of blacks participating in
highly organized sport, but noted that the different body types of black track and field athletes
is perhaps what accounted for their outstanding performances in particular events (30). In 1964,
writer Marshall Smith, in his provocatively titled Life magazine article “Giving the Olympics
an Anthropological Once-​Over” echoed some of what Tanner had expressed in his study (29).
Smith relied to a great extent on the views of well-​known anthropologists Carleton S. Coon
and Edward E. Hunt, Jr., both of whom were solidly in the camp of biological determinists by
stressing the importance of “inherited physical adaptations” and contending that the unique
body type of black athletes gave them a distinct advantage in certain sporting events (7). Finally,
in 1968, sportswriter Charles Mahar summed up the various arguments given up to that time
about the connection between race and sports performance in two articles in a five-​part series
on the black athlete in the Los Angeles Times. As the titles of his two articles suggest, “Blacks
Physically Superior? Some Say They’re Hungrier” and “Do Blacks Have a Physical Advantage?
Scientists Differ” Mahar essentially categorizes the arguments into those that took a biological
determinist approach to the debate and those that viewed the success of black athletes as
resulting from a combination of finely tuned physical skills, hard work, dedication, opportun-
ities, and a host of other social and economic factors (17).
Certainly a watershed episode in the debate regarding race and sport performance involved
an article written by sportswriter Martin Kane and the pointed and strongly worded responses
to that article by Harry Edwards, a professor of sociology at the University of California,
Berkeley, and architect of the proposed boycott of the 1968 Mexico City Olympic Games (9).
In 1971, Kane penned a detailed account of the debate in Sports Illustrated, a highly popular
journal in which he served as senior editor. Depending on the expertise of various individuals,
including coaches, medical doctors, and sport researchers, Kane made the case in the article

435
436

J. Nauright and D.K. Wiggins

“An Assessment of Black is Best” that the dominance of black athletes resulted from race-​based
psychological, historical, and physiological factors (15). Kane relied to a great extent on the
expertise of famous University of Indiana and US Olympic swimming coach James Councilman
who contended that the large number of white muscle fibers in black athletes was responsible
for their outstanding performances in sports that put a premium on speed and power (8).
Conversely, the preponderance of red muscle fibers in white athletes made them ideally suited
for those sports requiring endurance. Oddly, Councilman turns to socioeconomic reasons to
explain the few number of blacks in his sport, contending that blacks lacked the money and
access to facilities that were essential to participating and realizing success in swimming. For
whatever reason, he does not state in the Kane piece like he did in the previously mentioned
Marshall Smith essay in Life magazine of 1964 that it was difficult for black people to learn how
to swim because of their “lack of buoyancy,” one of several long-​held racial stereotypes used to
explain away the lack of black participants in the sport (8).
Edwards, obviously deeply troubled by the Sports Illustrated piece, quickly responded to the
article with a serious article of his own that countered in great detail Kane’s assertions. Edwards
was highly critical of Kane’s generalizations about the supposed special aptitudes and abilities
of a particular racial group, especially when considering, as W. Montague Cobb had postulated
years before, the fact that there were “more differences between individual members of any one
racial group than between any two groups as a whole.”This obviously precluded any arguments
that certain racial groups were uniquely built and destined to achieve success in certain sports.
Edwards asserted that Kane’s notion that black people possessed a unique psychological dispos-
ition in being able to relax under pressure was ludicrous, a deep-​seated stereotypical notion that
had no basis in fact. Edwards vehemently disputed Kane’s claim that hereditarily weaker black
individuals had been separated out during slavery and what was left was a physically superior
group uniquely predisposed to athletic success. Implied in Edward’s comment is that Kane’s
assertion regarding biological determinism and connection to superior athleticism was the same
racialist thinking that had been used for years to claim black intellectual inferiority. In the end,
Edwards asserted that a number of interrelated societal conditions and circumstances accounted
for the outstanding performances of black athletes as well as their commitment to the institu-
tion of sport that had traditionally been perceived to be open to them (9).
Kane’s article and Edwards’ response to it seems particularly important for at least three
different reasons. First of all, it took place during a period of time in which the lives and
careers of black athletes were receiving increased attention from scholars and the academic
community more generally. The involvement of black athletes in the larger civil rights struggle,
including the protests they waged during Olympic competitions and on predominantly white
university campuses, combined with a new-​found interest in the life and culture of black
Americans, expanded their public profile and resulted in a plethora of academic studies that
not only addressed questions about their excellence in certain sports but also such issues as
inequality in salaries and inability to secure coaching and upper level administrative positions.
Second, Edwards’ claim that the same racialist thinking used to account for black dominance in
sport was used to explain supposed black intellectual inferiority was certainly correct and not
proffered in a vacuum as the US was in the midst of a controversial debate about race, IQ scores,
and academic achievement (9). In 1969, just 2 years before the publication of Kane’s article,
Arthur R. Jensen, a professor of psychology at the University of California, Berkeley, published
a lengthy study in the Harvard Educational Review that reenergized the long-​standing debate
regarding the impact of race on the supposed differences between black and white intelligence.
“Genetic differences are manifested” noted Jensen, “in virtually every anatomical, physiological,
and biological comparison one can make between representative samples of identifiable racial

436
437

Race and sports performance

groups. There is no reason why the brain should be exempt from this generalization” (13).
Jensen’s views, along with others that took the same position like Stanford University physi-
cist and Nobel Laureate William B. Shockley, drew volatile responses from both black and
white people appalled by such racialist thinking and the disregard for the social, economic,
and environmental factors that play extraordinarily important roles in any differences between
and among men and women (28). Third, but certainly not least, Kane’s article and Edwards’
response to it makes patently clear an important fact about the debate over the impact of race
on sport performance that is not always acknowledged and that is almost the entire focus of the
discussions on black male rather than black female athletes. One simple reason for this, as sport
management scholar Jennifer Bruening and others have pointed out, is that black women are
largely invisible, never fully recognized for their singular experiences, regularly marginalized by
either being categorized in the same group as white women or black men (4).
The ensuing debate regarding race and sport performance would certainly have practical
implications and potentially deleterious effects. One prime example would be the impact
it would have on black athletes themselves. Although not responding uniformly to issues
surrounding the debate, some black athletes have bought into the notion that their race is
innately gifted physically and that it has resulted in superior sporting accomplishments. In
a 1977 Time magazine article titled “Sport: The Black Dominance,” football’s infamous O.J.
Simpson and baseball Hall of Famer Joe Morgan expressed such views. “We are built a little dif-
ferently” said Simpson, “built for speed—​skinny calves, long legs, high asses are all characteristic
of blacks.” Morgan was just as explicit as his football counterpart by noting: “I think blacks,
for physiological reasons, have better speed, quickness, and ability. Baseball, football, and basket-
ball put a premium on those skills” (31). More recently, in 2012 to be exact, multiple Olympic
gold medal-​winning track and field athlete Michael Johnson weighed in on the matter with
obvious conviction by talking about the impact of slavery on him and other members of his
race. “All my life,” said Johnson, “I believed I became an athlete through my own determination,
but it’s impossible to think that being descended from slaves hasn’t left an imprint through the
generations. Difficult as it was to hear, slavery has benefited descendants like me—​I believe there
is a superior athletic gene in us” (14).
Holding on to the belief they had genes that predisposed them to great sporting feats would
certainly give Simpson, Morgan, Johnson, and other black athletes a psychological advantage,
while having the opposite effect for white athletes who held the same views. Believing in “bio-
logical cultural destiny” could certainly provide a psychological edge for blackathletes in an
activity where the margins of victory are often very slim and usually determined by participants
with high confidence levels as well as finely tuned bodies and carefully honed athletic skills.
Conversely, believing in such a faulty supposition was of little help, in fact it could be a dev-
astating blow, to those black athletes that did not have the requisite physical skills to reach the
highest levels of sport that everyone expected of them. In addition, whether black athletes
were cognizant of the fact or not, the racialist thinking that undergird the debate regarding
race and sport performance was also used to discriminate against them, limit them to par-
ticular playing positions, and deny them opportunities to freely compete for coaching, man-
agerial, and upper-​level administrative positions. Perhaps there was no more famous example
of this then Los Angeles Dodger Vice President Al Campanis attempting to explain to host
Ted Kopell of Nightline in 1987 on the 40th anniversary of Jackie Robinson’s shattering of
the color line in Major League Baseball why there were still so few black managers and upper
level administrators in the sport. When Kopell specifically asked Campanis “Is there still that
much prejudice in baseball today?” the Dodger Vice President responded by saying “No, I don’t
believe it’s prejudice. I truly believe that they may not have some of the necessities to be, let’s

437
438

J. Nauright and D.K. Wiggins

say, a field manager, or perhaps a general manager.” When Kopell pressed him further on the
matter, Campanis continued spouting illogical racialist theories, noting for instance at one point
that “So, it might be—​why are black men, or black people, not good swimmers? Because they
don’t have the buoyancy.” Not unexpectedly, Campanis was forced to resign from his position
with Dodger owner Peter O’Malley noting that “Campanis’ statements on the ABC Nightline
show Monday night were so far removed from the belief of the Dodger organization that it was
impossible for him to continue in his duties” (32, 33).
Unfortunately, the belief on the part of people, obviously both black and white and from
every stratum of society, that black people were endowed with special physical talents was slow
to fade away as evidenced by the periodic spewing of such beliefs throughout the latter stages of
the 20th century. The pattern would become clear. Someone who took the view that race was
a biological phenomenon would make the assertion that black people had unique anatomical
and physiological gifts that accounted for their outstanding performances in particular activities
and this would then be countered by those who recognized the erroneous nature of this argu-
ment and that a combination of physical, sociological, economic, and cultural factors accounted
for the success of athletes irrespective of the race in which they identified. This would happen
with some regularity with those in the public eye who espoused the view that black athletes
achieved success in certain activities because of anatomical and physiological abilities unique to
their race, sometimes actually losing their jobs because of the racialist position they took. The
examples are many, but two that garnered much attention and controversy were the infamous
1988 interview conducted by Ed Hotaling with CBS prognosticator and sports broadcaster
Jimmy “The Greek” Snyder and Jon Entine’s attention-​g rabbing 2000 book Taboo: Why Black
Athletes Dominate Sports and Why We’re Afraid to Talk About It (10, 26).
On Martin Luther King Day (January 15) in 1988 at a restaurant in Washington, DC, local
newscaster Hotaling asked Snyder his opinion as to the progress being made by black athletes.
The response that Snyder gave to Hotaling’s question, a response he must have regretted until
the day he died, was laced with illogical assumptions, deep-​seated stereotypical notions about
race, and generalizations that were decidedly ahistorical in nature. He noted that black people
were far better athletes than white people since they had been “bred to be that way since the
days of slavery” and that if more black people found their way into coaching positions “there’s
not going to be anything left for the white people.” The origin of black dominance in sport
took place during the period of the Civil War when “the slave owner would breed his big black
with his big woman so that he could have a big black kid.” It is a fact, said Snyder, that black
athletes “jump higher and run faster” as a result of their “thigh size and big size.” In an apparent
sign of regret, he argued that white athletes would never be able to overcome these physical
disadvantages and achieve the same level of accomplishments as their black counterparts, par-
ticularly because they were lazy and less motivated to realize success in sport (26).
Not unexpectedly, Snyder’s comments drew harsh criticism from those in the popular press.
The editors of Sports Illustrated wrote that Snyder obviously had no understanding of history or
the institution of sport and that his broad generalizations were exactly the way in “which racial
stereotypes and prejudices are built.” Washington Post writer Carl Rowan compared Snyder to
Joseph Goebbels, Adolph Hitler’s cruel Minister of Public Enlightenment and Propaganda (25).
A cartoon in the Boston Globe showed Snyder being comforted by a hooded Ku Klux Klan
member. After just 2 days of intense and far-​reaching public condemnation of Snyder, CBS had
no choice but to fire him. Apparently, Snyder learned nothing from the Al Campanis debacle
that took place some 9 months earlier.
Snyder’s disturbing and offensive remarks also brought sport studies scholars, particularly
from sociology, even more squarely into the debate regarding race and sport performance. Harry

438
439

Race and sports performance

Edwards, for instance, did not mince words, calling Snyder ignorant and incompetent and regu-
larly spoke of the fallacy of the broadcaster’s statements in written form and through interviews
and other public forums. Jay Coakley, primarily through the multiple editions of his text Sport in
Society: Issues and Controversies, stepped up his already vehement denunciation of racial science,
arguing that the preponderance of black athletes in particular sports had to do with a combin-
ation of racial discrimination, characteristics of certain sports, and the motivation of individual
athletes. It also has to do, says Coakley, with how black athletes defined their chances for success
and those sports participated in by those successful athletes who served as role models during
their youth (5).
Beyond the US, discussions continued to either try and legitimate strong performance or
explain poor performance in certain sports by African and other black athletes. As late as 1996,
British Prime Minister John Major suggested Britain send “athletic missionaries” to Africa in
search of talent while in South Africa, white rugby officials tried to explain away the low
numbers of black rugby players at top levels by arguing they did not have experience playing
the sport. These assertions went counter to the more than 100-​year history of black rugby
in South Africa which developed alongside that of white rugby, though mostly in segregated
environments (22).
In 2000, Jon Entine raised the level of debate regarding race and sport performance to
an even more feverish pitch internationally with the publication of his controversial book
Taboo: Why Black Athletes Dominate Sports and Why We’re Afraid to Talk About It. Entine, a jour-
nalist and author who had helped produce, with the iconic American broadcaster Tom Brokow,
a 1989 1-​hour documentary on the topic, put together a book of over 300 pages that largely
sums up the many issues long associated with the debate. Perhaps most importantly, however,
Entine’s arguments and interpretations helped revive the bitter disputes that had always placed
in opposition the biological determinists and cultural theorists who had broached the topic.
Although asserting he was only concerned with acknowledging human diversity, not interested
in spewing racialist beliefs, and distancing himself from “The Bell Curve” that connected intel-
ligence and race, Entine’s belief in race as a biological phenomenon and his undying faith
in the objective and value-​neutral nature of science ultimately precluded him from seriously
considering the myriad sociological and cultural factors that largely accounted for those sports
in which they participated and, in some cases, dominated.Very troubling is Entine’s decision to
lump Africans and African Americans into one group and attempting to explain the differences
between sprinters and long-​distance runners based on such anatomical and physiological factors
as muscle fibers, muscle mass, lung capacity, body size, centers of gravity, and enzymes. Just as
disturbing is Entine’s claim that white athletes have been marginalized, false statements that black
athletes have achieved relatively much success in most sports, and inexplicable commitment to
racial science while at once disparaging the racist eugenics studies of the late 19th and early 20th
centuries, which he details in the book as being misplaced and without thorough vetting (10).
Like Al Campanis and Jimmy “The Greek” Snyder some 12 years earlier, Entine received a
rash of criticism for the arguments he put forward in Taboo. In spite of the criticism, however, it
did not put an end to the efforts of biological determinists to locate the genes that accounted
for the success of black athletes. The insistence on searching for unique athletic genes among
black athletes has not been an innocent endeavor, but evidence of the continual need to rank
people based on skin color so as to rationalize inequitable treatment and confirm the supposed
superiority of one race over another. There has been no comparable search for anatomical and
physiological differences among white athletes that could explain their domination of golf,
tennis, ice hockey, swimming, skating, skiing, field hockey, volleyball, and so many other sports.
No need to explain the overrepresentation of white participants in these sports by virtue of the

439
440

J. Nauright and D.K. Wiggins

fact that many people gloss over this reality, viewing the overwhelming success of individual
black athletes in a few selected sports as emblematic of the entire race.
Tellingly, much of the scholarly focus on the topic beginning with the publication of
Entine’s book has been on Kenyan distance runners. The extraordinary success of these athletes
in the recent past has generated much interest from scientists, an interesting but not surprising
juxtaposition in the debate regarding race and sports performance considering that for years
the focus was on finding the anatomical and physiological reasons for the seeming domin-
ance of African American sprinters and assumptions that the modest performances by African
American distance runners was merely the result of a lack of stamina and, by extension, the
heart. Examples of the thrust of the studies on Kenyon distance runners can be gleaned from
their titles, including headings ranging from “Kenyan and Ethiopian Distance Runners: What
Makes Them So Good” (35) and “Demographic Characteristics of Elite Kenyan Endurance
Runners” (23) to “Kenyan Dominance in Distance Running” (16) and “Dissociation Between
Running Economy and Running Performance in Elite Kenyan Distance Runners” (19).
John Bale and Joe Sang, writing in 1996, clearly outlined a multifaceted rationale for Kenyan
running success which scientists obsessed with racial explanations have largely ignored. Bale
and Sang demonstrate that nearly all top runners from Kenya are Kalenjin.The Maasai, who the
British originally thought would be more successful as athletes, live in similar environmental
conditions, yet very few have achieved success in distance running. The Kalenjin do live at
altitude, but, uniquely have a very high achievement principle in their culture. Bale and Sang
also demonstrate that many other African countries are underrepresented as are non-​African
countries where many people live at altitude, thus neither race nor geography can fully explain
successes. Furthermore, Bale and Sang show that Kenya was one of the first places where
Finnish methods of running coaching were taught such that by their first major international
competition at the 1954 Empire Games, Kenyan runners were competitive (3).
Despite no real evidence of a “magic racial gene,” scientists and observers continue to search
for explanation of success by black athletes in any sport where they are overrepresented or
achieve disproportionate success. The converse, as we stated, is never applied to white athletes
or the sports where they are overrepresented. Professor Yannis Pitsiladis, director of the Sub2hrs
marathon project and expert on genomics and sport performance, has argued there is no magic
gene thus agreeing with the assessment of Montague Cobb from the 1930s. Pitsiladis does
argue, however, that genetics matter in that athletes who have successful athletic parents and
grandparents are more likely to succeed than those who do not. Importantly though, there is no
skin color correlation. Scientists would do better to eliminate skin color from experimentation
and deracialize the science of human performance.

References
1. Anthony, A. White men can’t run. The Observer June 4, 2000. www.jonentine.com/​reviews/​observer.
htm. [Accessed April 17, 2018.]
2. Ashe, A.R. A Hard Road to Glory: A History of the African-​American Athlete Since 1946. New York: Amistad
Press, 1993.
3. Bale, J. and Sang, J. Kenyan Running. London: Frank Cass, 1996.
4. Bruening, J. Gender and racial analysis in sport: are all the women white and all the blacks men?
Quest 57: 330–​349, 2005.
5. Coakley, J. Sports in Society: Issues and Controversies. 12th edition. New York: The McGraw-​Hill
Companies, 2017.
6. Cobb, W.M. Race and runners. The Journal of Health and Physical Education 7: 3–​7, 52–​56, 1936.
7. Quoted in Smith, M. Giving the Olympics an anthropological once over. Life Magazine, October 23,
83, 1964.

440
441

Race and sports performance

8. Quoted in Kane, M. An assessment of black is best. Sports Illustrated, January 18, 72–​73, 1971.
9. Edwards, H. The sources of the black athletes superiority. The Black Scholar 3: 32–​41, 1971.
10. Entine, J. Taboo: Why Black Athletes Dominate Sports and Why We’re Afraid to Talk About It.
New York: Public Affairs, 2000.
11. Gleaves, J. and Dyreson, M. The Black Auxiliaries in American Memories: Sport, Race And
Politics in the Construction of Modern Legacies. The International Journal of the History of Sport
27: 2893–​2924, 2010.
12. Hoberman, J. Mortal Engines: The Science of Human Performance and the Dehumanization of Sport.
New York: The Free Press, 1992.
13. Jensen, A.R. How much can we boost I.Q. and scholastic achievement? Harvard Educational Review
39: 1–​123, 1969.
14. Quoted in Chase, C. Michael Johnson says slavery descendants run faster because of “superior
athletic gene”. Yahoo Sports, July 5, 2012. sports.yahoo.com/​blogs/​olympics-​fourth-​place-​medal/​
michaeljohnson-​says-​slavery-​descendents. [Accessed September 14, 2018.]
15. Kane, M. An assessment of black is best. Sports Illustrated, January 18, 72–​83, 1971.
16. Larsen, H.B. Kenyan dominance in distance running. Comparative Biochemistry And Physiology-​Part
A: Molecular & Integrative Physiology 136: 161–​170, 2003.
17. Maher, C. Blacks physically superior? Some say they’re hungrier. Los Angeles Times, March 24, 1968;
Maher, C. Do blacks have a physical advantage? scientists differ. Los Angeles Times, March 29, 1968.
18. Meade, G.P. The Negro in track athletics. Scientific Monthly 75: 366–​371, 1952; Abrahams, A. Race
and Athletics 44: 143–​145, 1952.
19. Mooses, M., Mooses, K., Haile, D.W., Durussel, J., Kaasik, P. and Pitsiladis, Y.P. Dissociation
between running economy and running performance in elite Kenyan distance runners. Journal of
Sports Sciences 33: 136–​144, 2014.
20. Nauright, J. Sport and the image of colonial manhood in the British mind: British physical
deterioration debates and colonial sporting tours, 1878–​1906. Canadian Journal of History of Sport
23: 54–​71, 1992.
21. Nauright, J. Colonial manhood and imperial race virility: British responses to colonial rugby tours.
In: Nauright J and Chandler TJL (Eds.), Making Men (pp. 121–​139). London: Routledge, 1996.
22. Nauright, J. Long Run to Freedom: Sport Cultures and Identities in South Africa. Morgantown, WV:
Fitness Information Technology, 2010.
23. Onywera, V.O, Scott, R.A, Boit, M.K and Pitsiladis, Y.P. Demographic characteristics of elite
Kenyan endurance runners. Journal of Sports Sciences 24: 415–​422, 2006.
24. Robinson, C. Forgeries of Memory and Meaning: Blacks and the Regimes of Race in American Theater
and Film before World War II. New Edition. Chapel Hill, NC: The University of North Carolina
Press, 2007.
25. Rowan, C. It didn’t start with Jimmy the Greek. Washington Post January 21, 1988; Range, P.R.What
we say, what we think. U.S. News & World Report 104: 27–​28, 1988.
26. Rowe, J. The Greek chorus: Jimmy the Greek got it wrong but so did his critics. The Washington
Monthly 20: 31–​34, 1988.
27. Runstedtler, T. Jack Johnson, Rebel Sojourner: Boxing in the Shadow of the Global Color Line. Berkeley
& Los Angeles, CA: University of California Press, 2013.
28. Pearson, R. Shockey on Eugenics and Race. Washington, DC: Townsend Publishers, 1992.
29. Smith, M. Giving the Olympics an anthropological once-​over. Life 57: 80–​84, 1964.
30. Tanner, J.M. The Physique of the Olympic Athlete: A Study of 137 Track and Field Athletes at The XVII
Olympic Games, Rome, 1960. London: G. Allen and Unwin, 1964.
31. Time. Sport: the black dominance. Time 109: 57–​60, 1977.
32. Uhlig, M. Racial remarks cause furor. New York Times, January 16, 1988.
33. Weinbaum, W. The legacy of Al Campanis. ESPN.com April 1, 2012.
34. Wiggins, D.K. Black athletes in white men’s games: race, sport, and American national pastimes. The
International Journal of the History of Sport 31: 181–​202, 2014.
35. Wilber, R.L and Pitsiladis, Y.P. Kenyan and Ethiopian distance runners: what makes them so
good? International Journal of Sports Performance 7: 92–​102, 2012.

441
442

31
THE SCIENTIFIC AND ETHICAL
CHALLENGES OF USING
GENETIC INFORMATION
TO PREDICT SPORT
PERFORMANCE
Andrew C. Venezia and Stephen M. Roth

Introduction
Children and adolescents are encouraged to participate in athletic competition for several
reasons. For many, the participation is simply an opportunity to engage in activities that are fun,
social, and healthy for mind and body. For others, whether internally or externally driven, their
goals and aspirations reach a much higher level, namely that of becoming an elite athlete with
the hope of securing a position on a professional or Olympic team. Behind these young athletes
are numerous stakeholders, including family members, coaches, and sponsors, who can be intim-
ately involved in the development of athletic potential, all with similar hopes of greatness for the
athlete, though perhaps for more self-​serving reasons. The culture and economy surrounding
athletic excellence has produced a demand for early talent identification, the process of rec-
ognizing a propensity for athletic excellence, to streamline the process of talent development
ensuring that resources are efficiently utilized and future stars have a chance to develop. But the
efficacy of talent identification strategies is limited due to the complexity of physical and psy-
chological development in children, as well as numerous other factors, including the sport, types
of evaluations, and interpretation of data (3, 6). Because the financial and social consequences
for helping to develop an elite athlete are so high, talent identification remains an area of great
interest for many in the world of sports.
Ideally, stakeholders would be able to predict future athletic performance using physiological,
psychological, and genetic characteristics, which would be measurable well before adulthood
with both high validity and predictive value. While the intricate balance between nature
(genetic predisposition) and nurture (environmental factors) is open to debate, it is generally
recognized that not all young athletes have the potential to progress to elite status, whether
due to limited innate talent or simply lack of training opportunities (time, access to facilities,
etc.). Coaches and sport organizations must therefore carefully select the targets for enhanced
training opportunities, hoping to focus the best of training resources on those athletes most
likely to have the innate traits needed for success. Early talent identification is an especially
more efficient approach to talent development compared to treating all young athletes equally
until a few athletes separate themselves from the pack in young adulthood. Instead of including

442
443

Prediction of sport performance

all young athletes in a talent development program, only the best young athletes are included,
which allows those athletes with presumably the most potential for elite status in adulthood
to be exposed to the highest level of training opportunities (3). While exceptions may exist,
in most athletic endeavors a certain level of genetic advantage is required to reach elite status
(17). Similarly, having a genetic advantage alone is not sufficient to become a champion unless
combined with an optimal training environment (17).With this in mind, one can see the utility
in identifying the youth athlete with the ideal genetic makeup and placing them in the ideal
environment to have the best chance of developing an elite athlete.
Although the entire process of developing an elite athlete is challenging, the talent iden-
tification process itself presents a unique set of challenges. By necessity, talent identification
models will vary by sport, yet most will contain a combination of tests that assess anthropo-
metric, physiological, and cognitive/​psychological variables, as well as certain sport-​related
technical skills (29). Beyond the difficulties of testing the complex combination of skills and
abilities required in certain sports, there are several major obstacles to identifying youth athletes
with the potential for success. For example, the physical and psychological development of
children does not follow a uniform trajectory. Biological age and chronological age can differ
dramatically, giving some precocious youth an advantage over later-​developing peers, though
those differences tend to minimize over time (29). As such, testing in children will necessarily
provide advantage to those who are more physically or mentally developed at that chrono-
logical age, regardless of the predictive value of those traits later in life. Because many youth
sports are separated by age groups, talent development programs tend to select athletes that
are born early in the selection year, resulting in more mature youth athletes being identified as
“talented” (47), a concept known as the relative age effect. Although biobanding, the process
of grouping youth athletes based on physical and psychological maturation instead of chrono-
logical age alone, is currently being explored and utilized by certain sporting organizations, this
concept is not fully optimized and requires more investigation to best determine its usefulness
in different sports (12). Obstacles such as these increase the likelihood of traditional talent
identification models resulting in tremendous error in the form of false positives and false
negatives. In contrast to most physiological and psychological traits, the DNA sequence does
not change over time or experience day-​to-​day variability depending on the health, training
status, or mental state of the athlete, and therefore potentially offers a steadfast glimpse at ath-
letic potential. Rather than relying on a snapshot of a child’s performance at age 9 years, for
example, a genetic analysis could provide insights into physical and mental traits that are likely
to be present in the fully developed young adult, regardless of the developmental trajectory
those traits take during childhood. Advances in genetic testing have now made it relatively
cheap and easy to evaluate gene sequence variation in individuals; the challenge is knowing
which gene sequences to examine and how to incorporate these into talent identification
programs. Of note, the fast-​developing field of epigenetics represents a unique complexity to
this idea, as epigenetic modification of the DNA sequence may impart influences on gene
expression not observable with a traditional genetic test. We will save discussion of this emer-
ging field for the future when additional evidence of the importance of epigenetics in this
context becomes available.
At any one time during the 2017 National Basketball Association (NBA) Finals, up to 40%
of players on the court had a father that also played in the NBA.With this information alone, we
cannot separate the influence of nature versus nurture in these athletes, yet this remarkably high
percentage certainly indicates a familial contribution of some kind. Indeed, twin and family
studies have shown that key sport-​related traits are heritable, such as height (38), somatotype
(30), aerobic capacity and trainability (4, 5), strength, power (10), muscle fiber type (22), and

443
444

A.C. Venezia and S.M. Roth

even elite athletic status (13).Yet despite knowing the heritability of many athletic traits and the
considerable research dedicated to finding genetic markers relevant to sport performance (and,
thus, talent identification), identifying specific genetic predictors of elite athletic performance
has been exceedingly challenging. This is especially true in sports such as basketball or soccer
that require a unique combination of aerobic and anaerobic endurance, anaerobic power, sport-​
specific skills, and cognitive abilities such as game understanding and decision-​making, in add-
ition to team dynamics.
As earlier chapters in the present book have discussed, several innovative, large-​scale genetic
investigations of sport-​related traits are currently underway and have the potential to elucidate
some of the current unknowns in the genetics of elite athletic potential (43). With these and
future studies utilizing new genetic tools, it seems possible that one day evidence-​based genetic
analysis during childhood or even prenatally could be efficacious and perhaps even ubiquitous
in talent identification programs. The incorporation of evidence-​based genetic analysis into
talent identification and development processes would likely improve outcomes compared to
current programs, and may even improve prediction of athletes predisposed to serious injury or
activity-​related mortality (41). However, this ability to make such predictions presents a set of
ethical issues that must be considered and directly addressed. In the remaining sections of the
present chapter, we aim to address these ethical issues, many of which are relevant beyond sport,
as the direct-​to-​consumer (DTC) genetic testing market is rapidly expanding in numerous
health-​and medicine-​related areas.

Potential for using genetic testing for talent identification

Although research over the past several decades has been devoted to identifying the heritability
and genetic factors underlying traits related to athletic performance (19, 25, 31, 34, 36, 37, 46),
we are still in the infancy of the field as new tools and large-​scale collaborative projects are being
developed. The intense focus on genetic predictors of athletic performance has led to the iden-
tification of several genetic markers and polygenic profiles associated with athletic performance;
however, the field is rife with underpowered studies and unreplicated findings. A recent inves-
tigation of available literature found 155 genetic markers associated with elite sport perform-
ance (1), yet even the most consistent of associated genetic markers account for only a small
variance (e.g., 1–​2% at most) in athletic performance and are limited to determining likelihood
of success in endurance or sprint/​power activities (27) rather than sport performance itself.
For example, the R allele of the alpha-​actinin-​3 (ACTN3) R577X variant is advantageous for
power-​oriented events (see Chapter 23). While this does not exclude the possibility of an indi-
vidual with the X allele from participating in power-​oriented events (26), there is certainly an
association between the R allele and elite power performance (28). A child with this allele is
hypothetically more likely to excel in events such as powerlifting and sprinting. Similarly, the
angiotensin-​converting enzyme (ACE) I/​I genotype is associated with greater endurance per-
formance while the ACE D/​D genotype is associated with strength and power performance
(27) and this genotype could be similarly used for prediction in young athletes (see Chapter 17).
Importantly, sports like basketball, boxing, soccer, rugby, and American football are characterized
by a combination of aerobic and anaerobic qualities, along with numerous other physical and
psychological characteristics. Using genetic profiling to predict success in sports with complex
physical and psychological requirements is a daunting task. The number of possible polygenic
profiles contributing to an elite boxer or soccer player is likely enormous. To add to the diffi-
culties of identifying polygenic profiles in elite athletes, we are limited by the number of “elite”
athletes on the planet, with even fewer available or willing to participate in genetic research.

444
445

Prediction of sport performance

Add to this limitation that sex and race are also important considerations, and the sample pool
is reduced even more. Breitbach et al. (6) suggest that utilizing the GWAS approach, roughly
100,000 elite athletes from a defined discipline would be necessary to explain 15% of variance
in elite performance in that specific sport discipline. Even if there were 100,000 elite athletes in
any specific sport discipline, 85% of the variance in elite performance would remain undefined.
Chapter 25 of this volume outlines some of the efforts of the field to improve the ability to
study elite athletes and more effectively identify contributing genetic factors.
Considering the above information, we do not believe that the available evidence is strong
enough to responsibly favor, or especially exclude, young athletes from participating in specific
sports based on their genetic makeup.This opinion is shared by several researchers in the field (8,
18, 24, 33, 41, 44, 45). In a 2015 consensus statement, Webborn et al. (44) argue against the use
of genetic testing for talent identification by stating “in the current state of knowledge, no child
or young athlete should be exposed to direct-​to-​consumer genetic testing to define or alter
training or for talent identification aimed at selecting gifted children or adolescents.” Similarly,
a position statement from the Australian Institute of Sport in 2017 (40) indicated that “despite
the correlation between some genes and elite athletic performance, there is no scientific evi-
dence for the predictive value of genetic profiling in sports performance.” Despite this general
consensus on the present inadequacies of genetic prediction in talent identification, the next
decades of collaborative high-​quality investigations with large populations of elite-​level athletes
may ultimately provide the ability to use polygenic prediction models to identify potential for
success. Moreover, advances in genetic technologies are leading to remarkable prenatal and
preimplantation screening tools to identify potential genetic disorders and select embryos for
certain traits (39). This preimplantation genetic diagnosis is primarily used to identify embryos
with genetic disorders to avoid transmission to offspring; however, this same technology could
be used to select specific advantageous polygenic profiles for athletic excellence. As such, dis-
cussion of the ethical issues surrounding such predictive tests is important.

Potential benefits of using genetic testing for talent identification

Assuming the scientific evidence advances to the point where talent can be predicted using
genetic techniques, there are several benefits to this approach. As with other talent identifi-
cation methods, early identification of athletes with potential for greatness provides young
athletes with the opportunity to dedicate years to sport-​specific deliberate practice.The volume
of 10,000 hours of dedicated training is often referenced and some have indicated that at least
10 years of dedicated practice is necessary to achieve peak performance in various endeavors,
including sports (14). Considering the age at which many athletes enter professional or Olympic
competition, the 10,000 hours and/​or 10 years of dedicated training needs to begin at a young
age. It is also recognized that family support is necessary for reaching the highest levels of ath-
letic competition (11). Parents may face several obstacles (financial, geographic, time limitations,
etc.) to providing resources necessary for success (coaching, equipment, etc.). Therefore, parents
interested in supporting an elite-​level athlete need to carefully select the activity with the most
likelihood for success. Stakeholders want to make certain that they are devoting their limited
resources to the activity that will prove to be most fruitful. If genetic screening identifies that a
child may be uniquely capable for success in a specific sport domain, parents may be more likely
to make sacrifices for the child to succeed in that domain.
Unlike traditional methods of talent identification, genetic approaches could allow for selec-
tion of athletic superiority well before any special physical attributes present themselves. As
outlined above, determining the potential for success in children is difficult given different

445
446

A.C. Venezia and S.M. Roth

developmental patterns. Genetic talent identification could avoid the relative age effect, where
more physically mature athletes are selected for talent development programs, since the genetic
code is not sensitive to the developmental variation of physical and psychological phenotypes.
Using heritability of traits to predict athletic potential most certainly already occurs without
the use of sophisticated genetic analytic tools. For example, height and arm span are important
contributors to success in basketball (20). Suppose we compare two young basketball players
who are equal in stature and skill, yet one has two parents with above-​average height and wing-
span, while the other has two parents with below-​average height and wingspan. A coach would
be wise to presume that the player with the “ideal” parents will develop into a more basketball-​
appropriate body type and will have more potential for future success. That is not to say that
the young athlete with the shorter parents cannot develop into an outstanding basketball player,
but they will likely not develop the ideal body type for optimal success in basketball and would
likely have to make up for the limitation with other physical, psychological, and skill-​related
qualities. This is an ideal scenario with each athlete having two parents with similar statures and
a sport with a certain body type associated with success. Other sports may not align so clearly
with a specific body type. For example, soccer, wrestling, judo, sprinting, and other sports are
not so clearly associated with a specific body type and the use of genetic screening along with
nongenetic talent identification techniques could prove useful for talent selection.
Another considerable benefit of genetic testing of young athletes is the identification of
individuals who are at increased risk of injury, since time lost to injury can derail the trajectory
of athletic development. Consider how losing months or years of deliberate practice due to a
major ligament injury during the teen years could significantly affect the necessary 10 years of
dedicated practice to the specific sport discipline. Moreover, time lost to injury can interfere
with success in team sports, making this a major concern for coaches and other stakeholders
in collegiate, professional, and Olympic team events. Some of these polymorphisms have been
described in Chapters 27 and 28. By identifying genetic determinants of soft tissue injuries,
for example, adequate focus on preventative training strategies can be provided for predisposed
individuals. However, while such testing may offer a valuable tool for reducing risks associated
with sporting activities, it certainly presents several ethical considerations. For one, how this
information is shared and used can be challenging. For example, can a third party (e.g., coaches
or team owners) request this genetic information and use it in recruiting or contract negoti-
ations? Parents and coaches likely lack the expertise and/​or training to fully understand the
genetic information and may not recognize the limitations of the test results. Suppose an ath-
lete has a 10% increased risk for developing a concussion or soft tissue injury. What does a 10%
increased risk mean in the context of sport and injury and is this worth specific attention (40) or
possible exclusion? Equally important is ensuring confidentiality and procedures for secondary
findings, a concern that will be addressed later in this chapter. What today is associated with
increased soft tissue injuries may be found to be associated with other conditions unrelated to
sport performance in the future.
Other conditions with more dire consequences may also be identified by genetic screening
and this may be a more appropriate use of the tool. In cases such as hypertrophic cardiomyop-
athy and other inherited cardiac conditions, genetic screening has the potential to save lives by
identifying individuals with an increased risk of sudden cardiac death. Other conditions such
as Marfan syndrome, which is associated with the tall, slender build and long limbs generally
associated with success in basketball and volleyball, is a genetic condition that increases the risk
of sudden cardiac death during high-​intensity exercise. Importantly, if genetic screenings are
to be used for identifying inherited life-​threatening cardiac conditions and Marfan syndrome,
they should be ordered by a medical professional upon symptoms or a family history of the

446
447

Prediction of sport performance

disease (41). In cases such as these, the medical professional can provide the necessary genetic
counseling to help the individual and/​or family understand the results of the genetic testing.
Increasing safety in sports is paramount and genetic screenings have the potential to help in
the prevention of injuries and sudden events that threaten the life of the athlete. However, the
available evidence is still insufficient to rely on genetic screening for determining injury predis-
position (41) and it is not likely that many sports medicine practitioners will have the necessary
level of expertise in medical genetics to fully understand and interpret the findings of a genetic
test (40). Moreover, mandating any type of health-​related screening of heritable traits carries
ethical concerns. As an example, the mandate by the National Collegiate Athletic Association
in the US that all student athletes need to be screened for sickle cell trait, provide evidence of
their sickle cell trait status, or sign a waiver to avoid screening has been met with controversy
and disapproval from several organizations, even though the goal of the program was designed
to enhance student athlete safety (15).This mandate, in particular, is complicated by the consid-
erably higher fraction of individuals of African descent carrying the sickle cell trait, thus poten-
tially opening doors for racial discrimination in the testing process.

Potential drawbacks of using genetic testing for talent identification

Despite some positive aspects of including genetic testing in talent identification programs,
there are a number of potential negatives to such use. For genetic talent identification strategies
to be successful, sampling and determination of the genetic profile needs to occur at a young
enough age to reap the benefits of targeted resources and training, which means such testing
necessarily occurs in children. In fact, prenatal or preimplantation (if in vitro fertilization) genetic
profiling is arguably optimal for predicting future likelihood of success, injury, or disease, and
although this tool is unlikely to be used for athletics anytime soon, it is currently feasible and
will likely become more commonplace for diagnosis of genetic disorders (39). Whether genetic
profiling occurs during childhood or prenatally, the identified “prodigy” can receive the familial
support, coaching, and resources necessary for elite proficiency. Because this determination is
made in childhood, however, the decisions made by parents and coaches may result in the loss
of the child’s autonomy and right to an open future (9). This is a complex issue concerning
genetic testing for elite performance and is not restricted to the competitive sports world. Early
identification of genetic predisposition for elite proficiency in any specific domain will cer-
tainly allow for mentors/​instructors/​coaches and parents to promote and encourage the devel-
opment of that talent. Coaches and parents are fundamental in the cultivation of talent since a
young child does not have the experience or maturity to carry the responsibility of knowing
what is best for them. Indeed, talent must be nurtured, and parents must do most of this nurt-
uring, yet children depend on parents to provide them with a wide breadth of experiences to
develop into a well-​rounded individual with the capacity to make their own life choices (9).
The potential for aggressive parental control over a child’s activities may be increased if genetic
screening indicates that the child is “uniquely” capable of achieving excellence, especially if the
parents do not recognize the limitations of these genetic predictions. The aggressive control of
athletic endeavors may not only come at the expense of other sporting activities but also at the
expense of other childhood experiences. According to Camporesi (9), it is perfectly reason-
able for parents to expose children to sporting activities within their own financial situation
and personal preferences but unreasonable to actively discourage other sporting opportunities.
Although early specialization is becoming more commonplace, research suggests that partici-
pating in several sports at a young age is the best strategy to develop athletic ability (11, 16, 21).
Early specialization has the potential to lead to overuse injuries and burnout (7, 21). Importantly,

447
448

A.C. Venezia and S.M. Roth

if the child achieves the ultimate goal of professional or Olympic status, the sacrifices made may
have been worth it; however, consider the enormous percentage of athletes who do not achieve
that level of success. For this large percentage of young athletes, they will have missed out on
several childhood experiences, potentially with little to show for their dedication.
Another challenge is the issue of consent in genetic testing. It can be argued that every indi-
vidual has the right to know their genetic profile since it belongs to them (23); however, this
is a complex issue, especially when considering talent identification in young athletes. For one,
youth athletes under 18 years of age do not have the capacity to provide informed consent.
Indeed, a recent position statement from the Australian Institute of Sport suggests that genetic
testing for sport is only appropriate if the individual being tested is 18 years of age (40). In fact,
due to the complexities and technical nature of genetic testing, it might be difficult for adults to
provide adequate informed assent for their child without a certain level of genetic counseling
prior to the genetic screening. Whether a legal adult or young athlete, all genetic screening
should be voluntary, requested by the individual to be tested (or medical professional in the
patient–​provider relationship), and free from any form of coercion. This may be difficult in the
case of young athletes who are at a significant power disparity and may feel pressured to submit
to a genetic test if requested by a coach, team owner, or parent. If genetic testing is included in
the process of talent identification for entry into talent development programs, young athletes
that decline participation in the testing may be discriminated against. Athletes should be allowed
to decline genetic testing and have no penalties or negative consequences as a result of that
decision.
If young athletes are undergoing medically supervised genetic screenings, parents, coaches,
and athletes should have access to experts and trained professionals who can sufficiently and
accurately explain the results as well as provide necessary genetic counseling. In fact, prior
to any genetic screening, individuals should undergo genetic counseling (40), which should
include information about the purposes and consequences of testing, limitations, potential
harms, etc. (35). Prescreening genetic counseling is helpful for providing enough information
for the potential subject to provide informed consent. Concerning genetic screening for risks
of sports injury, a trained professional can provide interpretation of the results and potential
courses of action based on the findings. Similarly, if testing for athletic potential, the trained
professional can provide facts and details concerning the relevance and limitations of the gen-
etic profile. Considering the status of the field at the time of writing this chapter, the role of the
genetic counselor would likely be to advise against this type of screening or thoroughly explain
the limitations of the results.
Although sport and exercise genetics researchers generally agree that the evidence is insuffi-
cient for using genetic tests for talent identification, several companies sell the promise of using
this technology to determine potential for athletic success and for designing individualized
exercise training programs. This promise comes in the form of DTC genetic testing. These
products can be ordered online and delivered to the home, providing the capacity to sidestep
the medical supervision of a physician or trained professional. The regulations for DTC tests
vary from country to country and, while the quality and regulation of these tests have generally
improved over the past several years (2), the quality control of tests for sporting potential is cur-
rently lacking. Although there is considerable turnover of companies in this industry, an investi-
gation in 2015 identified 39 companies selling DTC genetic tests for sport, fitness, or injury risk
(44). The majority of these tests did not indicate the specific DNA markers being tested. If we
consider the proposition that everyone has the right to know their own genetic code (23), tests
that do not report the specific genetic markers being tested do not provide the ability to know.
These tests only provide the buyer with an interpretation of findings, which cannot be verified

448
449

Prediction of sport performance

by an expert in the field. The 2015 investigation also found that the two most commonly
tested variants were ACE I/​D and ACTN3 R557X (44). While these two genetic markers do
have scientific support (as outlined in previous chapters), they are certainly not strong enough
inform life-​plans for young athletes. Because sport and exercise genetic screens do not qualify as
medical advice or disease risk assessment, these DTC tests may avoid the regulation that limits
the use of other health-​related genetic screenings. Therefore, the claims made concerning pre-
dictive potential of sport performance, response to training, and/​or injury risk are not under
the same scrutiny as tests for genetic risk of breast cancer or Alzheimer’s disease, for example.
Without such regulation, companies continue to market the ability to predict sport perform-
ance and training adaptations even though the scientific evidence does not support such con-
clusive claims. Although not all companies advertise to parents and coaches (42), the ability of
parents to purchase these tests and obtain a buccal cell sample from their child is uncomplicated.
Informed consent and genetic counseling can easily be avoided and the results of these screens
may be used to influence the sport participation and training decisions for a child even though
the tests have virtually no predictive capacity.
The potential for inappropriate discrimination and exclusion is a major concern in the
era of DTC genetic testing for sport performance. With the two most common markers for
athletic potential being ACTN3 R577X and ACE I/​D, which indicate some propensity for
success in endurance and power activities, one can imagine the number of children who
could be eliminated from participating in certain sports due to having the “wrong” geno-
type. Interestingly, a case report from 2007 indicated that an Olympic long jumper carried the
ACTN3 XX genotype (26), which should indicate limited potential for success in sprint and
power-​related sports (e.g., long jump). If this individual had underwent a genetic screening to
predict athletic potential as a child, he likely would have missed out on winning a gold medal
in the under-​17 category of the world championships and the opportunity to represent his
country in two Olympic Games. This athlete is an excellent case study to highlight the import-
ance of recognizing the limitations of using single genetic markers to guide sport participation
choices, the importance of genetic counseling, and the inappropriate marketing techniques of
some DTC genetic testing companies in the sport arena.
Another issue of concern with genetic testing is the shared nature of DNA in families.
A genetic test can provide information about family members who did not assent to obtaining
or to sharing genetic information with others. This is particularly true for parents and siblings,
especially monozygotic twins. These individuals may not have been involved in the consent
and genetic counseling process, nor would they have undergone a careful evaluation of the
implications of the testing. This is a mostly unavoidable consequence of genetic screening in
general and certainly extends to genetic testing for sport. And while the genetic test results
for sport performance may not appear inherently meaningful for family members, secondary
findings from genetic testing are a concern. We can use the example of the gene apolipoprotein
E (APOE) as a case in point for both unintended secondary findings and familial-​shared infor-
mation risks. The APOE ε4 allele has consistently been shown as a genetic risk factor for
late-​onset Alzheimer’s disease (AD), while the APOE ε2 allele is generally considered pro-
tective against AD. Unfortunately, some DTC companies have provided APOE genotyping to
inform individuals about their likelihood of suffering and recovering from a concussion, which
is an association that is far from being scientifically confirmed, with only limited investigations
focused on this question to date (see Chapter 27). In this case, a DTC test result focused on
a person’s risk for sport-​related injury is in fact far more informative for AD risk, and a quick
Internet search would show the linkage of APOE genotype with disease risk. There are con-
siderable psychological consequences to such an unintended discovery. Moreover, now the

449
450

A.C. Venezia and S.M. Roth

athlete who likely does not have the necessary training to understanding the full consequences
of the association of APOE genotype with AD also has information concerning AD risk of
parents, siblings, and future children. Could this change the individual’s behavior toward family
members? Will those family members want to know this information? Thus, both secondary
findings and shared genetic information are both potential hazards in any form of genetic
testing. APOE is an ideal and arguably extreme example, but even recent research on ACTN3
is finding that it is associated with exercise recovery and risk of injury (32). An individual may
provide consent for genetic testing to understand their propensity for success in a speed-​related
event but then find out that they are at increased risk for injury, information they may not have
been anticipating.

Conclusion
In this chapter, we have outlined a variety of scientific and ethical challenges to genetic testing
for sport performance prediction.While the present scientific evidence for such testing is highly
limited and numerous position stands argue against the use of genetic testing in talent identi-
fication programs, the expansion of DTC genetic tests and lack of regulation over such testing
in the area of sport performance mean that athletes and/​or their parents and other stakeholders
will have the option of pursuing such information with the potential for considerable negative
consequences. And while we can envision a day when the scientific evidence is strengthened and
the inclusion of genetic testing for talent identification is arguably justified, significant ethical
issues will remain in the use of such testing, particularly in young (or unborn) children. While
these are important conversations to have with parents, and some medical professionals may
be informed enough to have such conversations, the likelihood that parents will be informed
participants generally is unlikely. Ultimately, it will be the organizations that oversee talent iden-
tification and sport development that will need to provide oversight and protect young athletes
by preventing or carefully limiting the use of such genetic screening. As a field, we must push
for position stands and chapters like these to be provided to the leadership within these sport
organizations and federations to better ensure the safety and autonomy of young athletes, which
will ultimately benefit both the athletes and sport more generally by emphasizing an informed
and evidence-​based approach for talent identification.

References
1. Ahmetov II, Egorova ES, Gabdrakhmanova LJ and Fedotovskaya ON. Genes and Athletic
Performance: An Update. Medicine and Sport Science 61: 41–​54, 2016.
2. Allyse MA, Robinson DH, Ferber MJ and Sharp RR. Direct-​to-​Consumer Testing 2.0: Emerging
Models of Direct-​to-​Consumer Genetic Testing. Mayo Clinic Proceedings 93: 113–​120, 2018.
3. Bergeron MF, Mountjoy M, Armstrong N, Chia M, Côté J, Emery CA, Faigenbaum A, Hall G,
Kriemler S, Léglise M, Malina RM, Pensgaard AM, Sanchez A, Soligard T, Sundgot-​Borgen J, van
Mechelen W, Weissensteiner JR and Engebretsen L. International Olympic Committee Consensus
Statement on Youth Athletic Development. British Journal of Sports Medicine 49: 843–​851, 2015.
4. Bouchard C, An P, Rice T and Skinner JS. Familial Aggregation of VO2max Response to Exercise
Training: Results from the HERITAGE Family Study. Journal of Applied Physiology 87: 1003–​1008, 1999.
5. Bouchard C, Daw EW, Rice T, Pérusse L, Gagnon J, Province MA, Leon AS, Rao DC,
Skinner JS and Wilmore JH. Familial Resemblance for VO2max in the Sedentary State: The
HERITAGE Family Study. Medicine & Science in Sports & Exercise 30: 252–​258, 1998.
6. Breitbach S, Tug S and Simon P. Conventional and Genetic Talent Identification in Sports: Will
Recent Developments Trace Talent? Sports Medicine 44: 1489–​1503, 2014.

450
451

Prediction of sport performance

7. Brenner JS and AAP Council on Sports Medicine and Fitness. Sports Specialization and
Intensive Training in Young Athletes. Pediatrics 138: e20162148, 2016.
8. Camporesi S and McNamee MJ. Ethics, Genetic Testing, and Athletic Talent: Children’s Best
Interests, and the Right to an Open (Athletic) Future. Physiological Genomics 48: 191–​195, 2016.
9. Camporesi S. Bend it like Beckham! The Ethics of Genetically Testing Children for Athletic
Potential. Sport, Ethics and Philosophy 7: 175–​185, 2013.
10. Costa AM, Breitenfeld L, Silva AJ, Pereira A, Izquierdo M and Marques MC. Genetic
Inheritance Effects on Endurance and Muscle Strength: An Update. Sports Medicine 42: 449–​458, 2012.
11. Côté J. The Influence of the Family in the Development of Talent in Sport. The Sport Psychologist
13: 395–​417, 2009.
12. Cumming SP, Lloyd RS, Oliver JL, Eisenmann JC and Malina RM. Bio-​ banding in
Sport: Applications to Competition, Talent Identification, and Strength and Conditioning of Youth
Athletes. Strength and Conditioning Journal 39: 34–​47, 2017.
13. De Moor MHM, Spector TD, Cherkas LF, Falchi M, Hottenga J-​J, Boomsma DI and De
Geus EJC. Genome-​Wide Linkage Scan for Athlete Status in 700 British Female DZ Twin Pairs.
Twin Research and Human Genetics 10: 812–​820, 2007.
14. Ericsson KA. Deliberate Practice and Acquisition of Expert Performance: A General Overview.
Academic Emergency Medicine 15: 988–​994, 2008.
15. Ferrari R, Parker LS, Grubs RE and Krishnamurti L. Sickle Cell Trait Screening of Collegiate
Athletes: Ethical Reasons for Program Reform. Journal of Genetic Counseling 24: 873–​877, 2015.
16. Fransen J, Pion J, Vandendriessche J, Vandorpe B, Vaeyens R, Lenoir M and Philippaerts RM.
Differences in Physical Fitness and Gross Motor Coordination in Boys Aged 6–​12 Years Specializing
in One Versus Sampling More Than One Sport. Journal of Sports Sciences 30: 379–​386, 2012.
17. Georgiades E, Klissouras V, Baulch J, Wang G and Pitsiladis Y. Why Nature Prevails Over
Nurture in the Making of the Elite Athlete. BMC Genomics 18: 835, 2017.
18. Guth LM and Roth SM. Genetic Influence on Athletic Performance. Current Opinion in Pediatrics
25: 653–​658, 2013.
19. Hagberg JM, Rankinen T, Loos RJF, Pérusse L, Roth SM, Wolfarth B and Bouchard C.
Advances in Exercise, Fitness, and Performance Genomics in 2010. Medicine & Science in Sports &
Exercise 43: 743–​752, 2011.
20. Hoare DG. Predicting Success in Junior Elite Basketball Players –​The Contribution of
Anthropometric and Physiological Attributes. Journal of Science and Medicine in Sport 3: 391–​405, 2000.
21. Jayanthi N, Pinkham C, Dugas L, Patrick B and Labella C. Sports Specialization in Young
Athletes: Evidence-​Based Recommendations. Sports Health 5: 251–​257, 2013.
22. Komi PV, Viitasalo JH, Havu M, Thorstensson A, Sjödin B and Karlsson J. Skeletal Muscle
Fibres and Muscle Enzyme Activities in Monozygous and Dizygous Twins of Both Sexes. Acta
Physiologica Scandinavica 100: 385–​392, 1977.
23. Loi M. Direct to Consumer Genetic Testing and the Libertarian Right to Test. Journal of Medical
Ethics 42: 574–​577, 2016.
24. Loland S. Against Genetic Tests for Athletic Talent: The Primacy of the Phenotype. Sports Medicine
45: 1229–​1233, 2015.
25. Loos RJF, Hagberg JM, Pérusse L, Roth SM, Sarzynski MA, Wolfarth B, Rankinen T and
Bouchard C. Advances in Exercise, Fitness, and Performance Genomics in 2014. Medicine & Science
in Sports & Exercise 47: 1105–​1112, 2015
26. Lucia A, Oliván J, Gómez-​Gallego F, Santiago C, Montil M and Foster C. Citius and Longius
(Faster and Longer) with no Alpha-​Actinin-​3 in Skeletal Muscles? British Journal of Sports Medicine
41: 616–​617, 2007.
27. Ma F, Yang Y, Li X, Zhou F, Gao C, Li M and Gao L. The Association of Sport Performance with
ACE and ACTN3 Genetic Polymorphisms: A Systematic Review and Meta-​Analysis. PLoS ONE
8: e54685, 2013.
28. Papadimitriou ID, Lucia A, Pitsiladis YP, Pushkarev VP, Dyatlov DA, Orekhov EF, Artioli
GG, Guilherme JPLF, Lancha AH, Ginevičienė V, Cieszczyk P, Maciejewska-​Karlowska
A, Sawczuk M, Muniesa CA, Kouvatsi A, Massidda M, Calò CM, Garton F, Houweling PJ,
Wang G, Austin K, Druzhevskaya AM, Astratenkova IV, Ahmetov II, Bishop DJ, North
KN and Eynon N. ACTN3 R577X and ACE I/​D Gene Variants Influence Performance in Elite
Sprinters: A Multi-​Cohort Study. BMC Genomics 17: 285, 2016.

451
452

A.C. Venezia and S.M. Roth

29. Pearson DT, Naughton GA and Torode M. Predictability of Physiological Testing and the Role
of Maturation in Talent Identification for Adolescent Team Sports. Journal of Science and Medicine in
Sport 9: 277–​287, 2006.
30. Peeters MW, Thomis MA, Loos RJF, Derom CA, Fagard R, Claessens AL, Vlietinck RF and
Beunen GP. Heritability of Somatotype Components: A Multivariate Analysis. International Journal
of Obessity 31: 1295–​1301, 2007.
31. Pérusse L, Rankinen T, Hagberg JM, Loos RJF, Roth SM, Sarzynski MA, Wolfarth B and
Bouchard C. Advances in Exercise, Fitness, and Performance Genomics in 2012. Medicine & Science
in Sports & Exercise 45: 824–​831, 2013.
32. Pickering C and Kiely J. ACTN3: More than Just a Gene for Speed. Frontiers in Physiology
8: 1080, 2017.
33. Pitsiladis Y, Wang G, Wolfarth B, Scott R, Fuku N, Mikami E, He Z, Fiuza-​Luces C, Eynon
N and Lucia A. Genomics of Elite Sporting Performance: What Little We Know and Necessary
Advances. British Journal of Sports Medicine 47: 550–​555, 2013.
34. Rankinen T, Roth SM, Bray MS, Loos R, Pérusse L, Wolfarth B, Hagberg JM and Bouchard
C. Advances in Exercise, Fitness, and Performance Genomics. Medicine & Science in Sports & Exercise
42: 835–​846, 2010.
35. Rantanen E, Hietala M, Kristoffersson U, Nippert I, Schmidtke J, Sequeiros J and Kääriäinen
H. What is Ideal Genetic Counselling? A Survey of Current International Guidelines. European
Journal of Human Genetics 16: 445–​452, 2008.
36. Roth SM, Rankinen T, Hagberg JM, Loos RJF, Pérusse L, Sarzynski MA, Wolfarth B and
Bouchard C. Advances in Exercise, Fitness, and Performance Genomics in 2011. Medicine & Science
in Sports & Exercise 44: 809–​817, 2012.
37. Sarzynski MA, Loos RJF, Lucia A, Pérusse L, Roth SM, Wolfarth B, Rankinen T and
Bouchard C. Advances in Exercise, Fitness, and Performance Genomics in 2015. Medicine & Science
in Sports & Exercise 48: 1906–​1916, 2016.
38. Silventoinen K, Magnusson PKE, Tynelius P, Kaprio J and Rasmussen F. Heritability of
Body Size and Muscle Strength in Young Adulthood: A Study of One Million Swedish Men. Genetic
Epidemiology 32: 341–​349, 2008.
39. Vermeesch JR, Voet T and Devriendt K. Prenatal and Pre-​Implantation Genetic Diagnosis.
Nature Reviews Genetics 17: 643–​656, 2016.
40. Vlahovich N, Fricker PA, Brown MA and Hughes D. Ethics of Genetic Testing and Research in
Sport: A Position Statement from the Australian Institute of Sport. British Journal of Sports Medicine
51: 5–​11, 2017.
41. Vlahovich N, Hughes DC, Griffiths LR, Wang G, Pitsiladis YP, Pigozzi F, Bachl N and Eynon
N. Genetic Testing for Exercise Prescription and Injury Prevention: AIS-​Athlome Consortium-​
FIMS Joint Statement. BMC Genomics 18: 818, 2017.
42. Wagner JK and Royal CD. Field of Genes: An Investigation of Sports-​Related Genetic Testing.
Journal of Personalized Medicine 2: 119–​137, 2012.
43. Wang G, Tanaka M, Eynon N, North KN, Williams AG, Collins M, Moran CN, Britton SL,
Fuku N, Ashley EA, Klissouras V, Lucia A, Ahmetov II, De Geus E, Alsayrafi M and Pitsiladis
YP. The Future of Genomic Research in Athletic Performance and Adaptation to Training. Medicine
and Sport Science 61: 55–​67, 2016.
44. Webborn N, Williams A, McNamee M, Bouchard C, Pitsiladis Y, Ahmetov I, Ashley
E, Byrne N, Camporesi S, Collins M, Dijkstra P, Eynon N, Fuku N, Garton FC, Hoppe
N, Holm S, Kaye J, Klissouras V, Lucia A, Maase K, Moran C, North KN, Pigozzi F and
Wang G. Direct-​ to-​
Consumer Genetic Testing for Predicting Sports Performance and Talent
Identification: Consensus Statement. British Journal of Sports Medicine 49: 1486–​1491, 2015.
45. Williams AG, Wackerhage H and Day SH. Genetic Testing for Sports Performance, Responses
to Training and Injury Risk: Practical and Ethical Considerations. Medicine and Sport Science
61: 105–​119, 2016.
46. Wolfarth B, Rankinen T, Hagberg JM, Loos RJF, Pérusse L, Roth SM, Sarzynski MA and
Bouchard C. Advances in Exercise, Fitness, and Performance Genomics in 2013. Medicine & Science
in Sports & Exercise 46: 851–​859, 2014.
47. Yan X, Papadimitriou I, Lidor R and Eynon N. Nature versus Nurture in Determining Athletic
Ability. Medicine and Sport Science 61: 15–​28, 2016.

452
453

32
GENE DOPING: ETHICAL
PERSPECTIVES
Verner Møller and Rasmus Bysted Møller

Since the turn of the 21st century, concerns about the impact of gene therapy on sports have
gathered speed and a number of thought-​provoking books have been produced. Andy Miah’s
ground-​breaking work, Genetically Modified Athletes: Biomedical Ethics,Gene Doping and Sport
(2004) drew attention to many problems and prospects that potentially would make anti-​
doping impotent if the ethically laden endeavour should not become unethical. For instance,
he mentions “that muscle biopsy may be the only way to detect some forms of genetic modi-
fication in sport, which would be a significant challenge for anti-​doping policy given the
invasiveness of such a procedure” (Miah, 2004, p. 34). If Miah’s assertion is correct that muscle
biopsies are needed to detect gene doping this would certainly, as he implies, pose a huge threat
to the integrity of anti-​doping.
Miah offers a thorough analysis of the implication of gene therapy for the human future.
Money is not invested in gene technology with a view to develop Frankenstein monsters but
because it has medical potential. It is hoped that it can result in medical breakthroughs leading
to remedies for currently incurable diseases. If gene technology makes it possible to speed up
injured athletes’ recovery as well, should athletes then be forbidden to take advantage of that?
Such scenarios make it clear that “genetic enhancement cannot be characterized in the same
way as other ways of doping” he rightly maintains (Miah, 2004, p. 178). The magnitude of the
ethical dilemmas that occurs with gene technology is such that one easily understands why
Miah does not take upon himself to search for an ethically viable way for sport to negotiate the
future he foresees. Instead he ends his book by proposing a tentative solution which, as provoca-
tive as it appears, leaves the ethical problem unresolved. “Perhaps one solution to this ethical
dilemma,” he professes, “is to create distinct, genetically enhanced competitions and to continue
voluntary submission to anti-​doping testing procedures” (Miah, 2004, p. 178).
Rather than solving the ethical dilemma, splitting sports into two distinctly different types of
sport amplifies the ethical problem. First, accepting a category of genetically enhanced compe-
tition is to accept that this category of sport in effect is transformed into a gene-​tech laboratory
where elite sport’s competitive nature, pursuit of perfection, and will to win, tempts ambitious
athletes to accept the role as willing guinea pigs for medical and other companies interested
in human enhancement. The only way this “solves” the ethical dilemma is by throwing ethics
out of the window. Second, having an alternative category of sport for athletes who want to

453
454

V. Møller and R.B. Møller

compete without having their bodies genetically modified only accessible for individuals who
consent to having regular biopsies taken is equally indefensible. It implies that athletes who do
not want to have their bodies tampered with in order to be competitive in the gene-​doping
category are forced to accept to have their bodies tampered with by a doping control officer’s
muscle fiber-​cutting needle.
The profoundness of the ethical problems left unresolved by Miah’s proposed solution was
reflected in Claudio Tamburrini and Torbjörn Tännsjö’s edited book Genetic Technology and
Sport: Ethical Questions (2005). This collection written by scholars and scientists from various
disciplines and with disparate attitudes to the prospects of genetic enhancement is an excellent
introduction to the field. It is beyond the scope of this chapter to present the breadth and depth
of the themes the book comprises but it is worthwhile quoting Christian Munthe’s contri-
bution about the ethics of controlling gene doping because it so clearly illustrates why a gene
doping-​free sports category with consensual testing is unjustifiable:

Due to the apparent need for quite extensive and repeated sample-​taking, health risks
due to the repeated application of biopsy procedures to various parts of the body have
to be taken quite seriously. The same goes, of course, for the direct and rather obvious
inconvenience of the athlete who has to be subjected to such procedures. Any pro-
gramme for controlling gene doping would therefore have to consider the risks and
problems created by the sample-​taking alone as an ethically highly relevant factor that
needs to be balanced against the reasons for enforcing and upholding the ban on gene
doping.
(Munthe, 2005, p. 114)

The ethical challenges Munthe highlights were not foreign to the World Anti-​Doping Agency
(WADA). The agency, established in 1999 in the wake of the 1998 revelations of organized and
vast doping used at the Tour de France, understood that the fast expansion of scientific research
in gene therapy around the world meant that a spillover to sport might be on the horizon.
Acknowledging that pre-​WADA anti-​doping efforts had been reactive with the unfortunate
consequence that regulations and test procedures and methods were playing a frustrating game
of catch-​up, the agency understood it had to be proactive if it should not be caught off guard
by this new threat to sport. So, no more than 3 years after its foundation, the year before the
first World Anti-​Doping Code was drafted, WADA organized a meeting in New York where
representatives from the athletic community were brought together with scientists and scholars
to discuss the various aspects of the gene doping issue. Among the participants were Angela
Schneider and Theodore Friedmann who subsequently wrote an introductory book Gene
Doping in Sports: The Science and Ethics of Genetically Modified Athletes (2006). Here they explain
the basics of gene therapy and the potential applications of this emerging science to sports and
offer some ethical considerations related to the issue. What is particularly laudable about this
book is its sobering approach. It does not entertain the reader with exciting Brave New World
scenarios. On the contrary, it shows how embryonic this field of science is, and presents some
of the severe setbacks it has faced in the form of unexpected fatal side effects in experiments
carried out on patients suffering from severe genetic diseases. There are still many miles to go
before promising experiments on mice will translate to successful clinical treatment in humans.
As Friedmann and Schneider explain, “A great deal of additional basic and applied medical
research will be required before anyone can say with comfort that gene transfer, even for such
dire diseases [as Parkinson’s, Alzheimer’s and Lou Gehring’s] is safe and effective” (Schneider
& Friedmann, 2006, p. 35). Gene doping is in all likelihood even farther away because the

454
455

Gene doping: Ethical perspectives

present state of the technology of gene transfer for application in therapy is not sufficiently
advanced “to conduct a gene transfer application for athletic performance that would meet
the requirements for ethical research in human research subjects” (Schneider & Friedmann,
2006, p. 39f). However, this does not rule out hazardous experiments the authors warn. “To
do a scientifically poor and dangerous study in humans would be far less difficult, probably less
successful, and certainly perilous for the athlete” (Schneider & Friedmann, 2006, p. 41).
For readers who wants a broader introduction than this chapter offers to the field of gene
doping –​and the ethical issues that evolve with it –​the abovementioned books are commendable.

Gene doping unthinkable –​but not that unthinkable

Schneider and Friedmann’s understanding resonates with our own opinion. We share the view
that gene doping is not around the corner. In fact, as late as 2018 one of the authors of this
chapter published a book which preface announced that:

we should note that the future of sport might be revolutionised if the spectre of
gene doping becomes a reality, though for the moment we have no evidence that it
will. Should new undetectable gene-​based technologies emerge then anti-​doping will
become obsolete. Since we have not yet reached that point, we shall explore the state
of the system in its current form.
(Dimeo & Møller, 2018, p. viii)

When this was written, we were of course aware that research in gene technology is progressing
and reckon that in a more or less distant future it will probably bless humankind with marvelous
cures for severe diseases. However, in the late 1990s we listened to scientific lectures in which
experiments on mice were introduced that showed the effect of inactivation of the myostatin
gene. It was fascinating to see on slides the huge difference in muscle growth between a normal
mouse and the so-​called Schwarzenegger mouse. The problem with blocking the myostatin
gene, we learned, was that muscle growth continued and that it was not yet possible to reacti-
vate the myostatin gene when the desired muscle size had been reached. This convinced us that
gene doping was of scientific interest for futurists, but that gene doping at present was still too
speculative to be relevant to consider in a book dealing with the current anti-​doping crisis. Our
conviction was based on two things. First, there is a big difference between learning to inject
oneself with anabolic steroids, erythropoietin, cortisone etc. and to perform genetic engineering
safely. Second, if scientists with knowledge about gene technology offered to help athletes with
the process we thought it unlikely that athletes would accept the offer so long as the procedure
was not tried and tested, the outcome uncertain, and the risk immense. But these may well have
been too conservative assumptions based on too much faith in human ethics and rationality.
The state-​run doping program in the former East Germany exposed in the wake of the
German reunion in 1989 proved that some regimes can be reckless enough to sacrifice
their own citizens in pursuit of international fame and glory in the sports arena (Franke &
Berendonk, 1997). Despite the collapse of the Iron Curtain and the former Eastern bloc coun-
tries’ closed regimes being replaced by democracies, the 2015 revelations of organized doping in
Russia, which resulted in Russia’s exclusion from the 2018 Winter Olympics in PyeongChang,
indicates that states’ willingness to exploit their athletes for propaganda purposes is not a thing
of the past and may include willingness to work towards the creation of genetically modified
super-​athletes.The difference between doping in the East German dictatorship and post-​Soviet
Russia is the extent to which the enhancement measures transformed the athletes’, particularly

455
456

V. Møller and R.B. Møller

the females’, bodies. In the former East Germany, children and adolescents were unknowingly
administered potent steroids such as Oral Turinabol in quantities that resulted in visible mascu-
linization. This practice was coined “coerced doping” (Zwangsdoping) by the German sports
historian Giselher Spitzer (1998).
The doping practice that was revealed in Russia was different. According to whistle-​blower
Yuliya Stepanova, the state assisted athletes’ doping but she was informed what she was injected
with. Hence, she could reveal that she was given testosterone and other anabolic steroids, plus
erythropoietin (Ash, 2016). She might at the time have felt it difficult to refuse the injections
because if she did she might be expelled from the national team. But in case she had experienced
negative side effects such as visible signs of masculinization she could have walked away, which is
probably the reason why we do not see female athletes as bulky and masculine as we did during
the Cold War. We considered that so long as athletes’ agency was not suppressed by state coer-
cion, athletes would only (accept to) use doping up to the point where their health would be at
significant risk. Accordingly, we reasoned that gene doping, the outcome of which is uncertain
and the results irreversible, will put athletes off regardless of potential gains. However, a 2017
article in the magazine New Scientist gave reason to reconsider our confidence in reason.

Gene hacking
New Scientist is a popular science magazine established to inform laypeople of advances in
science. Its articles sometimes appear simplistic and sensationalist and has on occasion been
criticized by scientists. Some have even called for a boycott of the magazine and encouraged
scientists to stop working and writing for it in order to prevent the public from being misled
by its catchy covers and headlines (Whyevolutionistrue, 2009). Having said that, New Scientist is
generally sufficiently accurate and true to the science it covers which is why esteemed scientists
like Richard Dawkins consider it worthwhile to voice their criticism when they find the maga-
zine guilty of misrepresentation of a topic. Thus, we hold that its 2017 article about biohacking
cannot immediately be discarded as science fiction even though it may sound as such and give
cause for concern.
The article introduces clustered regularly interspaced short palindromic repeats (CRISPR),
which is a new and cheap technique to make changes to DNA sequence. It is considered a
big leap forward in medical science’s attempt to find remedies for incurable diseases. But the
CRISPR technology is not confined to medical laboratories. The article’s main character is
Josiah Zayner, a biochemist and former NASA employee who was the first person to edit his
own DNA by the use of this new method to genetic engineering. Livestreamed on Facebook,
he explained the technique and showed how to perform it on himself. He simply took out a
vial that contained a CRISPR-​edited copy of his own DNA where the myostatin gene had
been removed and injected himself. His intention was to increase muscle growth. Another aim
was to promote a do-​it-​yourself kit he sells through his private company, The Odin, by which
others can replicate his work. The author of the article quotes other so-​called biohackers who
embrace this advance in science. One wants to use Zayner’s kit to remedy his color blindness,
another wants to use genetic engineering to inject an extra copy of the gene follistatin, which
increases muscle growth much like myostatin does. So, Schneider and Friedmann’s predic-
tion that whereas gene doping based on authorized sciences that meets the requirements for
ethical research in human subjects may be far away, to carry out uncontrolled experiments is
far less difficult. In fact, the first movers have already begun. Whether it still holds true after
the invention of CRISPR that genetic modification is a perilous practice, as Schneider and
Friedmann presume, is an open question. Robin Lovell-​Badge a leading CRISPR researcher

456
457

Gene doping: Ethical perspectives

at the Francis Crick Institute in London, UK, is quoted as saying: “Zayner’s experiment was
‘foolish’ and could have unintended consequences, including tissue damage, cell death, or an
immune response that attacks his own muscle” (Pearlman, 2017, p. 22). Bioethicist John Harris
from the University of Manchester, UK, holds a different view similar to Miah’s bioscience
optimism. According to Harris, biohackers could help to accelerate the safe use of CRISPR
in humans, pointing out that there is “a long and noble history of both doctors and scientists
experimenting on themselves [that] has proven tremendously valuable in the public interest”
(Pearlman, 2017, p. 23). Günes Taylor at the Francis Crick Institute in London who also work
with CRISPR takes the middle ground. “Part of me is, like, ‘that is so awesome’… but it won’t
work.” In her opinion it will be more difficult than the biohackers think. “CRISPR has been
sold as a cure all, but actually getting it to do the thing that you want it to do successfully is more
complicated” she says, echoing Schneider and Friedmann (Pearlman, 2017).
For better or worse, the possibility of having one’s body changed by genetic manipulation
has become a reality and because the kit made for it is not technically a drug, “the FDA will not
yet regulate at-​home genome editing.” As a result, self-​experimentation is not illegal in the US.
Zayner is living evidence that there are dedicated persons who without the guarantee of a posi-
tive outcome, are willing to jeopardize their health in the hope they can break natural limits to
muscle growth. He is also proof that there are unauthorized experts willing to make the neces-
sary tools available for unauthorized people who may not be aware of the risks involved. The
combination of unscrupulous persons with sufficient knowledge about gene manipulation and
people dedicated to a cause who are willing to take risk are all that is needed for gene doping to
evolve. Hence it is understandable that WADA has taken a proactive approach to the issue and
in October 2017 announced that from 2018:

The following with the potential to enhance sports performance are prohibited:

1. The use of polymers of nucleic acids or nucleic analogues.

2. The use of gene editing agents designed to alter genome sequences and/​or the tran-
scriptional or epigenetic regulation of gene expression.
3. The use of normal or genetically modified cells.
(WADA, 2017, p. 6)

This means in effect that all forms of gene therapy or gene doping are banned from inter-
national competitive sport. The immediate problem for WADA in this regard is that it is much
easier to introduce a ban than find effective measures to enforce it. However, in what follows we
will argue that a blanket ban on gene therapy will be ethically indefensible in case gene therapy
actually becomes so potent, safe, and successful as the most optimistic futurists foresee, and why
such a scenario will spell and end to the meaning of anti-​doping.

The controversial case of Lionel Messi

As a child, Lionel Messi, one of the world’s best ever football players, was diagnosed with
growth hormone deficiency (GHD). Fortunately for him, growth hormone (GH) therapy for
that condition had been developed. Human growth hormone (hGH) was first isolated from the
human pituitary gland in 1956. This hormone had to be extracted from humans, so availability
was limited and expensive. Only the most severe cases were offered treatment. In 1960, the
American National Pituitary Agency was established with the purpose to coordinate collection
and extraction of the hormone. From 1963 to 1985, the agency registered no more than 7700

457
458

V. Møller and R.B. Møller

children in the US who were treated with GH extracted from humans. The number of chil-
dren treated worldwide was 27,000 (Ayyar, 2011). In 1972, the biochemical structure of hGH
was identified. This paved the way for the invention of recombinant human growth hormone
(rhGH). The first time the gene was successfully cloned was in 1979 and the American com-
pany Genentech developed rhGH in 1981 by a biosynthetic process (Ayyar, 2011). Hence
treatment was no longer limited by availability of hGH. When diagnosed, Messi was only 4
feet 3 inches tall (127 centimetres). Doctors predicted his final height would be 4 feet 3 inches
(140 centimetres) (Cooney, 2016). Thanks to the advances in medicine, Messi’s GHD could be
treated. For 3 years from 1998 at the age of 11, he had injections of rhGH. However, treatment
was still expensive. It cost more than his family could afford. His luck was that he was such an
outstanding talent that FC Barcelona agreed to pay for his treatment. It proved to be a fabulous
investment. Messi now stands 5 feet 7 inches tall (170 centimetres) and he has been instrumental
to the success of the club since his debut in 2004, winning nine Spanish Championships and
four UEFA Champions Leagues among a host of other trophies (FC Barcelona, 2017). It is
unlikely he would have become an equally successful footballer without treatment for his GHD.
At first glance, the Messi case is uncontroversial from an anti-​doping perspective because the
World Anti-​Doping Code allows use of banned substances for medical purposes. If an athlete is
diagnosed with a condition that needs medical treatment the person in question can be granted
a Therapeutic Use Exemption (TUE). The justification for this is the claim that necessary med-
ical treatment is not performance enhancing but performance enabling. The TUE system does
not allow athletes a competitive advantage, only sufficient treatment without which it would be
impossible for the athlete to participate in sports. The review process is rigorous. Four criteria
must be met for a TUE to be granted including that the “therapeutic use of the substance would
not produce significant enhancement of performance” (WADA, 2016). If clean sport is the ideal,
ideally the TUE system should be repealed. However, according to WADA it is not possible to
preclude athletes who suffer from various illnesses and conditions because it “would undermine
a fundamental value of sport that is the right of access and participation to sport and play, which
has long been recognized by numerous international conventions” (WADA, 2016, p. 4).
The problem for WADA is of course that the criterion, that treatment would not produce
any significant enhancement of performance, is impossible to determine. The reason why there
is zero tolerance for the vast majority of substances on the banned list is that ever so small
amounts found in a test may be residues of larger dosages taken in order to gain performance-​
enhancing effect. That is, allowance of small quantities can be exploited for doping purposes.
Moreover, it is impossible to establish that a performance-​enhancing substance administered
for health reasons does not have a significant performance-​enhancing effect on the person
who is treated. Obviously, if an athlete cannot breathe properly because he or she suffers from
asthma and thus would perform significantly below the level needed to be competitive in elite
sport, asthma medication has a significant performance enhancing effect on this person. So
what WADA has in mind must be that if, hypothetically, the athlete was not an asthmatic he
would perform at a higher or exact same level as he does with asthma medication but not on
a slightly lower level. However, since the introduction of the Code and the TUE system, some
asthma medications have proven so effective that athletes who have been granted a TUE due
to confirmed asthma/​airway hyper-​responsiveness have outperformed their nonasthmatic rivals
(Fitch, 2016).
If it is true that asthma medication is so effective that treatment unintentionally provides
the athlete patients a slight advantage, gene therapy that could once and for all remedy the
condition seems a fairer solution should such therapy become available. But with the new
WADA policy banning all kinds of gene doping and therapy, this fairer solution is ruled out. For

458
459

Gene doping: Ethical perspectives

example, athletes with asthma who want to exercise their fundamental right to participate in
sport and play will paradoxically be forced to live with their disease and fight its symptoms by
an inhaler, which has the side effect that it boosts their performance capacity.
One of WADA’s main concerns in relation to gene doping and the TUE system is that it will
continue to benefit the athlete when the medication is no longer needed for medical treatment.
Hence they emphasize that “TUEs are granted for a specific method or a substance with a
defied dosage and route of administration. They are also granted for a specific period of time
and do expire” (WADA, 2016, p. 3).
If we now return to the case of Messi, what immediately appeared uncontroversial seems
at closer inspection less clear cut. The treatment he received as a child did not prevent him
from playing football. He was playing football and identified as a rare talent so if he had been
denied a TUE this would not have violated his rights of access and participation to sport and
play. Moreover, it is hard to argue that his treatment did not significantly improve his perform-
ance ability when considering there is not a single player in world elite football nearly as short
as 4 feet 7 inches. The shortest is Élton José Xavier Gomes who was signed by Romanian side
Steaua Bucuresti in 2007 who stands 5 feet 0.63 inches (154 centimetres) tall (FootyFan, 2013).
In addition to this, it does not seem that Messi’s condition met the first of the four criteria that
all must be met for granting a TUE namely that the “athlete would experience significant health
problems without taking the prohibited substance or method” (WADA, 2016). It is true that
adults who suffer from GHD can experience symptoms such as reduced bone density, insulin
resistance, and heart problems, but these symptoms do not appear in childhood so it is unlikely
that Messi had experienced significant health problems when he began treatment (Toft, 2014).
If in the future it will be possible to treat GHD by gene therapy rather than by injections
with rhGH, a “new Messi” could not benefit from that according to WADA’s abovementioned
announcement. On the other hand, it would not make much sense to grant an athlete a TUE
for use of rhGH to treat GHD in the old-​fashioned style if another more efficient treatment
became available. And the problem is bigger than that.
Height is recognized as a general advantage in life and positively related to socioeconomic
status. A British study that evaluated genetic data of 396 genetic variants associated with height
from 120,000 persons aged between 40 and 70 found that tallness on average led to higher
levels of education, higher job status, and better income (Tyrrell et al., 2016). Most parents
wish the best for their children. Height is determined by genes, so it is predictable that if one
day it becomes possible for parents to safely boost their children’s growth by genetic modifica-
tion such therapy will be in demand. This has already been the case with rhGH. In 2003, the
US Food and Drug Administration (FDA) approved treatment with rhGH for children with
idiopathic short stature. That is, kids who are short without suffering from GHD. However,
shortness is a relative concept. In an interview with ABC News, pediatric endocrinologist Paul
Desroiers explains that since the FDA’s approval, ambitious parents with healthy kids regularly
turn up in the hope that he will give GH to their child. Desroiers gives a striking example: “Just
last week, the father of a young baseball player –​a 14-​year-​old who was already 5 feet 6 inches
(167 centimetres) tall –​expected [me] to prescribe recombinant growth hormone to add height
to his budding athlete.” The father wanted to make his kid big in the vain hope that he thereby
could turn him into a baseball star. “He was willing to pay more than $45,000 a year and didn’t
even bat an eyelash” (James, 2009).
When parents successfully begin to make their children grow taller by gene therapy, other
parents may be afraid their children will fall behind. Consequently, demand for this kind of
therapy will likely grow to the extent it becomes mainstream as we have seen with other body
improvement services such as botox injections, breast implants, and facelifts etc. However, when

459
460

V. Møller and R.B. Møller

it comes to height, the decision must be made by the parents before their child is of legal age.
Most of these hypothetical children will become taller than their parents and may benefit from
the therapy by getting better-​paid jobs than their shorter parents had. An infinitesimal number
of these children will be gifted by a unique talent for sport, whether it be football or something
else. It does not seem just to exclude these children from exercising their right to participate
in sport because of the decision their parents made to prevent them from being disadvantaged
in life in general without the intention to increase their ability to succeed in sport. It will also
be hard to find a valid argument for exclusion of a person who is standing 5 feet 7 inches tall
thanks to gene therapy decided by his or her parents from competing in sports with people
who are of similar size.
If WADA will indeed stand firm on their ban on all forms of gene therapy as the 2018
Prohibited List announces, they would have to develop ways to measure all international athletes’
genetic makeup to exclude the ones who have reached a certain height, strength, or stamina
by genetic modification. One should be forgiven for thinking that subjecting all international
athletes to gene testing raises serious questions about privacy rights. However, based on the
development of anti-​doping so far, there is nothing to suggest that respect for athletes’ privacy
rights will represent an obstacle for the introduction of this additional measure.
There is a general consensus that the fight against doping is in the public interest. It is backed
almost universally by the international sports federations. Almost all nations have signed the
WADA Code. The commitment to anti-​doping is further established by the United Nations via
UNESCO’s International Convention against Doping in Sport and by the Council of Europa
via its Anti-​Doping Convention. Based on this consensus, the European Court of Human Rights
has hitherto accepted every obligation forced upon athletes including the requirement to pro-
vide urine and blood samples and whereabouts information 24/​7 all year round. Because, as the
Court reasons, any “reduction or removal of the relevant obligations on athletes would inevit-
ably lead to an increase in doping and would be at odds with the consensus on the need for
unannounced testing as part of doping control” (WADA, 2018). In light of this, if WADA argues
that gene doping poses a threat to the integrity of sport, the introduction of gene testing, even
if it necessitated athletes’ submission to regular biopsies, would in all likelihood be condoned
by legal authorities no matter how unethical this would appear to ordinary people. So, it is not
because of judicial objections that advances in gene technology will spell an end to anti-​doping,
nor is it because gene modification may be undetectable. Since the formation of WADA there
is no indication that a doping substance or method has not been banned because of the fact that
there was no detection method in place. Gene doping does not represent anything novel in this
regard (Franks, 2014). The core of the problem for WADA is that the ban on gene therapy is not
anti-​doping. It is anti-​progress. The concept of gene doping is misleading. If this is not imme-
diately clear from the Messi example, let us conclude this chapter with a thought experiment.

Health protection and longevity

According to the World Health Organization, ischemic heart disease and stroke are the world’s
most common causes of death accounting for 15 million deaths globally in 2015 (World Health
Organization, 2017). There is a hereditary, that is, genetic, component to heart and circulatory
diseases. So, imagine that it became possible to reduce the risk of developing heart diseases by
gene-​editing human embryos with the only side effect being that heart strength and circulation
capacity improves significantly. Parents with a family history of heart failure and early deaths
would be eligible for treatment. However, children whose parents had accepted the treatment

460
461

Gene doping: Ethical perspectives

would not be eligible for international sports participation under the remit of WADA-​compliant
international sports federations. In effect, parents who would choose the protection of their
children’s health long term would thereby preclude them from competing in sports at elite level
should they be gifted with the required sports talent.
If children’s life expectancy can be increased by embryo editing it is predictable that the
practice will follow the same trend as has been the case with in vitro fertilization. The demand
for this treatment will grow. As a consequence, more and more athletes from the affluent world
will be excluded from elite sports due to their genetic makeup. And this without having done
anything wrong, being negligent, or even consented to the health prevention therapy that has
simply been forced upon them. Under such circumstances, their exclusion will be ethically
indefensible because it is discriminatory and punishes individuals for a condition that in no way
is their own fault.
At first glance, a TUE-​like acceptance of athletes who have been genetically modified as
embryos looks like a solution that would not compromise the ban on gene therapy performed
after birth. But if athletes modified as embryos shall be accepted in sport because they were
not responsible for the editing of their genes, children whose parents have subjected them to
gene therapy while they were under legal age should also be given exemption regardless that
their parents subjected them to gene therapy due to concerns about their height, strength, or
stamina. Otherwise these unfortunate children, whose parents did not edit their embryos but
waited till their babies were born and proved to be at a disadvantage in comparison with other
children, would be discriminated against and punished for something they were not legally able
to consent to.
If WADA in this future scenario would resist sports participation of involuntarily genetic-
ally modified athletes based on the premise that these modified individuals would undermine
the integrity of sport, the long-​term consequence will be that elite sport becomes a museum.
A spectacle of the past. Those who will be allowed to compete will be the natural bodied, rela-
tively weak with a below-​average life expectancy. The men and women whose parents secured
them strong physiques will, if they bother at all, watch in pity the unfortunates who struggle
unimpressively in the sports arenas for glory and fame, before they go out to do their own sports
at a much higher level, recreationally.

References
Ash, L. (2016, December 30). Yuliya Stepanova: what do Russians think of doping whistleblower? www.
bbc.com/​news/​magazine-​38406627.
Ayyar,V.S. (2011). History of growth hormone therapy. Indian Journal of Endocrinology and Metabolism 15(3),
162–​165.
Cooney, G. (2016, April 4). After yesterday’s revelations, it’s worth asking again: would Messi be Messi
without HGH? www.balls.ie/​football/​after-​yesterdays-​revelations-​its-​worth-​asking-​again-​would-​
messi-​be-​messi-​without-​hgh-​329558.
Dimeo, P. & Møller,V. (2018). The Crisis of Anti-​Doping: Causes, Consequences, Solutions. London: Routledge.
FC Barcelona. (2017). Honours. www.fcbarcelona.com/​football/​card/​honours-​football.
Fitch, K. (2016). The World Anti-​Doping Code: can you have asthma and still be an elite athlete? Breathe
12(2), 148–​158.
FootyFan. (2013). Top 10 shortest football players in the world. www.sportskeeda.com/​football/​top-​10-​
shortest-​players-​in-​football/​4.
Franke,W.W. & Berendonk, B. (1997). Hormonal doping and androgenization of athletes: a secret program
of the German Democratic Republic government. Clinical Chemistry 43(7), 1262–​1279.
Franks,T. (2014, January 14). Gene doping: sport’s biggest battle? www.bbc.com/​news/​magazine-​25687002.

461
462

V. Møller and R.B. Møller

James, S.D. (2009, September 15). Growth hormones on rise in healthy kids. http://​abcnews.go.com/​
Health/​growth-​hormones-​healthy-​kids-​increase/​story?id=8571628.
Miah, A. (2004). Genetically Modified Athletes: Biomedical Ethics, Gene Doping and Sport. London: Routledge.
Munthe, C. (2005). Ethics of controlling genetic doping. In C. Tamburrini & T. Tännsjö (Eds.), Genetic
Technology and Sport: Ethical questions (pp. 107–​125). London: Routledge.
Pearlman, A. (2017). My body, my genes. New Scientist 236, 22–​23.
Schneider, A. & Friedmann, T. (2006). Gene Doping in Sport: The Science and Ethics of Genetically Modified
Athletes. San Diego, CA: Elsevier.
Spitzer, G. (1998). Doping in der DDR: Ein historischer Überblick zu einer konspirativen Praxis: Genese –​
Verantwortung –​ Gefahren. Köln: Sport und Buch Strauss.
Toft, D.J. (2014, May 12). Growth hormone deficiency symptoms. www.endocrineweb.com/​conditions/​
growth-​disorders/​growth-​hormone-​deficiency-​symptoms.
Tyrrell, J., Jones, S.E., Beaumont, R., Astley, C.M., Lovell, R., Yaghootkar, H., et al. (2016). Height, body
mass index, and socioeconomic status: mendelian randomisation study in UK Biobank. BMJ 352.
World Anti-​Doping Agency. (2016, September 27). WADA releases frequently asked questions (FAQs)
on Therapeutic Use Exemptions (TUEs). www.wada-​ama.org/​en/​media/​news/​2016-​09/​
wada-​releases-​frequently-​asked-​questions-​faqs-​on-​therapeutic-​use-​exemptions-​tues.
World Anti-​ Doping Agency. (2017, September 29). The World Anti-​ Doping Code International
Standard Prohibited List 2018. www.wada- ​ a ma.org/ ​ e n/ ​ m edia/ ​ n ews/ ​ 2 017- ​ 0 9/​
wada-​publishes-​2018-​list-​of-​prohibited-​substances-​and-​methods.
World Anti-​ Doping Agency. (2018, January 18). WADA welcomes ECHR decision to back
Whereabouts rules. Retr ieved from www.wada-​ a ma.org/​ e n/​ m edia/​ n ews/​ 2 018-​ 0 1/​
wada-​welcomes-​echr-​decision-​to-​back-​whereabouts-​rules.
World Health Organization. (2017, January 2). The top 10 causes of death. www.who.int/​mediacentre/​
factsheets/​fs310/​en/​.
Why Evolution is True. (2009, March 21). The New Scientist has no shame –​again. whyevolutionistrue.
wordpress.com/​2009/​03/​21/​the-​new-​scientist-​has-​no-​shame-​again/​.

462
463

33
ENHANCING EVOLUTION:
THE TRANSHUMAN CASE
FOR GENE DOPING
Andy Miah

Introduction
The conventional view on gene doping is that it should be treated like other forms of doping,
both in scientific and ethical terms, as evidenced by its inclusion within the World Anti-​Doping
Code back in 2003 (World Anti-​Doping Agency (WADA), 2003). The basis for this position
is that, like other forms of doping, gene doping involves the use of medical technologies for
nontherapeutic biological modifications, as described in WADA’s initial terminology banning
gene doping, which described it as:

M3: Gene Doping

The non-​therapeutic use of cells, genes, genetic elements, or the modulation of gene
expression, having the capacity to enhance athlete performance.
(WADA, 2003)

Over the years, this terminology has been extended and clarified towards the following defin-
ition of gene doping, which removes the term “non-​therapeutic,” due to the term “enhance”
providing adequate determination of unacceptable use:

M3: Gene Doping

The following, with the potential to enhance sport performance, are prohibited:
1. The use of polymers of nucleic acids or nucleic acid analogues.
2. The use of gene editing agents designed to alter genome sequences and/​or the tran-
scriptional or epigenetic regulation of gene expression.
3. The use of normal or genetically modified cells.
(WADA, 2018, p. 6)

In this respect, gene doping violates crucial characteristics of scientific and medical ethics, which
at best, will permit the use of experimental techniques in service of a therapeutic need, but which
would not be allowed merely to enhance an athlete’s performance. The argument underpinning
this stance is that medical interventions are inherently risky and taking such risks can only be

463
464

A. Miah

justified when there are no other means of treating a person who suffers from some illness, injury,
or disease. In this sense, medical treatment is a practice of necessity rather than a practice of prefer-
ence. It is the only means by which we have the potential to alleviate biological suffering.
This principal concern about medicine being misused is articulated at the beginning of the
World Anti-​Doping Agency’s Prohibited List, which states:

Any pharmacological substance which is not addressed by any of the subsequent

sections of the List and with no current approval by any governmental regulatory
health authority for human therapeutic use (e.g. drugs under pre-​clinical or clinical
development or discontinued, designer drugs, substances approved only for veterinary
use) is prohibited at all times.
(WADA, 2011)

This stipulation means that, for the sports world to endorse gene doping, it would need to
advocate a position that contradicts worldwide codes of conduct in scientific research and
medical care. This is especially challenging when the sports industries rely heavily on the
endorsements of medical professionals to ensure that their events remain sufficiently safe. In
many sports, athletes must pass a medical, physical examination to demonstrate that they are
fit for competition, before they are allowed to compete. Indeed, in sports where this does
not happen, it is due to lack of resources, but an ideal system is one where the health risks
an athlete encounters can be monitored continuously and where medical interventions can
occur as soon as possible, once a risk becomes apparent. These characteristics of elite sport’s
medicalized infrastructure are reinforced when considering that those working within
sports governance are often people involved with the medical professions. For example,
the former International Olympic Committee (IOC) President, Dr. Jacques Rogge and
many celebrated athletes who have gone on to be involved in the administration of sport
have medical or scientific expertise. The history of anti-​doping may even be described as a
movement progressed by key individuals within the medical professions.
In this respect, sports –​especially elite sports competitions –​should be understood socio-
logically and culturally as industries that are regulated, not just by medical concerns pertinent
to the practice of sport, but industries that are engaged with wider healthcare priorities. Indeed,
the pursuit of IOC member and long-​term anti-​doping advocate Arne Ljungqvist to ensure
all Olympic bidding nations have a criminal law against doping speaks to the manner in which
many doping substances derive from an illegal black trade and where such use coheres with
wider concerns about drug abuse within society at large.
As such, advocating doping –​or promoting gene doping –​is incompatible with the duty
of care taken by medical professionals towards athletes, a commitment which is central to the
execution of a sports event. These circumstances are made all the more complicated because
elite adult sport is intimately connected with the development of sporting experiences in
childhood. While there are reasonable distinctions to be made between these two stages of
life, their continuity through sport makes it all the more challenging to hold a position on
gene doping –​or doping generally –​that is anything other than supportive of prohibition or
deterrence.
This position is made manifest in the deliberations of sport authorities explicitly via the
WADA Code, which identifies three areas of interest when determining whether or not a sub-
stance, method, or technology should be considered a doping practice. Thus, a new candidate
for inclusion within the list must engage two of three conditions to be worthy of consideration:

464
465

Enhancing evolution

• 4.3.1 A substance or method shall be considered for inclusion on the Prohibited List
if WADA, in its sole discretion, determines that the substance or method meets
any two of the following three criteria:
• 4.3.1.1 Medical or other scientific evidence, pharmacological effect or experi-
ence that the substance or method, alone or in combination with other
substances or methods, has the potential to enhance or enhances sport
performance;
• 4.3.1.2 Medical or other scientific evidence, pharmacological effect or experi-
ence that the Use of the substance or method represents an actual or
potential health risk to the Athlete;
• 4.3.1.3 WADA’s determination that the Use of the substance or method violates
the spirit of sport described in the introduction to the Code.
(WADA Code, 2015)

When confronted with a new means of altering an athlete’s biological circumstances, WADA
tests against these parameters to determine its degree of concern. Together, the three criteria
provide scope for a range of performance enhancements to be acceptable, but are sufficiently
broad to address a range of examples, from hypobaric chambers to the use of gene doping or
steroid use.
Aside from the third criteria, the emphasis on medical and scientific evidence reinforces
the points outlined above, which explain how the problem with the sports world endorsing
any form of performance enhancement that involves medical substances or method has to do
with the ethical commitments made by –​and thus, legal restrictions upon –​health professions
allied to the practice of sports. In this respect, it is a general conclusion that sports are incom-
patible with the pursuit of human enhancement, where this involves the use of scientific and
medical techniques, substances, or methods, which aim to provide a performance effect that
exceeds therapy. Indeed, even therapeutic use of such things requires special approval, if an ath-
lete expects to use them while competing.
However, there is an alternative view, which provides scope for sports to embrace gene
doping, while maintaining a hard line on other forms of doping, and this article outlines the
basis of that position. The argument requires sports –​and society more widely –​to embrace a
nontherapeutic model of health, whereby the role of medicine is not just to treat patients towards
the achievement and maintenance of a biostatistical norm of health, but where it embraces
a definition of health that is epitomized by the transhumanist conviction that our biology
must evolve alongside technology, not independent of it. By implication, this commitment also
involves foregoing the idea that healthcare should be focused just on repair, arguing instead that
a healthier population would emerge from circumstances where human enhancement is cen-
tral to health maintenance. This aspect of enhancing evolution also allows us to endorse gene
doping, without needing to embrace all other forms of doping. In this respect, gene doping is
valued partly for its capacity to take evolution to a new level of functioning. We can call this
the germ line proposition and, while somatic cell line modifications would not have any evo-
lutionary impact, there are also reasons to endorse these kinds of enhancements, particularly
where research can lead to safer forms of performance enhancement than current methods
which we call doping.
The starting point for this position derives from what may be described as the western model
of healthcare, although I offer some caution when using such terminology. In this case, I use
the term western model to describe a system of approaching medicine and healthcare as leading

465
466

A. Miah

inexorably to the pursuit of immortality and indefinite health span, where aging is, increasingly,
treated as a disease to be resisted, rather than an inevitable consequence of life (Gladyshev &
Gladyshev, 2016; Janaca et al., 2017).While such notions may seem to be located in the realm of
science fiction, there are credible, tangible examples of scientific research and healthcare policy,
which expose these aspirations (Bulterjs et al., 2016).
In so doing, this paper questions the scientific and ethical basis for anti-​doping work; arguing
that the pursuit of human enhancement –​through sport or other means –​ought not be
described as medicine at all, but as some new category of human intervention, which consists of
our intervening with or disrupting the course of evolution. At the very least, the implications of
these circumstances should lead WADA and sports organizations to rearticulate their approach
to gene doping specifically and human enhancement generally. Such repositioning would
involve reorienting itself around the changing circumstances within our society and embracing
the wider commitment that this implies to support the principle that humanity embraces the
pursuit of perfecting technology, in order to optimize our capacity to evolve.
In making this case, I recognize that many anti-​doping advocates will not take these terms as
being the starting point for a debate about the ethics of doping. Instead, discussions often focus
on what kinds of practices sports are, from which we may then ascertain whether enhancements
are consistent with such criteria. Consequently, before returning to the medical model as the
basis for my reasoning, I will outline why it is that I think that the ethics of sport do not matter
to resolving the doping dilemma.

Why appeals to the “spirit of sport” don’t resolve the doping dilemma
The World Anti-​Doping Code outlines that its mandate is interwoven within a range of critical
concerns that seek to protect the “intrinsic value of sports” (WADA, 2015), which it collectively
describes as the “spirit of sport.” It states:

Anti-​doping programs seek to preserve what is intrinsically valuable about sport. This
intrinsic value is often referred to as “the spirit of sport.” It is the essence of Olympism,
the pursuit of human excellence through the dedicated perfection of each person’s
natural talents. It is how we play true. The spirit of sport is the celebration of the
human spirit, body and mind, and is reflected in values we find in and through sport,
including:
• Ethics, fair play and honesty
• Health
• Excellence in performance
• Character and education
• Fun and joy
• Teamwork
• Dedication and commitment
• Respect for rules and laws
• Respect for self and other Participants
• Courage
• Community and solidarity
Doping is fundamentally contrary to the spirit of sport.
(WADA, 2015)

466
467

Enhancing evolution

While each of these elements resonates with what one may conventionally think of as the
values of sports, taken individually, they do not make a compelling case against doping, nor are
uniquely characteristic of sport at all. For instance, one could envisage a system of pro-​doping
which preserves fair play –​even enhances it.Thus, the sports world may just allow all athletes to
use whatever enhancements they wish (or even just the ones currently prohibited by WADA). If
every athlete operates by the same rule that stipulates they are free to use any such means, then
all will be competing by the same rules and fairness will prevail.While there may still be athletes
who seek to find other means as yet unimagined by the scope of the WADA Prohibited List, this
would not be dramatically different from the present circumstances. Indeed, it may be far better
as there may be far fewer alternatives by which an athlete could try to cheat.
Similarly, “excellence in performance” need not be jeopardized by doping, as, for instance,
the use of blood doping, would still require an athlete to train their body and minds to a com-
parable degree as an athlete who does not use blood doping. The gain in their performance
would not lead to a situation where the pursuit of excellence is diminished, especially if one
characterizes this pursuit as the perfection of athletic skills, where this is evidenced by being the
most proficient athlete. While this may be too narrow a characterization of sporting excellence,
Berg (2015) notes that the definition of excellence is malleable and changes depending on the
social context. While Berg also goes on to note that there may be a deeper characterization of
excellence which has to do with integrity, this does not help us terribly in resolving the doping
dilemma as, if doping were made legal, there would be no loss of integrity in using such means.
One could also easily imagine athletes working as a “team” while also doping. Indeed, one
might even find a reasonable articulation of how being able to dope effectively involves being
part of a highly sophisticated team of pro-​doping scientists, with whom they would have to
work carefully to achieve the optimal form of enhancement. Such endeavours may even involve
significant sacrifice or “courage” and “character” to achieve these goals.
In sum, any one of these characteristics may be supported in a climate where doping is
allowed, but the most challenging one to uphold is “health.” I am skeptical that health pro-
motion can be argued as intrinsic to the value of sport –​at an elite level so many players
are injured either through incident or over-​training/​competing –​and even the IOC Medical
Commission recognized that there is greater prevalence of some health risks resulting from elite
sports participation, compared to regular citizenship (IOC, 2009). Nevertheless, the incom-
patibility of health promotion with nontherapeutic interventions makes pro-​doping untenable,
which is why my focus in presenting an alternative response to gene doping relies heavily on
this dimension of the problem. In short, determining that doping would be acceptable requires
approaching health from a very different perspective than is conventionally held, but this is a
perspective which I will argue to be intrinsic to the logic of medicine, though we often do not
acknowledge this. Essentially, this is to say that the end goal of healthcare is the eradication of
the need for medicine or the medical services that we utilize across our lives. When healthcare
reaches this endpoint, then it will cease to be necessary. At most, it will become something else
other than direct care.

The purpose of healthcare is changing

In 2001, the Human Genome Project published the first draft of its findings in Nature, show-
casing 90% of the genome map. It was quickly accompanied by many discussions about what
may become of our world if we are able to intervene genetically in either the eradication of
disease and illness or through our ability to pick, choose, or enhance our genetic characteristics.
At the time –​and still –​the excitement was broadly focused on the gains that could be made

467
468

A. Miah

towards diagnostic discoveries or therapeutic interventions, though even here there have been
controversies over the eradication of illness or disease. For example, disability scholars have
argued against “ableism” (Wolbring, 2008), which is the unjustified presumption that disability
is something that needs fixing –​through genetic means or otherwise –​as opposed to some-
thing that societies should seek to accommodate. Even stronger concerns are found in how the
eradication of biological dysfunction could lead to a eugenic ideology becoming reinforced
within society, whereby unintelligent and socially damaging views may become commonplace,
such as determining a person’s worth on the basis of their genome. Some have even argued that
removing genetic dysfunctions by tinkering with our genes could have a negative evolutionary
impact, just because we do not yet fully comprehend our genes (Kozubek, 2017).
In any case, since the publication of the Human Genome Project, full genome sequencing
has changed dramatically in terms of cost and the methods by which altering the genome
would take place. Whereas the first human genome map cost around $2.5 billion and 20 years
to produce, it now costs less than $1000 and takes just a few weeks to process (Ray, 2016).
A future where every citizen has their own genome mapped continually across their lifetime
to minimize health risks and costs seems a highly likely scenario. While the number of thera-
peutic interventions has been limited to date, scientists are optimistic that more successes will
follow. However, these limitations in present application are not really the crucial detail in this
discussion. Rather, the more relevant aspect of this story to our present discussion about what to
do with gene doping in sport has to do with the purpose of science, medicine, and healthcare,
namely, the complete alleviation of biological suffering.
Whether it is mental or physical, the logic of the western model of health is to drive
humanity towards circumstances where, across our entire lives, we actually have no need for
medicine whatsoever. The end goal of healthcare is for us to get to a point where there is no
need to see a doctor, nurse, or to enter a hospital. No need for medication, for treatment, or
rehabilitation; where we are healthy and where this good health persists over our entire lives.
These characteristics describe the ideal healthcare system, but in a world where this is achieved,
the parameters by which we judge the worth of scientific experimentation are also different. It
will cease to be necessary to focus just on repair or maintenance, leaving open the possibility
that biology may be modified to transcend such limits and norms.
Of course, for anybody working within science or medicine, these circumstances are a long
way off. For many years to come, healthcare will be far more focused on health maintenance
than human enhancement. Yet, it is crucial that the goals of medicine are clear before we can
then consider how best to achieve them. As such, I argue that the end point of healthcare is for
nobody to need healthcare at all.This claim is central to my argument, as it frames how one later
approaches what society wants from biology and what is sought from its modification through
technology, which is an intermediary step towards embracing gene doping in sport. However,
before imagining a range of radical new transhumanist interventions, it is helpful to consider
something more familiar, such as vaccinations. Such medical interventions are consistent with
the ideas outlined above. Thus, a vaccination should be regarded in a similar manner as a way of
supplementing inadequacies within our genomes to cope with exposure to viruses that would
be catastrophic to our survival. People vaccinate themselves against such vulnerabilities to pre-
vent the possibility of becoming ill, even if the chances of this occurring are remote. In this
sense, healthcare works in a pre-​emptive way responding to the risks within our environment.
For sport –​and life more generally –​such risks require a more radical enhancement of our
biology to ensure such aims are met.
The specifics of these changes are characterized neatly in Buchanan et al. (2000), who
describe how science is moving humanity from circumstances where biological characteristics

468
469

Enhancing evolution

are defined mostly by chance to circumstances where they are defined by choice. Choices to
enhance humanity are made all the more critical because people make prior decisions to put
themselves in positions of high biological risk, whether that is through the pursuit of sport
or any number of other lifestyle choices, which are made in order to live the kind of life
that people deem to make life worth living. Whether our activities involve deciding to go to
space or going deep underwater, the extremities of these experiences require people to seek
enhancements in order to be fit for purpose. Equally, humanity’s desire to evolve or transcend
its biological limits –​typified by the pursuit of elite sports –​provides a foundation for pursuing
radical biological interventions, such as gene doping. Admittedly, I am mindful of the fact that
this argument may lead a critic to ask why this position does not simply open the door to all
forms of doping. My response to this has to do with the characteristics of the technology and
how it compares to all other doping methods. Gene doping’s capacity to modify us in a more
precise manner, but also because of its impact on our evolutionary trajectory, justifies it being
treated differently ethically.
Humanity’s interest in pursuing such alterations is evidenced by the many ways in which
people pursue aesthetic or functional enhancements throughout their lives –​choices made
just because they consider that such changes will permit the enjoyment of the kind of life that
they seek to live. The specifics of those choices may be worthy of vigorous debate, but then so
too might one debate the merit of dedicating a life to elite sport, which some would argue to
be mostly in the service of a capitalist economy which exploits the many (those athletes who
struggle to make ends meet) in order to benefit the few (those athletes and institutions who are
winners economically).
There may be further reasons to question such choices, especially where they may lead to a
range of health risks. For instance, one would want to ensure that such decisions are not made
under unreasonable duress or in the context of an excessively coercive environment (recog-
nizing that many choices occur within climates where such behaviors are reasonably coercive,
such as the coercive environment of sports training which requires athletes to train full time
in order to be competitive). To feel reassured about such choices, one must acknowledge that
they are not made in isolation, but that our choices are always shaped by a range of societal
influences. While one should always interrogate the value of such influence, our determin-
ation of values always operates within such conditions. There is no sense of self or interests
outside of our social circumstances, no choices that operate outside of a social context. There
is no inherent value to a new mobile phone, car, house, or even having a child, which can be
explained in and of itself, without recourse to some wider framework of values which is located
within a social network. In this sense, questioning the legitimacy of human enhancement, or
the value of leading a transhumanist existence, on the basis of concerns about being pressured
to make limiting decisions has only limited force.

What transhumanism means for gene doping

The arguments outlined above explain how humanity’s pursuit of a long and healthy life is
interwoven with the fabric of modern society and how this commitment drives us towards the
transhumanist aspiration to continually remake the species through means arising from intellec-
tual discoveries. Indeed, it recognizes that the proliferation of the human species is explicable
in large part due to technological innovation. Humanity’s evolutionary course has been shaped
by its utilization of technology. Whether it is the discovery of penicillin, an understanding of
hygiene, the development of artificial intelligence, or indeed, bikes, bakelite (the first synthetic
plastic), and bulbs (Bijker 1995), humanity has sought such means to redefine its existence

469
470

A. Miah

through technological advancement. On these terms, technology is defined in a very broad

sense to describe the application of humanity’s ideas to the natural world in a way which
transforms its characteristics. The breeding of animals, agriculture, tool development, science,
industry, medicine, and cartography, among many other examples, are all manifestations of this
journey.
Indeed, sports themselves are manifestations of this transhumanist trajectory. The present
value system that operates around elite sports elevates the importance of competition, winning,
and transcendence through the recording of results, world records, personal bests, and using
science to remove inhibitors to performance. These dimensions of modern sport are not simply
unintended consequences of the intrinsic values of sports, but have become sport’s intrinsic
values. Loland (2002, 2003) describes this view of sport’s value as the thin theory of sport,
which elevates the quantitative evaluation of performance. However, Loland argues that this is
an impoverished characterization of sport’s value, arguing instead on behalf of a thick theory,
which champions the development of a wider range of sporting virtues, such as effort or
personal development.Yet, Loland overlooks the fact that the pursuit of thinner priorities does
not jeopardize the enjoyment of his thicker values. The thick theory is not incompatible with
the thin theory. Indeed, it is apparent in the exhibition of contemporary elite sports that athletes
are able to apply both sets of values simultaneously in their conduct of competition.

What should sports do about gene doping right now?

Based on the current state of policy discourse in elite sports, it is highly unlikely that, any time
soon, the world of sport will embrace the transhumanist aspiration to use technology to con-
tinually expand the potential of our biology, and so bring about an age of precautionary pro-​
doping. Even preparing for gene doping remains somewhat contested, with the WADA Health,
Medical and Research Committee working group in Gene and Cell Doping determining in
August 2017 that “gene doping did not appear to represent an immediate threat” (WADA,
2017).This view is hard to understand given the volume of research, commentaries, and discus-
sion that have taken place about the imminent prospects of gene doping in sport, which have
surrounded each Olympic Games at least since Athens 2004. Nevertheless, the present limited
success of novel therapeutic genetic interventions means that it is not likely any time in the next
10 years to create a situation where either therapeutic interventions translate into enhanced
humans, or where specific selective or editing choices could permit a new kind of superhuman
athlete to emerge.
Thus, many of the ideas expressed in this chapter will seem somewhat removed from the
present-​day complexity of dealing with doping. After all, the present-​day reality that faces anti-​
doping has to do with maintaining a level playing field when competitors try to cheat by means
of doping. This chapter does not propose a solution to that problem. Indeed, I rather think that
this is an inherent feature of competitive, rule-​bound environments and it makes no sense to
complain about such characteristics, or to expect them to be absent within some imagined,
ideal scenario. For as long as we permit free will, we are likely to be stuck with behavior that
deviates from rule keeping.
Yet, the world of sport does need to get its thinking right on the future of genetic
interventions and their impact on the state of humanity. Part of this process involves moving
on from theoretical philosophy and precautionary policymaking –​namely, the prohibition of
things that are not yet here –​towards scenario planning and differentiating between forms of
gene doping challenges. Many of the discussions so far have been theoretical. They involve the
world of sport –​or philosophers –​trying to establish sound principles by which sport can be

470
471

Enhancing evolution

governed to protect its so-​called intrinsic value (WADA, 2015).Yet, this theoretical work must
be accompanied by some realistic expectations about how genetic technology may find its way
into society. For instance, there are at least four kinds of gene doping, which we are likely to see
impact upon the world of sport:

1) The conventional doping athlete.

2) The genetically identified athlete.
3) The genetically pre-​selected athlete.
4) The genetically enhanced descendent athlete.

Getting sport policy right means getting it right for society at large, and not just for the short
term. For years, various theorists have been developing legal definitions that take into account
the imminent biological complexity of future humans. For example, Miah (2003) considers the
rights of the genetically modified athlete, arguing that sport policy on the use of gene doping
would need to be consistent with wider societal entitlements to genetic modification, even
if the consequence of such changes is enhancement. Alternatively, UNESCO even adopted
a Universal Declaration on the Human Genome and Human Rights (1997) to prepare for
an era where genetically modified persons may find their rights curtailed were it not for an
expanded definition of what is recognized as genetically human. To this end, sports must plan
for these future humans, to ensure it creates a climate of inclusion for people who are genet-
ically enhanced and to ensure humanity can take advantage of technology in order to preserve
and enrich the species and maximize its chances of survival.
Of course there is an alternative to the value system I describe above. Indeed, it is likely that
many people would consider humanity’s pursuit of transhumanism to be misguided where,
instead, our focus should be simply on enabling people to live a reasonably healthy life, where
one accepts the natural consequences of our biology’s fallibility, vulnerability, and eventual
degradation. Parens (1995) redescribes Martha Nussbaum’s characterization of the “fragility
of goodness” (1986) when talking about the “goodness of fragility” as a basis for determining
which kinds of things we should seek to modify versus the kinds of things we should seek to
leave and accept as being part of the natural circumstances of our existence. Indeed, Parens goes
on to suggest that life may be more impoverished by this loss of fragility.
While I accept that it is prudent to ready oneself for a life that will involve some degree
of biological suffering brought about by our imperfect genomes, or by the inevitability of our
decline and eventual death, we should not mistake our need to prepare psychologically for such
circumstances for a willful desire for them to exist. Understood on this level, one must look to
a broader sense of humanity’s aspirations beyond simply the needs of individuals to find a way
to get through life, in order to derive a more comprehensive and prudential view of what things
matter to us. In this manner, one can draw on the evidence of humanity’s pursuit of eradicating
disease, or of trying to pursue a longer healthier life through scientific discovery and medical
intervention, as the basis for determining our wider interests in ideas that we currently regard
to be unrealistic, improbable, or utopian.
A further caveat to this conversation is to note that the circumstances I have described
would be mistakenly characterized as utopian. The shift from chance to choice brings with
it a number of new burdens, not least of which is the fact that people will have many more
decisions to make about their existence within these circumstances. For example, people may
need to consider which types of characteristics they design into the genomes of their progeny
and, equally, which characteristics they seek to remove. These are both highly burdensome and
highly controversial choices to make and it may be preferable to leave things to chance, rather

471
472

A. Miah

than be confronted with the division that will likely arise from having the capacity to make
such choices. Yet, I am optimistic that, with the fullness of time, humanity will find a way to
make sensible and effective decisions in this regard and further consider that this predicament is
precisely what distinguishes humanity as a species.
Of course, among any population there will be those who make choices that are ill conceived
or based upon some dubious logic, which will surely bring about the exact opposite conditions
they seek to promote. This is especially worrisome in cases where people are modifying genes
in such a way that might affect the human gene pool. Indeed, at this point in our history, it
is apparent that we have no good system for thinking through the basis on which we should
make such decisions, or even where such decisions should be made as individuals or where they
should be determined by a wider governing authority, which restricts elements of that freedom.
There are some examples already where this has been considered. For instance, various coun-
tries permit genetic sex selection so as not to pass on a debilitating genetic disease, but not all
countries are yet comfortable with permitting such choices for family balancing, or because
prospective parents might prefer to have a boy or a girl. Indeed, one of the reasons for not
allowing this is that there is some understanding how this may create a climate in which people
are allowed to impose their own gendered prejudices onto future generations in a way that is
challenging both in terms of population balance, but also because it reinforces unintelligent
stereotypes about the worth of either sex.
On this matter alone there are already words of caution for how we may think about some-
thing like gene doping. Prospective parents who might seek to modify or select an embryo on the
basis of its being able to pursue some kind of athletic ability, would likely misconstrue the potential
of our genes to determine our abilities. Such freedom would also –​if indeed it were possible –​
promote a particular range of abilities, which is so narrow in function as to undermine the agency
of the life that is created through such choices. One could quite reasonably envisage that a person
born from such choices and subject to their limitations would feel that their chances to flourish in
life had been hampered due to their having been modified for a specific purpose in mind, rather
than for a broader sense of what might be a life where a range of choices are available to them.This
critique of gene doping would be particularly important in cases where one choice may mean
that another choice is frustrated.While a child could hardly blame their parents for being very tall
when the parents themselves are very tall, there may be more foundation for complaint when their
phenotype has been chosen, regardless of those parental characteristics.
In pursuing these elements of the conversation, I am especially mindful of how far removed
this will seem from the present-​day realities of anti-​doping authorities whose challenges focus
mostly on trying to determine whether there are effective tests by which the sports industries
could even discover evidence of gene doping. Instead of attending to this immediate challenge,
my discourse has focused on the wider questions that surround the pursuit of testing at all and
the questions that enhancement more generally create. These parallel conversations are brought
together when thinking about the underpinning rationale and interests of anti-​doping.
What I want from anti-​doping authorities is for their codes to recognize that the rules are
incompatible with the wider pursuit of health that we see operating within society at large
today, which encompasses investment into finding safer forms of human enhancement. The
rules also must make provision for addressing the prospect of there being people within our
society who have been born with genetic modification within their ancestry or having been
born themselves as a result of genetic modification or selection. The fundamental values of
anti-​doping should expand to recognize and acknowledge that there would be no discrimin-
ation towards such people, so as to ensure that sports remain open to how such technologies
may become inextricable from what we think of as the means by which good health is possible.

472
473

Enhancing evolution

These would be quite minor modifications to the World Anti-​Doping Code, but would allow
it to maintain its pursuit of anti-​doping as it presently is, while also ensuring that it recognizes
its mission to be subject to wider trends in human society, where the pursuit of health beyond
our species’ typical functioning has become a valued pursuit of humanity.
Alongside this is the further commitment to seek safer forms of performance enhancement
within sport and to ensure athletes have access to and information about such means. In this
sense, there are also implications for how anti-​doping authorities approach education which is,
presently, entirely predicated on the idea that athletes must be educated in order to ensure they
do not dope.Yet, instead of framing the value of an education program as merely a means towards
fostering “anti-​doping behaviors,” as the WADA Education program describes its purpose, it
should seek to nurture a critical awareness of the complexities of performance enhancement and
their wider context, as described within this chapter. It should seek to engage athletes on matters
wider than sport, not least because many athletes who go through the elite sports system end
up in some form within health promotion industries. Pedagogically, education should never be
about indoctrination; it should never begin with a narrowly prescribed outcome as its conse-
quence. Rather, education programs within anti-​doping should seek to foster critical thinking,
not just about performance enhancement, but also its governance. This is of urgent need, espe-
cially given the controversies that surround the administration of sports that we have seen in
recent years.This more critical approach to anti-​doping education would not just work towards a
great achievement for the present anti-​doping system and its priorities, it would also be of much
greater service to humanity more widely.

References
Berg, A. (2015). The ethos of excellence. Journal of the Philosophy of Sport, 42, 233–​249.
Bijker, W.E. (1995). Of Bicycles, Bakelites, and Bulbs: Toward a Theory of Sociotechnical Change. London:
MIT Press.
Buchanan, A., Brock, D.W., Daniels, N., & Wikler, D. (2000). From Chance to Choice: Genetics & Justice.
Cambridge: Cambridge University Press.
Bulterijs, S., Hull, R.S., Björk, V.C.E., and Roy, A.G. (2015). It is time to classify biological aging as a
disease, Frontiers in Genetics, 6, 205.
Gladyshev, T.V. & Gladyshev, V.N. (2016). A disease or not a disease? Aging as a pathology. Trends Mol
Med, 22, 995–​996.
International Olympic Committee (2009). The International Olympic Committee (IOC) Consensus
Statement on Periodic Health Evaluation of Elite Athletes. https://​stillmed.olympic.org/​media/​
Document%20Library/​ O lympicOrg/​ N ews/​ 2 0090716-​ T he-​ I OC-​ C onsensus-​ S tatement-​ o n-​
Periodic-​Health-​Evaluation-​of-​Elite-​Athletes/​EN-​Health-​Evaluation-​of-​Elite-​Athletes-​2009-​
report-​1448.pdf#_​ga=2.261766285.1075817988.1529839491-​1215944620.1528100467.
Janaca, S., Clarke, B., and Gems, D. (2017). Aging: natural or disease? A view from medical textbooks, in
Vaiserman, A.M. (Ed) Anti-​aging Drugs: From Basic Research to Clinical Practice., pp. 11–​34. London: The
Royal Society of Chemistry.
Kozubek, J. (2017, January 9). How gene editing could ruin human evolution. Time. http://​time.com/​
4626571/​crispr-​gene-​modification-​evolution/​.
Loland, S. (2002). Fair Play in Sport: A Moral Norm System. London: Routledge.
Loland, S. (2003). Technology in sport: three ideal-​typical views and their implications. Idrotts-​Forum.
www.idrottsforum.org/​articles/​loland/​loland_​2.html [Accessed June 23, 2018]
Miah, A. (2003). Genetic modification (GM) in sport: legal implications. Contemporary Issues in Law, 6,
207–​226.
Nussbaum, M. (1986). The Fragility of Goodness. New York: Cambridge University Press.
Parens, E. (1995).The goodness of fragility: on the prospect of genetic technologies aimed at the enhance-
ment of human capacities. Kennedy Institute of Ethics Journal, 5, 141–​153.

473
474

A. Miah

Ray, T. (2016). With $999 whole-​genome sequencing service, Veritas embarks on goal to democratize
DNA information. Genome Web. www.genomeweb.com/​sequencing-​technology/​999-​whole-​
genome-​sequencing-​service-​veritas-​embarks-​goal-​democratize-​dna#.WzJm7RMbPOS. [Accessed
June 23, 2018.]
UNESCO. (1997). The Universal Declaration on the Human Genome and Human Rights. http://​
unesdoc.unesco.org/​images/​0012/​001229/​122990eo.pdf. [Accessed June 23, 2018.]
Wolbring, G. (2008). Why NBIC? Why human performance enhancement? Innovation: The European
Journal of Social Science Research, 21, 25–​40. https://​doi.org/​10.1080/​13511610802002189.
World Anti-​Doping Agency. (2003). Prohibited List. Montreal: WADA.
World Anti-​Doping Agency. (2011). Prohibited List. Montreal: WADA.
World Anti-​Doping Agency. (2015). World Anti-​Doping Code. Montreal: WADA.
World Anti-​ Doping Agency. (2017). Minutes of the Health, Medical and Research Committee.
Montreal: WADA. www.wada-​ama.org/​sites/​default/​files/​resources/​files/​minutes_​hmr_​committee_​
august_​29-​30_​2017_​final.pdf. [Accessed June 23, 2018.]
World Anti-​Doping Agency. (2018). Prohibited List. Montreal: WADA.

474
475

34
THE ETHICS OF SEX
TESTING IN SPORT
Jaime Schultz

Sport is one of the few remaining institutions to mandate sex segregation, but defining sex is
not as clear cut as it may seem. Since at least the 1930s, athletic authorities have been consumed
with the problem of how to distinguish between women and men. More accurately, they have
been consumed with which characteristics best determine femaleness. In the process, they have
relied on a range of “sex tests”: physical examinations, sex chromatin tests, genetic analyses,
and hormonal assessments, all of which have disqualified women with differences of sex devel-
opment (DSD) that may or may not relate to athletic potential (48).1 In their efforts to level
the playing field for women athletes, sport policymakers have generated more questions than
answers when it comes to solving the riddles of sex.

An early history of sex verification, 1936–​1966

In the late 18th and early 19th centuries, men established elite sports organizations, including the
International Olympic Committee (IOC) and the International Amateur Athletic Federation
(IAAF, now the International Association of Athletic Federations), to govern male athletes and
their competitions. Early organizers were not interested in providing opportunities to women.
Their general attitude was that women were too weak or frail to engage in the rigors of highly
competitive sport. Those women who proved otherwise, the logic continued, must not be
“real” women.
In the 1920s, the IOC and the IAAF grudgingly relented to pressures to add events for
women, but progress was slow and limited. Track and field, one of only four Olympic discip-
lines open to women at the 1928 Games in Amsterdam, The Netherlands, especially troubled
organizers who argued that the events were “profoundly unnatural” for female competitors and
would cause them to “develop wholly masculine physiques and behavior traits” (39). Critics
openly disputed the sex/​gender of several women in the early days of athletics and in 1936,
IOC member Avery Brundage “roundly recommended that all women athletes entered in the
Olympics be subjected to a thorough physical examination to make sure they were really 100%
female” (34). That same year, at the Olympic Games in Berlin, Germany, officials reportedly
demanded that American Helen Stephens prove her sex after her superlative performance in
the 100-​meter race. Stephens’s biographer writes that the athlete’s response was that reporters

475
476

J. Schultz

could “check the facts with the Olympic committee physician who sex-​tested all athletes prior
to competition” (14).
The story of Heinrich Ratjen makes it unlikely that physicians tested “all athletes” before
the Games. At age 17, Ratjen, who competed under his given name Dora, took fourth in the
1936 Olympic women’s high jump. Ratjen was not a “gender fraud,” as many accounts have
unjustly characterized him. Rather, he was someone who did not fit neatly into the sex/​gender
binary (18). The details of his life are murky and were only made public in 2009 after the
Department for Sexual Medicine at Kiel University Hospital posthumously released Ratjen’s
file to a German newspaper.The records revealed that a midwife declared Ratjen female at birth
and his parents raised him as such. Two years after the 1936 Berlin Games, a German police
officer questioned Ratjen’s sex –​judging his feminine clothing incongruent with his mascu-
line features. A subsequent medical examination reclassified Ratjen as male, and he lived the
remainder of his life as a man (1).
Ratjen would have likely failed an Olympic Games sex check in 1936, yet there is no
evidence to suggest that any of his contemporaries doubted his eligibility to compete as a
woman. However, several athletes in the 1930s and 1940s underwent female-​to-​male gender
confirmation surgeries (6). Their stories further stoked anxieties about the “femininity” of
women athletes and authorities escalated their discussions about the need for compulsory sex
verification (37).
Although World War II halted most major international sports competitions, the conflict
did little to quell apprehensions about the sex of women athletes. To the contrary, in the post-​
war era, the outstanding achievements of women “from Communist countries” who, were
“of questionable femininity” (according to Western journalists) only intensified those concerns
(46). In 1946, the IAAF required all women who competed in women-​only events to present
a physician’s letter confirming their sex. The IOC instituted the same requirement at the 1948
Olympic Games in London, though neither organization defined the parameters of femaleness.
Soon, the growing significance of sport in the Cold War era made it clear that affidavits from
team and family physicians could no longer suffice –​the documents were too easily falsified in
the quest for athletic supremacy.
In the 1960s, sports authorities grew increasingly concerned with two issues –​doping and
sex –​and experimented with different ways to test for both (29). In 1966, the IAAF initiated its
first standardized, precompetition sex test by requiring female athletes at the British Empire and
Commonwealth Games in Kingston, Jamaica, to undergo gynecological examinations. IAAF
officials did not tell the women what to expect, as pentathlete Mary Peters described in her
autobiography. Just before the competition, officials forced her to undergo what she called “the
most crude and degrading experience I have ever known in my life. I was ordered to lie on the
couch and pull my knees up … Presumably they were searching for hidden testes” (36).
Perhaps sympathetic to the distress the exams caused competitors, the IAAF changed its
procedures later that year at the European Athletics Championships in Budapest, Hungary.
There, female athletes went through a visual inspection or “nude parade.” As Time reported,
“The examination, as it turned out, was perfunctory. Lined up in single file, the 234 female
athletes paraded past three female gynecologists. ‘They let you walk by,’ said one competitor
afterward. ‘Then they asked you to turn and face them, and that was it’” (38). All participants
passed the survey, although there were five women who did not attend the competition. Italy’s
Maria Vittoria Tria declared that her religious convictions did not allow her “to undress in front
of unknown people” (38).The other athletes, one from Romania and three from Russia, did not
explain their absences and abrupt retirements from sport, which all but validated speculations
about their “manly” appearance and skills (52).2

476
477

The ethics of sex testing in sport

The chromosome formula, 1967–​1988

Athletic officials next turned to the buccal smear, which was, at the time, a relatively new
cytological test. Specifically, technicians looked for the presence of the second inactive X
chromosome or Barr body. Women who tested “positive” for the Barr body earned their
“certificate of femininity” or “GV [gender verification] card,” a license they could submit
at future competitions to avoid retesting (16). Women who tested “negative” for the Barr
body could opt for additional examinations –​a full chromosome analysis and, if neces-
sary, a gynecologic exam, after which authorities ruled on their eligibility. More often,
though, athletes who failed at the first stage withdrew from competition to avoid additional
humiliation (8).
The IAAF first introduced the sex chromatin test at the 1967 European Cup in Kiev, Ukraine –​
the “Sex-​Check Track Meet,” as the press called it (42). Polish sprinter Ewa Kłobukowska, who
passed the visual inspection the previous year, became the buccal smear’s first athletic casualty
when authorities determined she had “one chromosome too many” (presumed to be 46,XY/​
47,XXY mosaicism) (46, 47). At 21 years old, Kłobukowska could no longer compete in the
sport to which she had devoted her life. “It’s a dirty and stupid thing to do to me,” she said at
the time. “I know what I am and how I feel” (40).
Despite a brief protest from Kłobukowska and her compatriots, members of the IOC’s newly
established Medical Commission lauded the laboratory-​based assessment as “simpler, objective
and more dignified” than earlier verification procedures (16). The commission experimented
with “the chromosome formula” at the 1968 Winter Games in Grenoble, France, and officially
introduced it for all female participants competing in women’s sports at the 1968 Mexico City
Summer Games (35). As Eduardo Hay, chief of testing for the IOC’s Medical Commission
explained, “The investigation for femininity of the athletes participating in the Olympic Games
verifies that the athletes are competing on an equal basis considering their physical status. In
the cases of intersexuality or hermaphroditism, the athlete must be barred from competition
in order to insure fair play” (16). In telling language, a 1968 issue of the IOC’s newsletter
proclaimed that from that point on, “All women athletes participating in the Games will be
controlled” (25).
From the beginning, members of the international medical community denounced sex
testing as “grossly unfair,” arguing that “no single index or criterion can signify the appropriate
sex for an individual.” The sex chromatin test, anatomist Keith L. Moore stressed in 1968,
“should not be used as absolute criteria of sexual identity” (33). As early as 1969, specialists
refused to administer the exam on the grounds that it was scientifically unreliable and ethically
unsound (13). Danish researchers issued a memorandum in 1972 in which they asserted that,
“the International Olympic Committee has made its own definition of sex” and recommended
that, “the use of the test should be cancelled” (50). Noted geneticist Albert de la Chapelle called
the test “not only inaccurate but also discriminatory in that it excludes women who should be
allowed to participate” (8). Even so, these condemnations did little to shake the IOC’s confi-
dence in the procedure, and sex testing persisted for the next 30 years.
It is impossible to know how many women have been disqualified by the results of their
sex tests. In the wake of Ewa Kłobukowska’s humiliation in 1967, officials pledged to keep the
results confidential. Experts estimate that the examinations have eliminated one or two women
at each Olympic Games, but this does not accurately represent the number of women athletes
with DSDs –​only those who make it to the Olympic stage. It says nothing of women in the
lower or recreational echelons of sport. Potential female Olympians who do not neatly align
within the sporting parameters of femaleness are likely to be ruled out at the local, regional, or

477
478

J. Schultz

national levels and do not make it to the international stage. Even women not directly affected
by the tests might avoid sport if they suspect they might not pass.
Because organizations pledged confidentiality, it was only when a woman openly criticized
or challenged her results that the public learned of her sex-​test “failure.” This was the situation
for Spanish hurdler María José Martínez-​Patiño, who received her “GV card” in 1983 but
neglected to bring it to the 1985 World University Games in Kobe, Japan. Authorities tested
her again and were troubled by the results (which backed critiques that the tests were unreli-
able). They advised Martínez-​Patiño to fake an injury and drop out of the race. She complied.
Two months later, she received a letter reading, “Karyotype is decided 46, XY” (30). Genetically
speaking, and much to her astonishment, the IAAF had classified Martínez-​Patiño as male.
Upon further testing, Martínez-​Patiño learned –​for the first time –​that she also has com-
plete androgen insensitivity syndrome (CAIS), in which her cells do not have typical androgen
receptors and, consequently, cannot respond to circulating testosterone secreted by internal
testes. There may be high or “male” levels of androgens in her body, but this does not affect
her athletic aptitude. Although women with CIAS appear to be genetically male, they do not
develop the strength, musculature, or body types associated with male levels of testosterone;
neither can they benefit from the use of anabolic steroids or exogenous testosterone (that is,
testosterone that originates outside an individual’s body). “I could hardly pretend to be a man,”
Martínez-​Patiño wrote of the ordeal. “I have breasts and a vagina. I never cheated. I fought my
disqualification” (30). After nearly three years, and with the support of Dr. de la Chapelle, the
European Athletics Association allowed her to compete as a woman, though the trial signifi-
cantly and irreversibly damaged her life and athletic career (31).
As a result of Martínez-​Patiño’s case and other mounting objections, the IAAF dropped the
sex chromatin test in 1988 and substituted in its place a “medical examination for the health
and well-​being of all athletes (men and women).” This would, according to IAAF-​affiliated
researchers, “ensure satisfactory physical status for competition and would, of course, include
simple inspection of the external genitalia” (29). Four years later, the IAAF recognized that con-
temporary sports clothing and urine samples, collected in the presence of anti-​doping officials,
made the “health screen” unnecessary for identifying anomalous or ambiguous genitalia. As of
May 1992, a federation spokesperson declared, “We have no femininity list—​the file is closed”
(53). And yet, the file was not entirely closed, but rather left ajar for officials “to be able to
address the occasional anomalies that do surface,” the policy explained (21).

The search for SRY, 1991–​1999

Despite the IAAF’s change in protocol, IOC officials remained “wedded to the notion that
gender testing was necessary to prevent masquerading males from infiltrating female-​only
events” (49). In 1991, members of the IOC’s Medical Commission replaced the sex chromatin
test with DNA analysis of the Y chromosome using polymerase chain reaction amplification,
looking specifically for the SRY gene (sex-​determining region Y) to define an individual’s sex.
In other words, female athletes could have a Y chromosome, but not a specific gene located
on that chromosome if they wanted to compete in women-​only events. As before, experts
criticized the efficacy of this test –​indeed, of any test –​for determining sex (7, 41).
Genetic testing did not last long. At the 1996 Centennial Olympic Games in Atlanta,
Georgia, 8 women out of 3387 (1 in 423) expressed the SRY gene.3 Additional examinations
found that seven of the women had androgen insensitivity syndrome (four complete, three par-
tial), and one woman had 5α-​reductase deficiency, a disorder in which testosterone cannot be

478
479

The ethics of sex testing in sport

converted to its active metabolite, dihydrotestosterone; she had also undergone a gonadectomy
(10). Authorities cleared all eight women to compete as women at the 1996 Games.
That all of the SRY women ultimately competed, in addition to the considerable costs
and time the analyses required, added support to those who called for the elimination of
sex-​testing procedures. The Norwegian medical community collectively refused to admin-
ister the examinations at the 1994 Winter Games in Lillehammer, and the Norwegian parlia-
ment subsequently banned genetic testing for the purpose of sex verification in sport (11). By
1996, nearly every major medical society had called for the end of these practices. Additional
recommendations from the IOC World Conference on Women and Sport and the IOC’s
Athletes’ Commission finally convinced the IOC’s Executive Board to concede. The 2000
Summer Games in Sydney, Australia, marked the first the first time in more than three decades
that female Olympians could compete without first proving their sex.

Suspicion-​based testing, 2000–​2011

Still, officials could resurrect the tests at any time. Like the IAAF before it, the IOC resolved
to intervene when someone cast aspersions on the sex of a female competitor. In the event
of “any ‘suspicion,’ ” according to the guidelines, “or if there is a ‘challenge’ [to her sex]
then the athlete concerned can be asked to attend a medical evaluation” (21). Instead of
comprehensive sex testing, authorities targeted only those women of ambiguous femin-
inity. This became standard operating procedure for most national and international sports
organizations until 2011.
Shortly after India’s Santhi Soundarajan finished second in the 800-​meter race at the 2006
Asian Games in Doha, Qatar, she found herself ensnared in this very procedure. Although
her birth certificate affirmed her as female and she identified and lived her entire life as such,
the Indian Olympic Association determined that Soundarajan “did not possess the sexual
characteristics of a woman” (20). Without explaining what those characteristics were, officials
stripped her of her silver medal and banned her from future competitions for what they called
a “Games rule violation.” Soundarajan attempted suicide not long after her expulsion. As she
explained to reporters, “I am physically and mentally totally broken” (5).
South Africa’s Caster Semenya likewise found herself caught in the crosshairs of suspicion-​
based testing at the 2009 track and field World Championships in Berlin, Germany. As the
18-​year-​old Semenya cruised to victory in the 800-​meter race, Australian journalists published
unsubstantiated rumors that she “has a chromosomal abnormality that gives her both male and
female characteristics” and that her “testosterone levels were three times the normal level for
a woman” (15). IAAF officials did not comment on the specifics of Semenya’s case, but they
did confirm that they had subjected her to “gender verification” procedures amid “concerns
that she does not meet the requirements to compete as a woman” (44). The federation allowed
Semenya to retain her 800-​meter title and prize money, but requested she withdraw from sub-
sequent competitions while a medical committee, which included a gynecologist, endocrin-
ologist, psychologist, internal medicine specialist, and a “gender expert” reviewed her tests (43).
In July 2010, nearly a full year after Semenya’s last race, IAAF administrators announced
they had accepted “the conclusion of a panel of medical experts that she can compete with
immediate effect” (22). She was noticeably slower upon returning to the track, leading to specu-
lation that, in the interim, she had undergone some type of treatment to reduce her testos-
terone (44). Although Semenya has never addressed that rumor, the IAAF’s succeeding policy
on hyperandrogenism seemed to support the theory.

479
480

J. Schultz

The testosterone question

In April 2011, an IAAF-​appointed panel of experts, working closely with the IOC’s Medical
Commission, released a 28-​page document titled Regulations Governing Eligibility of Females with
Hyperandrogenism to Compete in Women’s Competition. The new policy replaced “the IAAF’s pre-
vious Gender Verification Policy and the IAAF has now abandoned all reference to the termin-
ology ‘gender verification’ and ‘gender policy’ in its Rules” (23). The test was no longer about
sex, the federation maintained; it was about health, fair play, and testosterone.
Hyperandrogenism, as defined in the IAAF’s document, is “the excessive production of
androgenic hormones (Testosterone).” It quantified “excessive” as women with 10 or more
nanomoles of functional testosterone per liter of serum (nmol/​L). This threshold is above the
“normal” reference range for women (0.12–​1.79 nmol/​L) and at the low end for “normal” men
(7.7–​29.4 nmol/​L) (2). Hyperandrogenic women (those who were not androgen insensitive)
had two options: drop out of sport or agree to a “prescribed medical treatment” that would
“correct” their “disorder” (23).
Indian sprinter Dutee Chand refused both. Instead, she challenged the Sports Authority of
India’s 2014 decision that disqualified her based on a diagnosis of hyperandrogenism. Chand
took her case before the Court of Arbitration for Sport (CAS), where adjudicators ruled
that there was “insufficient evidence about the degree of advantage that androgen-​sensitive
hyperandrogenic females enjoy over non-​hyperandrogenic females.” CAS gave the IAAF 2 years
to produce a conclusive “correlation” between high endogenous testosterone and enhanced
athletic performance. Failure to provide evidence to this effect would void the Federation’s
policy (9).
In 2017, scientists declared that they had found the evidence. In a study funded by the IAAF
and the World Anti-​Doping Agency (WADA), researchers divided track and field competitors –​
men and women –​into three groups based on their testosterone. No significant patterns
emerged for men, but women with the highest levels of testosterone outperformed women
in the lowest tertile in five events: the 400-​meter race (in which high-​testosterone females
performed 2.7% better than low-​testosterone females), 400-​meter hurdles (2.8%), 800-​meter
race (1.8%), pole vault (2.9%), and hammer throw (4.5%) (2).
Based on this study, the IAAF approved new regulations in 2018 that only apply to “rele-
vant athletes” who compete internationally in “restricted events”: the 400 meters, 400-​meter
hurdles, 800 meters, 1500 meters, and 1-​mile races. Curiously, the 2017 study did not find that
testosterone affected performance in the 1500-​meter and mile races. Even more curiously, the
two events that showed the widest performance gaps, the pole vault and hammer throw, are not
included in the restricted events. “Relevant athletes,” as stipulated in the 2018 regulations, are
women with testosterone >5 nmol/​L (half the threshold of the 2011 policy) who are diagnosed
with a specific DSD, including 5α-​reductase type 2 deficiency, partial androgen insensitivity
syndrome, 17β-​hydroxysteroid dehydrogenase type 3 deficiency, congenital adrenal hyperplasia,
3β-​hydroxysteroid dehydrogenase deficiency, ovotesticular DSD, or any other genetic condi-
tion involving disordered gonadal steroidogenesis. These athletes must lower their testosterone
to below 5 nmol/​L for 6 months prior to competition. Those who refuse or fail to do so can
compete in a nonrestricted event, in male events, or in “any applicable intersex or similar clas-
sification that may be offered” (24).
The logic is even more confounding when juxtaposed with testosterone-​related regulations
for trans athletes. In 2015, the IOC released a document that outlined the results of a Consensus
Meeting on Sex Reassignment and Hyperandrogenism. It is important not to conflate women
with hyperandrogenism and DSDs with transgender athletes. Still, the proposed eligibility

480
481

The ethics of sex testing in sport

requirements both isolate testosterone levels as the determining factor for who counts as a
woman. The IOC no longer requires “surgical anatomical changes as a pre-​condition” for trans
athletes to compete as the sex with which they identify. But a trans woman athlete (one who
transitions from male to female) “must demonstrate that her total testosterone level in serum
has been below 10 nmol/​L for at least 12 months prior to her first competition.” She must
remain below that threshold throughout her athletic career (26). The effect of this seemingly
progressive rule is to reinforce the idea that testosterone levels define sex, which strengthens the
proposed policy on hyperandrogenism.
Whereas officials wrestle over the “requirements to compete as a woman,” they appear
unconcerned with identifying any requirements to compete as a man. According to the IOC’s
2015 consensus, trans male athletes (who transition from female to male) “are eligible to com-
pete without restriction.” The document provides no exclusionary provisions for trans men,
including the use of exogenous testosterone (26). Under different circumstances, the use of
exogenous testosterone is illegal; it is doping. But WADA, which guides most organizations’ pol-
icies on performance-​enhancing drugs (including the IOC’s), allows athletes to take a banned
substance if they can prove they have a medical necessity through what is called a Therapeutic
Use Exemption (TUE).
To that end, WADA’s 2016 policy recognizes that “androgen therapy is essential for comple-
tion of the anatomical and psychological transition process in FtM [female-​to-​male] athletes.”
As such,“FtM athletes may be granted a TUE,” though the specific eligibility criteria “is entirely
left to the different sporting federations and organizations” (54). Hyperandrogenic and trans
women must suppress their endogenous testosterone, but trans men are permitted the use of
exogenous testosterone.
What is more, cis men (those affirmed male at birth and who identify as men) who drop below
the “normal” male-​typical testosterone threshold can petition for a TUE for exogenous testos-
terone if a physician diagnoses them with an “androgen deficiency.” And although most men
fall within the expansive 7.7–​29.4 nmol/​L range, sports organizations do not impose an upper
limit for male athletes’ endogenous testosterone in the same way they do for hyperandrogenic
and trans women.
At the same time, WADA’s policy makes it clear, in bold font, that a “TUE for androgen
deficiency should not be approved for females” (54). Even if a woman’s functional testosterone
registers below the female-​typical range, she is not granted the same opportunity as a man to
raise her testosterone to levels commensurate with her same-​sex competitors.This becomes not
just an issue of sport performance, but can also present significant health concerns.

The impossibility of fair

The discussion of endogenous and exogenous testosterone highlights the complexities of
defining sex for the purpose of elite sport competition (2, 17, 27). It also underscores the com-
plexities, perhaps the impossibilities, of determining what is fair. Authorities have long argued
that sex testing is necessary “to insure fair play” and “to create a level playing field in female
sport” (16, 19).Yet, sport is inherently unfair. The playing field is irredeemably unlevel. As other
authors in this book point out, there are genetic variations that predispose athletes for enhanced
performance, but authorities do not rule these athletes ineligible to compete or ask them to
suppress any potentially advantageous condition those variations might confer. It is only when
a condition challenges the ways we define sex that policymakers intervene.
On the issue of inequality, we might further consider the social, environmental, and class-​
based inequities that advantage certain athletes over others. Even on the specific issue of

481
482

J. Schultz

hyperandrogenism, a number of women diagnosed with the condition come from “rural or
mountainous regions of developing countries,” as one study concluded (12). These women
may not have the same access to healthcare, much less the advanced sport training, facilities,
equipment, or competitive opportunities as their international peers. Is that fair?
Debates about the need for sex testing invariably lead to considerations about the necessity
of sex segregation in sport. Researchers estimate that there is, on average, a 10% performance
gap between elite male and female athletes –​that men perform about 10% better than women
in the highest ranks of competition. This varies according to the sport. The difference between
the best men and women’s times in the 800-​meter freestyle swimming event is 5.5%, but in
weightlifting the top men outperform the top women by 36.8% (51). To eliminate sex segre-
gation would put most women athletes at a disadvantage. Still, the dilemma remains: By what
criteria should we define sex? Even more, do those criteria account for the performance gap?
There are no conclusive answers to these questions. Some researchers have proposed that
instead of sex segregation, we divide competitors by “sports sex” that would be “determined by
biological parameters … and not linked to gender categories, but rather only focused on valid
in relation to athletic performance” (32). The word “sex” in “sports sex” is misleading, then.
Instead, the proposal might amount to something akin to the use of “sport classes” in parasport,
whereby evaluators classify athletes according to impairment and activity limitation. But there
are problems with this system as well. Any attempt to equalize the field of play will cause con-
troversy. This much is evident in the history of sex testing, and it is a history that is unlikely to
be brought to a conclusion any time soon.

Notes
1 Sex testing goes by several alternative terms, including sex control, sex determination, gender testing,
gender verification, and femininity control. In this chapter, I refer to sex as the biological characteristics
associated with male and female, and gender the culturally constructed characteristics associated with
masculinities and femininities.
2 Administrators at the 1967 Pan American Games in Winnipeg, Manitoba, also subjected women to a
“nude parade,” an experience 16-​year-​old American shot putter Maren Seldler described as “hideous”
and “humiliating” (28).
3 At the 1992 Olympic Games in Barcelona, Spain, technicians screened 2406 female athletes. Five
women tested positive for the SRY gene, but the IOC provided no further information about the
women’s eligibility (45).

References
1. Berg S. How Dora the man competed in the women’s high jump. Spiegel, 15 Sept. 2009. www.
spiegel.de/​international/​germany/​1936-​berlin-​olympics-​how-​dora-​the-​man-​competed-​in-​the- ​
woman-​s-​high-​jump-​a-​649104.html. [Accessed 13 Nov. 2017.]
2. Bermon S, Garnier P. Serum androgen levels and their relation to performance in track and
field: mass spectrometry results from 2127 observations in male and female elite athletes. British
Journal of Sports Medicine 51: 1309, 2017.
3. Bermon S, Garnier PY, Hirschberg AL, Robinson N, Giraud S, Nicoli R, Baume N, Saugy
M, Fénichel P, Bruce SJ, Henry H. Serum androgen levels in elite female athletes. Journal of Clinical
Endocrinology Metabolism 99: 4328–​4335, 2014.
4. Bermon S, Ritzén M, Hirschberg AL, Murray TH. Are the new policies on hyperandrogenism
in elite female athletes really out of bounds? American Journal of Bioethics 13: 64, 2013.
5. Bhowmick N, Thottam J. Gender and athletics: India’s own Caster Semenya. Time 1 Sept. 2009.
www.time.com/​time/​world/​article/​0,8599,1919562,00.html. [Accessed 24 Jan. 2010.]
6. Change of sex. Time 28: 39–​40, 1936.

482
483

The ethics of sex testing in sport

7. Cotinot C, Pailhoux E, Jaubert F, Fellous M. Molecular genetics of sex determination. Seminars

in Reproductive Medicine 20: 157–​68, 2002.
8. de la Chapelle A. The use and misuse of sex chromatin screen for “gender identification” of female
athletes. Journal of the American Medical Association 256: 1920–​1923, 1986.
9. Dutee Chand v.Athletics Federation of India and the International Association of Athletics Federation,
CAS 2014/​A/​3759.
10. Elsas LJ, Hayes RP, Muralidharan K. Gender verification at the centennial Olympic games.
Journal of the Medical Association of Georgia 86: 50, 1997.
11. Elsas LJ, Ljungqvist A, Ferguson-​Smith MA, Simpson JL, Genel M, Carlson AS, Ferris
E, de la Chapelle A, Ehrhardt AA. Gender verification of female athletes. Genetics in Medicine
2: 249–​254, 2000.
12. Fénichel P, Paris F, Philibert P, Hiéronimus S, Gaspari L, Kurzenne JY, Chevallier P,
Bermon S, Chevalier N, Sultan C. Molecular diagnosis of 5α-​reductase deficiency in 4 elite young
female athletes through hormonal screening for hyperandrogenism. Clinical Endocrinology Metabolism
98: E1055–​E1059, 2013.
13. Ferris EA. Gender verification testing in sport. British Medical Bulletin 48: 692, 1992.
14. Hanson SK. The Life of Helen Stephens: The Fulton Flash. Carbondale, IL: Southern Illinois University
Press, 2004.
15. Hart S. Caster Semenya “is a hermaphrodite,” tests show. Sydney Herald 11 Sept. 2009. www.tele-
graph.co.uk/​sport/​othersports/​athletics/​6170229/C ​ aster-​Semenya-​is-​a-​hermaphrodite-​tests-s​ how.
html. [Accessed 13 Nov. 2017.]
16. Hay E. Sex determination in putative female athletes. Journal of the American Medical Association
4: 39–​41, 1972.
17. Healy ML, Gibney J, Pentecost C, Wheeler MJ, Sonksen PH. Endocrine profiles in 693 elite
athletes in the postcompetition setting. Clinical Endocrinology 81: 294–​305, 2014.
18. Heggie V. Testing sex and gender in sports; Reinventing, reimagining and reconstructing histories.
Endeavour 34: 157–​63, 2010.
19. Ingle S. Sebastian Coe: IAAF right to seek court ruling over hyperandrogeism issue. Guardian 12 Aug.
2017. www.theguardian.com/​sport/​2017/​aug/​12/​sebastian-​coe-​iaaf-​hyperandrogenism. [Accessed
13 Nov. 2017.]
20. Indian runner fails gender test, loses medal. ESPN, 18 Dec. 2006. http://​sports.espn.go.com. [Accessed
16 Jan. 2007.]
21. International Association of Athletics Federations. IAAF Policy on Gender Verification
(Technical Report). 2006. www.iaaf.org.
22. International Association of Athletics Federations. Caster Semenya may compete. 6 July 2010.
www.iaaf.org/​news/​iaaf-​news/​caster-​semenya-​may-​compete. [Accessed 12 July 2012.]
23. International Association of Athletics Federations. IAAF Regulations Governing Eligibility of
Females with Hyperandrogenism to Compete in Women’s Competition (Technical Report). 2011.
www.iaaf.org/​medical/​policy/​. [Accessed 12 July 2012.]
24. International Association of Athletics Federations. Eligibility Regulations for the Female
Classification (Athletes with Differences of Sex Development). 23 April 2018. www.iaaf.org/​about-​
iaaf/​documents/​rules-​regulations#collapseregulations. [Accessed 1 Sept. 2017].
25. International Olympic Committee. The work of the Medical Commission. Olympic Review
10: 269, 1968.
26. International Olympic Committee. IOC Consensus Meeting on Sex Reassignment and
Hyperandrogenism (Technical Report). Nov. 2015 https://​stillmed.olympic.org/​. [Accessed 17
July 2017.]
27. Karkazis K, Jordan-​Young R. Debating a testosterone “sex gap.” Science 348: 858–​869, 2015.
28. Larned D. The femininity test: a woman’s first Olympic hurdle. WomenSports, July: 9, 1976.
29. Ljungqvist L, Simpson JL. Medical examination for health of all athletes replacing the need for
gender verification in international sports: The International Amateur Athletic Federation plan.
Journal of the American Medical Association 267: 850–​852, 1992.
30. Martínez-​Patiño MJ. Personal account: a woman tried and tested. Lancet 366: S38, 2005.
31. Martínez-​Patiño MJ, Mateos-​Padorno C, Martínez-​Vidal A, Sánchez Mosquera AN, García
Soidán JL, Díaz Pereira MD, Touriño González CF. An approach to the biological, historical
and psychological repercussions of gender verification in top level competitions Journal of Human
Sport and Exercise 5: 307–​321, 2010.

483
484

J. Schultz

32. Martinez-​Patino MJ, Vilain E, Bueno-​Guerra N. The unfinished race: 30 years of gender verifi-
cation in sport. Lancet 388: 541–​543, 2016.
33. Moore KL.The sexual identity of athletes. Journal of the American Medical Association 205: 163–​164, 1968.
34. Olympic Games. Time 28: 46, 1936.
35. Olympics require sex test. New York Times 30 June: 48, 1968.
36. Peters M with Wooldridge I. Mary P.: Autobiography. London: Paul, 1974, pp. 55–​56.
37. Pieper LP. Sex Testing: Gender Policing in Women’s Sports. Urbana, IL: University of Illinois Press,
2016, 29–​31.
38. Preserving la difference. Time, 16 Sept.: 74, 1966.
39. Rand FR. Olympics for girls? School & Society 30: 194, 1929.
40. Records of Polish girl sprinter who flunked sex test barred. New York Times 26 Feb.: 50, 1968.
41. Ritchie R, Reynard J, Lewis T. Intersex and the Olympic Games. Journal of the Royal Society of
Medicine 101: 395–​399, 2008.
42. Russians win sex-​check track meet. Washington Post 16 Sept.: D2, 1967.
43. Schultz J. Caster Semenya and the “question of too”: sex testing in elite women’s sport and the issue
of advantage. Quest 1: 228–​43, 2011.
44. Schultz J. Qualifying Times: Points of Change in U.S. Women’s Sport. Urbana, IL: University of Illinois
Press, 2014.
45. Serrat A, García de Herreros A. Determination of genetic sex by PCR amplification of Y-​
chromosome-​specific sequences. Lancet 341: 1593, 1993.
46. Sex test disqualifies athlete. New York Times 16 Sept.: 28, 1967.
47. Simpson JL. Gender verification in competitive sports. Sports Medicine 16: 305–​315, 1993.
48. Simpson JL, Ljungqvist A, Ferguson-​Smith MA, de la Chapelle A, Elsas LJ, Ehrhardt AA,
Genel M, Ferris EA, Carlson A. Gender verification in the Olympics. JAMA 284: 1568–​1569, 2000.
49. Stephenson J. Female Olympians’ sex tests outmoded. JAMA 276: 178, 1996.
50. Strömgren E. A memorandum on the use of sex chromatin investigation of competitors in women’s
divisions of the Olympic Games. 3 Feb. 1972, IOC files folder: “Danish Report, “Sex Tests,” 1972.
The International Olympic Committee’s Historical Archives, Lausanne, Switzerland.
51. Thibault V, Guillaume M, Berthelot G, El Helou N, Schaal K, Quinquis L, Nassif H, Tafflet
M, Escolano S, Hermine O, Toussaint JF.Women and men in sport performance: the gender gap
has not evolved since 1983. Journal of Sports Science and Medicine 9: 214–​223, 2010.
52. Teitel J. Faster, higher, stronger. Toronto Life Dec.: 73, 1983.
53. Vines G. Last Olympics for sex test? New Scientist 135: 39–​42, 1992.
54. World Anti-​Doping Agency. TUE Physician Guidelines, Medical Information to Support the
Decisions of TUE Committees: Female-​to-​Male (FtM) Transsexual Athletes (Technical Report).
March 2016. www.wada-​ama.org. [Accessed 13 Nov. 2017.]

484
485

SECTION 7

Conclusions

485
486
487

35
EXERCISE GENOMICS,
EPIGENOMICS, AND
TRANSCRIPTOMICS:
A REALITY CHECK!
Claude Bouchard
Over the last three decades or so, several books have dealt with the highly specialized topic of
genetics and exercise. The first effort was more than 20 years ago when there was only min-
imal data to report across a range of topics of relevance to exercise (8). Since then, a few other
volumes have appeared, reflecting an incremental growth in content but with only a modest
paradigm shift in the underlying science (5, 20, 22). As one could expect, the present effort
under the editorial guidance of Drs. Timothy Lightfoot, Monica Hubal, and Stephen Roth is
the most up-​to-​date and comprehensive of these publications. It allows us to understand the
progress we have made but also the territory that has been only superficially explored as well as
the many issues and research questions that we have not even begun to address.
The editors have asked me for a commentary and concluding remarks probably because
I have been an actor and a witness of the evolution of research in the field of exercise genetics.
I launched my first study in the area in 1978. It took the form of several fitness and exercise-​
related traits that were measured in nuclear families and some of their extended relatives, pairs
of monozygotic and dizygotic twins, adopted siblings, and foster parents who were recruited
as part of the Quebec Family Study (6, 16). This was followed by several exercise training
experiments performed on pairs of identical twins (9, 15, 21, 23). Two major conclusions that
have stood for the last three decades were reached in these studies: first, the heritability of car-
diorespiratory fitness and exercise-​related traits is moderate and not as high as was suggested
by earlier studies; and second, there is a genetic component to human variability in trainability.
Incidentally, these conclusions were later strongly supported by the findings of the multicenter
HERITAGE Family Study (2, 4).
However, the research landscape became more complicated when, beginning around 1980,
my laboratory tried to identify DNA variants associated with cardiorespiratory fitness and other
exercise-​related traits in the sedentary state or their responsiveness to exercise training (3, 7,
10, 11). Our efforts followed closely the progress at the time in research designs and technolo-
gies used in human genetics to illuminate the genetic architecture of complex traits. Soon, it
became clear that association studies based on candidate genes or restriction fragment length
polymorphisms, or linkage studies based on panels of microsatellite markers to identify quan-
titative trait loci were insufficient to conclude about genes and alleles of relevance to complex,
multifactorial exercise phenotypes. The reasons for these failures to deliver in a reproducible

487
488

C. Bouchard

manner were not originally obvious. It took the advances in genome-​wide association studies
(GWAS) and the Encyclopedia of DNA Elements Consortium (ENCODE) project to high-
light how challenging the genomic architecture of human complex traits truly is and to define
the major reasons the earlier research paradigms failed to deliver (1). This view has been re-​
enforced by the recently launched human cell initiative (www.humancellatlas.org​) and advances
in the Genotype-​Tissue Expression (GTEx) project (www.gtexportal.org).
The early exercise genetic studies assumed that the relation between genotype and pheno-
type was rather direct and simple, that the regulation of gene expression in a given tissue was
dependent on a small number of regulatory elements, and that an allele of interest would
have a large effect size on a relevant exercise trait. In the end, it turned out that none of these
three assumptions were true. Thus complex traits are influenced by genes and alleles with small
effect sizes; the regulation of transcription, translation, and other cellular processes is widely
distributed and highly complex; and pervasive redundancy at gene, protein, and pathway levels
makes prediction even more challenging.

The human genome in numbers

An illustration of this enormous degree of genomic complexity can be obtained from the
findings of the ENCODE project, resulting from the contributions of hundreds of scientists
from around the world (www.genome.gov/​10005107/​the-​encode-​project-​encyclopedia-​
of-​dna-​elements). As partially summarized elsewhere (1), less than 2% of the human genome
sequence encodes an estimated 20,687 protein-​coding genes and about 80% of the genome is
transcribed into other encoded products involved mainly in the regulation of gene expression
and other cellular events. Interestingly, only about 10% of the single nucleotide polymorphisms
(SNPs) commonly genotyped in GWAS reside within protein-​coding sequences. The human
genome harbors almost 3 million protein-​binding sites along its DNA, which include transcrip-
tion factor-​binding sites, recombination sites, and restriction sites, but only a few thousands may
be relevant to a given cell type.The more than 1600 or so transcription factors identified to date
have been shown to bind at DNA sites representing about 8% of the genome. There are about
8800 small RNAs, including greater than 2000 miRNAs targeting large numbers of mRNAs,
and more than 9600 long noncoding RNAs of more than 200 nucleotides in length being
transcribed variously depending on cell type, and most participate in the regulation of transcrip-
tion and translation. One last set of numbers to illustrate the inherent structural complexity of
the genome: human DNA encodes about 70,000 promoter regions, 400,000 enhancer regions,
and multiple repressor sites. These numbers are lower for a given cell type or tissue, and the
relevant sequences are often at a substantial genomic distance from the genes they are thought
to regulate. However, DNA is twisted and folded in ways that sequences located at long phys-
ical distance away from each other on the linear DNA can be actually physically close to one
another when wrapped around nucleoproteins. Thus, in the aggregate, a complex network of
regulatory molecules and DNA binding sequences are involved in the widely distributed regu-
lation of less than 21,000 protein-​coding genes.
A few additional numbers can help us appreciate the extraordinary complexity of the human
genome. There are on average as many as 250 DNA protein-​binding sites per protein-​coding
gene, each transcription factor could relate to an average of 15 protein-​coding genes, and there
are on average about 23 promoter and enhancer regions per protein-​coding gene.The biological
instructions encoded in the human genome are vastly more complex than we thought.This fact
alone should give us pause when designing new studies or when attempting to conclude about

488
489

Genomics, epigenomics, and transcriptomics

the functional relevance of DNA variants, epigenetic marks, or transcript abundance, taken in
isolation, for a given exercise trait.
Additionally, cytosines in CpG dinucleotides can be methylated to form 5-​methylcytosine
under the action of DNA methyltransferases. Altering the methylation status of cytosines in a
gene promoter region has the potential to influence gene expression.There is growing evidence
that there is a dose–​response relation between the number of methylated CpG dinucleotides
in a given promoter/​regulatory region and the expression level of the targeted gene. It has
been shown that there are about 28 million CpG sites and more than 30,000 genomic regions
where these sites are concentrated into CpG islands of up to 1500 base pairs long in the human
genome. Of course, one also needs to incorporate the array of histone chemical modifications
to cytosine methylation levels in order to draw a more complete picture of the impact of given
epigenetic signatures on gene expression.

Genomic architecture of complex traits

The impetus provided by the impressive wave of GWAS publications on complex traits over the
last decade or so has nurtured a major shift in our understanding of the genomic architecture
of these traits. Unfortunately, adequately powered GWAS data are not yet available for most
exercise-​related traits, mainly because it is very difficult to generate appropriately large samples
of subjects in which laboratory-​based exercise trait measurements were obtained. Despite the
lack of solid data on exercise-​related traits, there are many lessons that we can borrow from large
studies of other complex traits to appreciate what is in all likelihood the genomic architecture
landscape of exercise traits of interest. One excellent case from which we can derive useful
lessons is in the biology of the body mass index (BMI) as illuminated by GWAS.
For decades, it was assumed in some quarters that the heritability of BMI was quite high, over
80% by some estimates. There were divergent views but they did not receive much attention.
More recently, two studies conducted on large sample sizes with large panels of SNPs have
reached the conclusion that the BMI heritability based on hundreds of SNPs was of the order
of 27% with small standard errors (13, 25). Common variants characterized by minor allele
frequencies greater than 0.4 accounted for about 80% of the heritability (25). One study from
Iceland showed that by combining information on close and distant relatives with a panel of
SNPs, the heritability of BMI reached 42% with a standard error of about 2%. Such data are
not available yet for exercise traits measured in the sedentary state or in response to standardized
exercise programs but it would not be surprising if the true heritability of exercise traits would
be lower than originally reported from genetic epidemiology and statistical genetics approaches.
Recent reports have begun to define pathways and biological systems represented by the
many SNPs associated with individual differences in BMI (14, 18, 24). As is typically observed
for complex traits, BMI common variants are characterized by small effect sizes, but they still
account for most of the SNP heritability. In the largest study to date, Turcot et al. searched for
rare (minor allele <1%) and low (minor allele <5%) frequency variants associated with BMI in
718,734 individuals from 125 studies (24). They observed 14 rare coding variants in 13 genes,
with effect size about ten times larger than what is typically seen in common SNPs. However,
because these variants are not very prevalent in human populations, they account for only a
very small fraction of the variance in BMI in populations, but they can have a large impact on
individuals who happen to be carriers of these high-​r isk BMI alleles. These findings are con-
cordant with what has been reported for rare, low, and common frequency variants affecting
human height (19).

489
490

C. Bouchard

Unpublished findings based on the HERITAGE Family Study strongly suggest that the most
important effects of common genomic variants can be imputed to alleles in noncoding regions
of the genome (Ghosh S, Bouchard C, et al, unpublished). A bioinformatics exploration of the
biology underlying a panel of SNPs associated with VO2max in the sedentary state (at P <10
× 10−5) reveals that DNA variants located in transcription factor binding sites, promoter and
enhancer sites, and CpG-​r ich areas were substantially more prevalent compared to variants in
other genomic regions.

Epilogue
Overall, these observations and our current understanding of the extraordinary complexity of
the human genome support the view that the trilogy of DNA sequence variants, epigenetic
signatures, and transcript abundance of relevant genes are all participating in the modulation
of cardiorespiratory fitness and other exercise traits of interest. However, in spite of all the pro-
gress made in genomics and bioinformatics, it remains an extraordinary challenge to define
causal relationships between DNA variants, epigenetic events or gene expression profile, and a
relevant phenotype. This is true even when adequate statistical power, strong research designs,
and state-​of-​the-​art physiological and behavioral phenotyping combined with cutting-​edge
molecular and high-​throughput omics technologies are used. Establishing a clear and unam-
biguous genotype–​phenotype connection is fraught with difficulties.
We believe that the field of exercise biology and genetics must urgently address two key
questions. First, are there DNA sequence variants, epigenetic marks in specific cell populations,
or gene expression profiles in some tissues that influence the intrinsic level of activity in chil-
dren and adults? In a nutshell, having a reasonable picture of the biology and genetics of the
intrinsic and spontaneous levels of activity would go a long way in advancing efforts to develop
more comprehensive explanations for the fact that some people enjoy being physically active
whereas most adults do not care about exercising regularly. One cannot rule out, depending on
the nature of the biological and genetic explanations, that compensatory approaches aimed at
enhancing the interest for a physically active lifestyle could arise from such advances. This topic
is of great importance and it does not receive the attention it deserves as we have been arguing
recently in a consensus paper (17).
The second question can be framed as follows: What are the molecular transducers of adap-
tation to acute and chronic exposures to exercise and are they influenced by DNA sequence
variants, epigenetic marks, and other molecular substrates? The global topic of the molecular
transducers of adaptation to exercise (endurance and resistance exercise) is currently being
addressed by the Molecular Transducers of Physical Activity Consortium (MoTrPAC), com-
prising more than 20 laboratories working in an integrated and coordinated manner with
funding support by the Common Fund of the US National Institutes of Health (www.motrpac.
org​). The research is based on animal models and human experimentations with a focus on
acute and chronic endurance exercise, and acute and chronic resistance exercise programs.
A sample of children and sets of trained endurance and resistance athletes are also incorporated
in the research program of MoTrPAC.
To conclude: we have made progress. But in some areas, the progress is superficial and will
not endure the test of time. Given the complexity of the human genome and the overarching
principles stipulating that DNA variants are characterized by small effect sizes, that the regula-
tion of gene expression is heterogeneous and widely distributed, and that biological redundancy
is pervasive, it should not be surprising that the science of exercise genomics is still in its infancy.
In fact, one should not be surprised if most of the observations reported in this volume turn

490
491

Genomics, epigenomics, and transcriptomics

out to be partially wrong or totally false once all the evidence is in. However, this is not as bad
as it may look and should not be a source of discouragement: if exercise genomics lags behind,
it is likely to be caused by the fact that it is underfunded in general and that it has been next
to impossible to put together the resources necessary to investigate a myriad of questions and
test hypotheses with appropriate study designs, under adequate experimental and reproducible
conditions.
Another reason not to be discouraged by contradictory facts and negative observations is
that, as a community of scientists, we should be thriving in the presence of the unknown.
Scientists are by definition striving to reduce the level of ignorance regarding a specific topic.
Ignorance is the driver of science as creatively proposed by Stuart Firestein from Columbia
University, New York City (12). Greater progress is achieved when an experimental observa-
tion contradicts the prevailing theory or view. It forces a revision of the theory and encourages
the formulation of new hypotheses. Fully accepting that exercise scientists are best positioned
to recognize the level of ignorance characteristic of exercise genomics in its current incarnation
could actually be a productive stand. It could lead to a greater level of doubt on isolated and
perhaps poorly reproducible facts so that the focus shifts to higher-​order levels of generaliza-
tion and theoretical models. So, let’s collectively focus on eradicating what is poor science and
on identifying clearly what is unknown in exercise genomics with the aim of devising stronger
and more comprehensive models that we can submit to experimentation aimed at reducing our
level of ignorance.

References
1. Bouchard C. Exercise genomics –​a paradigm shift is needed: a commentary. Br J Sports Med
49: 1492–​1496, 2015.
2. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, Perusse L, Leon AS, and Rao
DC. Familial aggregation of VO(2max) response to exercise training: results from the HERITAGE
Family Study. J Appl Physiol (1985) 87: 1003–​1008, 1999.
3. Bouchard C, Chagnon M, Thibault MC, Boulay MR, Marcotte M, Cote C, and Simoneau
JA. Muscle genetic variants and relationship with performance and trainability. Med Sci Sports Exerc
21: 71–​77, 1989.
4. Bouchard C, Daw EW, Rice T, Perusse L, Gagnon J, Province MA, Leon AS, Rao DC,
Skinner JS, and Wilmore JH. Familial resemblance for VO2max in the sedentary state: the
HERITAGE Family Study. Med Sci Sports Exerc 30: 252–​258, 1998.
5. Bouchard C and Hoffman EP. Genetic and Molecular Aspects of Sports Performance, Vol. XVIII of
the IOC Encyclopeadia of Sports Medicine, an IOC Medical Commission Publication. Chichester, West
Sussex: Wiley-​Blackwell, 2011.
6. Bouchard C, Lortie G, Simoneau JA, Leblanc C, Theriault G, and Tremblay A. Submaximal
power output in adopted and biological siblings. Ann Hum Biol 11: 303–​309, 1984.
7. Bouchard C and Malina RM. Genetics of physiological fitness and motor performance. Exerc Sport
Sci Rev 11: 306–​339, 1983.
8. Bouchard C, Malina RM, and Perisse L. Genetics of Fitness and Physical Performance. Champaign,
IL: Human Kinetics, 1997.
9. Bouchard C, Tremblay A, Despres JP, Theriault G, Nadeau A, Lupien PJ, Moorjani S,
Prudhomme D, and Fournier G.The response to exercise with constant energy intake in identical
twins. Obes Res 2: 400–​410, 1994.
10. Chagnon YC, Allard C, and Bouchard C. Red blood cell genetic variation in Olympic endurance
athletes. J Sport Sci 2: 121–​129, 1984.
11. Dionne FT, Turcotte L, Thibault MC, Boulay MR, Skinner JS, and Bouchard C. Mitochondrial
DNA sequence polymorphism, VO2max, and response to endurance training. Med Sci Sports Exerc
23: 177–​185, 1991.
12. Firestein S. Ignorance: How It Drives Science. Oxford: Oxford University Press, 2012.

491
492

C. Bouchard

13. Ge T, Chen CY, Neale BM, Sabuncu MR, and Smoller JW. Phenome-​wide heritability analysis
of the UK Biobank. PLoS Genet 13: e1006711, 2017.
14. Ghosh S and Bouchard C. Convergence between biological, behavioural and genetic determinants
of obesity. Nat Rev Genet 18: 731–​748, 2017.
15. Hamel P, Simoneau JA, Lortie G, Boulay MR, and Bouchard C. Heredity and muscle adapta-
tion to endurance training. Med Sci Sports Exerc 18: 690–​696, 1986.
16. Lesage R, Simoneau JA, Jobin J, Leblanc J, and Bouchard C. Familial resemblance in maximal
heart-​rate, blood lactate and aerobic power. Hum Hered 35: 182–​189, 1985.
17. Lightfoot JT, De Geus EJC, Booth FW, Bray MS, M DENH, Kaprio J, Kelly SA, Pomp D,
Saul MC, Thomis MA, Garland T, Jr., and Bouchard C. Biological/​genetic regulation of physical
activity level: consensus from GenBioPAC. Med Sci Sports Exerc 50: 863–​873, 2018.
18. Locke AE, Kahali B, Berndt SI, Justice AE, Pers TH, Day FR, Powell C, Vedantam S,
Buchkovich ML, Yang J, Croteau-​Chonka DC, Esko T, Fall T, Ferreira T, Gustafsson S,
Kutalik Z, Luan J, Magi R, Randall JC, Winkler TW, Wood AR, Workalemahu T, Faul JD,
Smith JA, Zhao JH, Zhao W, Chen J, Fehrmann R, Hedman AK, Karjalainen J, Schmidt
EM, Absher D, Amin N, Anderson D, Beekman M, Bolton JL, Bragg-​Gresham JL, Buyske
S, Demirkan A, Deng G, Ehret GB, Feenstra B, Feitosa MF, Fischer K, Goel A, Gong J,
Jackson AU, Kanoni S, Kleber ME, Kristiansson K, Lim U, Lotay V, Mangino M, Leach IM,
Medina-​Gomez C, Medland SE, Nalls MA, Palmer CD, Pasko D, Pechlivanis S, Peters MJ,
Prokopenko I, Shungin D, Stancakova A, Strawbridge RJ, Sung YJ, Tanaka T, Teumer A,
Trompet S, van der Laan SW, van Setten J, Van Vliet-​Ostaptchouk JV, Wang Z, Yengo L,
Zhang W, Isaacs A, Albrecht E, Arnlov J, Arscott GM, Attwood AP, Bandinelli S, Barrett
A, Bas IN, Bellis C, Bennett AJ, Berne C, Blagieva R, Bluher M, Bohringer S, Bonnycastle
LL, Bottcher Y, Boyd HA, Bruinenberg M, Caspersen IH, Chen YI, Clarke R, Daw EW, de
Craen AJM, Delgado G, Dimitriou M, et al. Genetic studies of body mass index yield new insights
for obesity biology. Nature 518: 197–​206, 2015.
19. Marouli E, Graff M, Medina-​Gomez C, Lo KS, Wood AR, Kjaer TR, Fine RS, Lu Y,
Schurmann C, Highland HM, Rueger S, Thorleifsson G, Justice AE, Lamparter D, Stirrups
KE, Turcot V, Young KL, Winkler TW, Esko T, Karaderi T, Locke AE, Masca NG, Ng MC,
Mudgal P, Rivas MA, Vedantam S, Mahajan A, Guo X, Abecasis G, Aben KK, Adair LS,
Alam DS, Albrecht E, Allin KH, Allison M, Amouyel P, Appel EV, Arveiler D, Asselbergs
FW, Auer PL, Balkau B, Banas B, Bang LE, Benn M, Bergmann S, Bielak LF, Bluher M,
Boeing H, Boerwinkle E, Boger CA, Bonnycastle LL, Bork-​Jensen J, Bots ML, Bottinger
EP, Bowden DW, Brandslund I, Breen G, Brilliant MH, Broer L, Burt AA, Butterworth AS,
Carey DJ, Caulfield MJ, Chambers JC, Chasman DI, Chen YI, Chowdhury R, Christensen
C, Chu AY, Cocca M, Collins FS, Cook JP, Corley J, Galbany JC, Cox AJ, Cuellar-​Partida
G, Danesh J, Davies G, de Bakker PI, de Borst GJ, de Denus S, de Groot MC, de Mutsert
R, Deary IJ, Dedoussis G, Demerath EW, den Hollander AI, Dennis JG, Di Angelantonio
E, Drenos F, Du M, Dunning AM, Easton DF, Ebeling T, Edwards TL, Ellinor PT, Elliott P,
Evangelou E, Farmaki AE, Faul JD, et al. Rare and low-​frequency coding variants alter human
adult height. Nature 542: 186–​190, 2017.
20. Pescatello LS and Roth SM. Exercise genomics. In: Molecular and Translational Medicine, edited by
Pescatello LS and Roth SM. New York: Humana Press, 2011, p. 179–​230.
21. Prud’homme D, Bouchard C, Leblanc C, Landry F, and Fontaine E. Sensitivity of maximal
aerobic power to training is genotype-​dependent. Med Sci Sports Exerc 16: 489–​493, 1984.
22. Roth SM. Genetics Primer for Exercise Science and Health. Champaign, IL: Human Kinetics, 2007.
23. Simoneau JA, Lortie G, Boulay MR, Marcotte M, Thibault MC, and Bouchard C. Inheritance
of human skeletal muscle and anaerobic capacity adaptation to high-​intensity intermittent training.
Int J Sports Med 7: 167–​171, 1986.
24. Turcot V, Lu Y, Highland HM, Schurmann C, Justice AE, Fine RS, Bradfield JP, Esko T,
Giri A, Graff M, Guo X, Hendricks AE, Karaderi T, Lempradl A, Locke AE, Mahajan A,
Marouli E, Sivapalaratnam S, Young KL, Alfred T, Feitosa MF, Masca NGD, Manning AK,
Medina-​Gomez C, Mudgal P, Ng MCY, Reiner AP, Vedantam S, Willems SM, Winkler
TW, Abecasis G, Aben KK, Alam DS, Alharthi SE, Allison M, Amouyel P, Asselbergs
FW, Auer PL, Balkau B, Bang LE, Barroso I, Bastarache L, Benn M, Bergmann S, Bielak
LF, Bluher M, Boehnke M, Boeing H, Boerwinkle E, Boger CA, Bork-​Jensen J, Bots ML,
Bottinger EP, Bowden DW, Brandslund I, Breen G, Brilliant MH, Broer L, Brumat M,

492
493

Genomics, epigenomics, and transcriptomics

Burt AA, Butterworth AS, Campbell PT, Cappellani S, Carey DJ, Catamo E, Caulfield MJ,
Chambers JC, Chasman DI, Chen YI, Chowdhury R, Christensen C, Chu AY, Cocca M,
Collins FS, Cook JP, Corley J, Corominas Galbany J, Cox AJ, Crosslin DS, Cuellar-​Partida
G, D’Eustacchio A, Danesh J, Davies G, Bakker PIW, Groot MCH, Mutsert R, Deary IJ,
Dedoussis G, Demerath EW, Heijer M, Hollander AI, Ruijter HM, Dennis JG, Denny JC, Di
Angelantonio E, Drenos F, Du M, Dube MP, Dunning AM, Easton DF, et al. Protein-​altering
variants associated with body mass index implicate pathways that control energy intake and expend-
iture in obesity. Nat Genet 50: 26–​41, 2018.
25. Yang J, Bakshi A, Zhu Z, Hemani G, Vinkhuyzen AA, Lee SH, Robinson MR, Perry JR,
Nolte IM, van Vliet-​Ostaptchouk JV, Snieder H, LifeLines Cohort S, Esko T, Milani L,
Magi R, Metspalu A, Hamsten A, Magnusson PK, Pedersen NL, Ingelsson E, Soranzo
N, Keller MC, Wray NR, Goddard ME, and Visscher PM. Genetic variance estimation with
imputed variants finds negligible missing heritability for human height and body mass index. Nat
Genet 47: 1114–​1120, 2015.

493
494

36
AFTERWORD –​ CLOSING
THE LOOP: OBSERVATIONS
AND CONCLUSIONS
J. Timothy Lightfoot, Monica J. Hubal, and Stephen M. Roth

Introduction
As we close this volume, we believe that a few points need to be made –​a sort of summary if
you will. Because the authors in each chapter have done such a great job of summarizing the
field in each of their areas, this chapter will be rather more of a 30,000ʹ overview of the whole
field. It was not long ago that many of us hoped that understanding the underlying systems gen-
etics of any trait would be as easy as understanding the gene polymorphisms present in different
individuals, a view captured and forever immortalized in pop culture in the movie “Gattaca.”
However, one thing that is clear from the previous chapters is that we are not working in the
“genetics” of 40 years ago; i.e., to paraphrase an automobile advertisement from several years
ago, “This is not your Father’s (or Gattaca’s) genetics!”
As succinctly summarized by Dr. Bouchard’s closing words (Chapter 35) and as demonstrated
in almost every chapter, complexity reigns supreme in the systems genetics of exercise and sport.
The literature has roundly dispensed with the idea that only one or a few genes control any
one exercise-​related phenotype; rather, it has become widely accepted that because all exercise
phenotypes are complex traits that are made up of numerous different physiological pathways,
the genetic mechanisms that control most –​if not all –​exercise phenotypes are made up of
hundreds, if not thousands, of genes, each contributing a small amount to the overall regulation.
Adding to this complexity, as mentioned in Chapters 7 and 9, Xu and Garland, Jr. have provided
experimental proof (10) that some exercise phenotypes may be the result of multiple different
genetic architectures –​a confirmation of the “multiple independent evolutionary solution”
hypothesis in genetics. Thus, the exercise scientist who is interested in studying exercise and
sport systems genetics must know that the potential genetic mechanisms influencing exercise
and sport phenotypes are likely to be complex, and will be difficult to not only understand,
but also to translate and apply in practice. In short, the world presented in the movie “Gattaca”
is now viewed as hopelessly naïve and outdated, especially considering the new findings and
concepts in systems genetics that are being added to the literature on an almost daily basis. But
as David Epstein pointed out in the Foreword, there are cases where a single gene variant can
markedly affect sport performance, and as such, there will probably always be a few exceptions
to the “complexity rule” that we are seeing in exercise and sport systems genetics.

494
495

Afterword – closing the loop

Writing during a time in history where politicians routinely question whether science is
truthful or relevant, scientists often take offense at any critique of their findings. However, it is
often forgotten that science only thrives when “what we think we know” is heavily scrutinized
and respectfully challenged within the scientific community. As an example of examining care-
fully what we think we know, early in the development of this project, in one of the numerous
outlines that we composed, one of us wrote the following sentence, “There is no doubt that
genetics plays a role in all exercise behaviors and little doubt that exercise behaviors are driven
by numerous genes with small effect sizes.” With the paragraphs of this chapter and throughout
much of this volume, you would think that the previous sentence would be entirely defens-
ible. However, as is appropriate, there have been several articles recently (5, 7) and at least one
formal scientific debate that have questioned whether genetics is actually involved in deter-
mining exercise phenotypes (in particular, these articles question the “nonresponder to training”
phenotype established originally by Bouchard and Rankinen (3) and covered extensively in
Section 3 of this volume). While the editors of this volume believe that the authors of those
questioning papers do not appreciate or allow for the complexity of genetic regulation, none-
theless, these types of discussions are invaluable to ensure that the science within this field is on
a solid foundation (and we look forward to including in the next edition of this volume the
upcoming published debates regarding this topic). In short, debate, critique, and examination of
our ideas and findings are critical to advancement of the field.
Along the lines of a 30,000ʹ overview, we would be remiss if we did not recognize what may
be tremendous advances coming to the field in the next few years (and probably before the next
edition of this volume) due to the rapid embrace of clustered regularly interspaced short pal-
indromic repeats (CRISPR)-​Cas9 technologies (4). While the final impact of CRISPR-​Cas9
on all systems genetics is unknown at this point, the potential of such a low-​cost, precise, and
relatively easy system of altering genomes will certainly be applied to exercise phenotypes soon,
if not being applied already. For example, early results were recently announced that CRISPR
technologies reversed a form of muscular dystrophy in a canine model (1). Additionally, given
that some countries are already allowing CRISPR-​Cas9 experimentation on nonviable embryos
(e.g., 9), one can easily imagine that this technology may be applied to viable embryos in the
future. This development will potentially forever change sport and exercise systems genetics,
and may bring sport gene-​doping (or -​editing) to the forefront quicker than we can philo-
sophically deal with it. Readers and scientists alike should continue to watch the literature for
developments in this exciting area.
It is a given that within each of the exercise phenotypes that have been covered within this
volume, there are specific nuances, findings, and conclusions that can be made. Certainly, pro-
gress within each of the phenotypes covered (loosely grouped into the book sections) have been
uneven, probably because of the scientists involved and the resources within each area that have
been brought to bear on the topic. As such, we have provided an overall summary below of
each section of the book, to provide overall take-​home messages for each of the exercise/​sport
phenotypes that have been covered in this volume.

Summary –​Section 1: General systems genetics

The biggest take-​home messages of the first section of the book –​which dealt with a general
overview of current systems genetics knowledge –​begin with the thought that systems gen-
etics is rapidly advancing with new ideas, techniques, and concepts being added to the literature
almost daily. Systems genetics is a dynamic field and marrying it with exercise and sport science
should provide fertile research grounds for many years.While many exercise scientists may want

495
496

J.T. Lightfoot et al.

to ignore systems genetics of exercise and sport, it is incredibly important that we consider
systems genetics so that we can further understand how activity, exercise, and sport provide posi-
tive health outcomes. As Dr. Booth and colleagues noted in Chapter 1, understanding the “gen-
etic symphony” underpinning physiology is critical as we all continue to face health challenges
brought about by our improving quality of life and the concomitant reduction in physical
activity/​exercise in our daily lives. As such, translational efforts can serve to identify those who
are at higher risk of being inactive with a potential of providing differing exercise prescriptions
to those who are at risk. Further, there are already robust human modeling techniques that can
be used to study these types of genetic interventions to aid in the application of systems genetics
results to human physiology.
From a general systems genetics standpoint, the next 5 years should be exciting. The
application of CRISPR-​ Cas9 approaches to a wide variety of physiological questions
(4) promises to illuminate many pathways that may play major roles in exercise and sport
responses. However, the complexity of understanding how all the systems genetics pieces
“fit” together to influence a phenotype will remain a challenge to those who are interested
in determining how physiology is regulated by systems genetics mechanisms. Even the new
appreciation for the role of exosomes and their ability to deliver miRNA to sites other than
where they are produced promises to provide potential mechanisms controlling many of the
normal physiological responses. The professional interested in studying or applying systems
genetics needs to watch the literature closely for additional concepts that may apply to their
physiological interests.

Summary –​Section 2: Systems genetics of physical activity

Some would argue that the most critical health-​related issue currently facing humanity is the
increase in the number of chronic conditions and diseases caused by physical inactivity. The
major take-​home messages from the chapters relating to physical activity are that: 1) there is
a significant amount of literature –​well over 50 papers at this point –​that conclusively show,
in both humans and animal models, that systems genetics play a large role in regulating phys-
ical activity and physical inactivity; 2) physical activity and physical inactivity have genetic
underpinnings which may be different depending upon whether one is studying activity or
inactivity; and 3) the systems genetics influence on physical activity/​inactivity likely has both
peripheral (e.g., skeletal muscle) and central (e.g., brain) control components, with environ-
mental factors such as diet and exposure to toxicants altering the underlying systems genetics
control of activity/​inactivity. There can be little argument that, as noted in a recent consensus
paper (6), “choosing not to investigate these mechanisms would be irresponsible and would
hinder the science or understanding the causes of this critical health issue.”
The primary challenge facing investigators considering the systems genetics of physical
activity is the need for more hypothesis-​driven, causative studies. Correlational/​associative
studies have provided critical observations, including the suggestion of multiple genes and/​
or other genetic factors involved in regulating physical activity. However, there have been few
attempts to validate the suggested genetic factors and provide causative data supporting many
of these factors. Providing causative data is incredibly difficult and may be nearly impossible
in human studies; however, the number of solid animal models available provides an avenue
through which to provide initial causative data. In this same vein, there needs to be transla-
tional studies that include both humans and animal models. There have been minimal attempts
using this approach, with the existing data significantly flawed; however, this issue does not
predicate future, well-​designed, and properly interpreted translational studies. In fact, the lack

496
497

Afterword – closing the loop

of translational data available to bridge the voluminous animal data and the voluminous human
data is a critical gap in the literature.
Another challenge that has faced investigators in this area has been the relative lack of
funding. Scientific reviewers have been reluctant to accept that systems genetics plays a signifi-
cant role in the determination of physical activity/​inactivity levels; however, given the extensive
literature basis cited in Chapters 5–​10 as well as recent efforts to provide consensus statements in
the area (e.g., 6), we hope that scientists seeking funding in this area will be able to more easily
make a convincing case that these mechanisms are worth funding and investigation.
Within the next 5 years, we believe that several different avenues within the systems genetics
of physical activity/​inactivity will begin to be investigated. As mentioned previously, we antici-
pate that the application of CRISPR-​Cas9 systems will be used to supplement the minimal
amount of in vivo gene silencing that has already been done in the area to provide more causa-
tive data and better gene targets. These types of approaches will not only provide more causa-
tive data, but will help investigators further understand the role of newer concepts that have
not been investigated in the field such as the role of exosomes and epigenetics in determining
physical activity/​inactivity. Further, as noted in both Chapters 4 and 6, there are already initial
efforts to examine the genetics of affective mood and the role that these pathways might play in
physical activity/​inactivity.We anticipate that these efforts toward understanding the genetics of
affective mood will surely blossom over the next 5 years and the topic will deserve a chapter of
its own in future editions of this volume. Lastly, and maybe most critically, we believe that over
the next 5 years there will be efforts to understand why there has been variability in the genetic
results related to physical activity to date –​especially in the human genome wide association
study (GWAS) results –​and whether this variability is a result of multiple independent evolu-
tionary solutions for physical activity (10) or if there are other factors that need to be considered
in the research paradigms used.
In summary, it is clear that systems genetics play a role in regulating physical activity and
physical inactivity. Research efforts in these areas have thrived over the past 20 years and new
concepts and tools will only increase the progress in this area into the future.

Summary –​Section 3: Systems genetics of exercise endurance and

trainability
Exercise traits related to endurance are highly complex even at the organ level, integrating
organ systems such as lung, skeletal muscle, and the cardiovascular system. Much of Section 3 is
devoted to explaining endurance-​related traits from a multitude of perspectives, from a general
phenotype primer in Chapter 12 to more focused chapters on the anthropological evolution of
the endurance phenotype (Chapter 11) and underlying mitochondrial physiology and genetics
(Chapter 16). Two chapters explore heritability estimates of endurance-​related traits, splitting
the literature into animal-​(Chapter 13) and human-​based (Chapter 14) evidence, both pro-
viding high estimates of heritability of baseline endurance, with lower estimates of heritability
for response to training, indicating that contributions of environmental factors to variance in
training responses are substantial.
There is great interest in optimizing aerobic fitness across many different populations, as it is
so tightly linked to overall health, longevity, and athletic performance. At one end of the spec-
trum, there is always pressure on elite athletes to find any advantages over their competitors.
Recently, Kenya’s Eliud Kipchoge smashed the world record for the marathon, running 2:01:39
in the 2018 Berlin Marathon. A year earlier, Kipchoge had run 2:00:25 as part of Nike’s Breaking
2 project, running with several advantages impossible in a real marathon (such as staggered

497
498

J.T. Lightfoot et al.

pacing). After the Berlin marathon, there was much press about what factors limit human per-
formance capacity, such as lactate threshold,VO2max, and mechanical efficiency in a given task.
Obviously, genetic variation plays a role in endurance performance, but not in a simplistic way.
An example of a simplistic view would be the influence of a mutation in the angiotensin-​
converting enzyme (a key part of blood flow regulation) on endurance performance. Despite
over 1000 studies on the ACE I/​D variant, there is still no consensus about its effects on overall
endurance performance (summarized in Chapter 17). Most likely, there are thousands of genetic
variants affecting endurance components that sum to affect performance, many of which likely
interact among the key biological pathways related to endurance. Initial efforts to identify gene
variants specific to elite athletes have fallen short, likely due to lack of statistical power.
At the other end of the spectrum are people with low cardiorespiratory fitness due to aging,
lack of activity, or other reasons.This end of the spectrum is much more populated than the elite
athlete end, and diseases and disabilities related to low fitness drive most healthcare costs in many
developed countries. There is a great deal of interest in understanding endurance-​related traits
at the genetic and molecular levels to drive the generation of new therapeutics, such as exercise
“mimetics” that may help treat these inactivity-​related diseases and help reduce the incredible
healthcare costs associated with them. Though there is great debate about whether an exercise
mimetic could provide all the positive benefits of actual exercise (e.g., 2), therapeutics, such as
AICAR, that affect systems related to endurance exercise have been used to promote health
(8). It is hoped that with current efforts such as the National Institutes of Health Molecular
Transducers of Physical Activity Consortium (MoTrPAC; www.MoTrPAC.org), which was
described extensively by Dr. Bouchard in Chapter 35, we will better understand the molecular
and genetic determinants of endurance-​related traits and will be better able to develop and
deliver targeted therapeutics.

Summary –​Section 4: Systems genetics of muscle mass,

strength, and trainability
Section 4 explored exercise-​related traits generally associated with resistance (strength) training,
highlighting strength and size variability among people at baseline and following training.These
traits fall into the general pattern we see with many exercise-​related phenotypes, in that classic
heritability studies tell us that they should be strongly linked to genetic background. This con-
cept is summarized expertly by Dr. Martine Thomis in Chapter 18, stating that “twin and
family studies have indicated that genetic factors significantly (>0.50–​0.70) contribute to the
observed differences in muscle mass and strength characteristics.” Each of the authors in Section
4 were tasked with exploring both strength and size at baseline and these responses following
exercise training. Similar to the endurance-​related traits highlighted in Section 3, heritability
estimates are higher for baseline size and strength compared to responses to training.
At least from traditional SNP studies, associations between candidate gene variants and
strength or size traits sometimes overlap between baseline and post-​training changes, but often
do not, suggesting a partially unique genetic basis of training response. However, we have
mentioned many times throughout the book that there are many limitations within the trad-
itional SNP association studies (including lack of power, varied intervention parameters, lack of
ethnic diversity, etc.) that limit our ability to extrapolate their findings, meaning that we still do
not fully understand genetic versus environmental drivers of strength and size.
One thing that we can be sure of is that determinants of strength and size, whether at base-
line or following training, are much more complex than we once thought, as are the traits them-
selves (Chapters 19 and 20 provide primers for strength and size, in addition to an excellent

498
499

Afterword – closing the loop

review of neuroendocrine drivers of these traits in Chapter 21). Given a 50–​70% heritability for
strength and size, much work was devoted in the 2000s to identifying a few gene variants that
could explain large amounts of variation in each phenotype. A prime example of these efforts
are the findings regarding mutations in the myostatin gene, which provide dramatic changes
in muscle size and body composition in animal models. Dr. Dustin Hittel summarizes the
exploration of myostatin to date, highlighting the complexity of regulation of myostatin-​related
pathways (see Figure 22.1) and noting that common variants of the myostatin gene with large
effect sizes simply don’t exist in humans, though modification of myostatin’s effects pharma-
cologically could be of use in certain clinical populations. It is possible that newer efforts like
MoTrPAC will identify new gene variants associated with muscle strength and size, but efforts
to date have explained only a fraction of the variation present among humans. MoTrPAC and
other efforts are emphasizing multiscale omics approaches that should shed more light on the
molecular underpinnings of strength and muscle size.

Summary –​Section 5: Systems genetics of sports performance

In the bulk of Section 5, we see the challenges of expanding our genetic analysis to the level of
“performance” rather than specific underlying traits. Performance is arguably the hardest trait
to measure and study, because it represents the contributions of many underlying traits that
combine in specific ways for a specific outcome in a specific activity. What equates as “maximal
performance” in sprinting varies from soccer, which varies from cricket, which varies from
wrestling, and on and on. Moreover, measuring elite-​level performance has the added challenge
that elite performers, by definition, are rare and thus unique, making research studies difficult
to pursue and far-​reaching conclusions difficult to come by. Researchers are addressing these
questions through consortia and other means, but there is a clear conundrum facing researchers
focused on the systems genetics of exercise and sport: how to address the remarkable complexity
of performance while necessarily focusing on the “narrower” analyses of specific traits, be it grip
strength, stroke volume, or lactate threshold. Those detailed studies are needed, but they may
not enhance the broader understanding of performance as they are done independently of the
broader gene networks and other variables impacting the more complex performance traits.
While the studies to date clearly support a genetic contribution to performance, in par-
ticular from the twin and family studies outlined, we’ve made little progress in identifying the
underlying systems genetics factors involved. More complex GWAS and direct DNA sequen-
cing approaches are beginning to emerge and will require replication and extension. The many
environmental contributors to performance will necessarily complicate this task, but that makes
replication even more important.What remains to be seen is whether, as outlined in the conclu-
sion of Chapter 26, E.O.Wilson was right those many years ago when he effectively argued that
there are so many possible combinations of exceptional genes that predicting their appearance
in an individual is nearly impossible. And the corollary to that thought is knowing whether
there are so many unique combinations that looking for similarities across even hundreds of
elite athletes within the same sport may ultimately yield hundreds of specific, unique, genetic
patterns, all of which predispose to elite performance and which may be reflective of the “mul-
tiple independent evolutionary solutions” hypothesis mentioned earlier.
In considering the chapters across Sections 2–​5, the field appears to be caught in the middle
space between the single gene studies of the 2000s and the more advanced genome-​wide omics
techniques of present day, with an inherent desire to identify specific genes while knowing we
are pursuing variation in extensive genomic networks instead. Those latter studies are remark-
ably complex, requiring the combined power of multiple omics analyses within a carefully

499
500

J.T. Lightfoot et al.

studied (and sufficiently large) cohort. While these studies are emerging in a variety of fields
(e.g., cancer and diabetes), our ability to successfully pursue systems genetics studies in exer-
cise and sport is particularly challenged by the nature of our complex exercise interventions,
differences in measurement techniques, and inherently small sample sizes.That funding agencies
in some countries are less inclined to support sport-​related research is an added barrier to the
establishment of large studies and international consortia capable of addressing these complex
issues. But these constraints are being addressed in a variety of ways and the number of consortia
to emerge in the past few years is remarkable. There is clearly optimism that the emerging tools
in genomic analysis coming primarily from disease-​oriented fields will be equally beneficial to
the field of exercise and sport.
Although slightly unique in context from the other chapters, Chapters 27 and 28 are notable
for their implications for our field to better identify individuals predisposed to sport-​related
injury, as well as improve prevention and treatment approaches. That genetic predisposition
exists in both brain and soft tissue injuries is clear, though the details (as with all other traits in
the field) remain to be clarified. And yet these findings yield additional complexities for sport;
while genetic testing and genetic technologies to improve performance are generally frowned
upon (as shown in Section 6), their potential use for injury prevention and recovery often result
in an opposite perspective. Navigating those different viewpoints won’t be easy, as one person’s
injury recovery is another’s performance enhancement.
One area that is clearly lacking in this volume is an examination of epigenetics in exercise
and sport. While we recognized the importance of the topic, we felt the state of the literature
was too underdeveloped to warrant a dedicated chapter at the time we designed the book.That
will surely change over the next few years and the findings may illuminate unique aspects of
systems genetics in exercise that will help dissect the unique interaction of genes and environ-
ment in sport, perhaps with implications for heritability.

Summary –​Section 6: The ethics of systems genetics in exercise and sport

While the previous sections have focused on the discovery of new insights into the systems gen-
etics of exercise and sport sciences, Section 6 of the book takes a much different approach and
examines some of the ethical considerations stemming from those research findings. Despite
scientists’ best intentions (or lack of interest), their findings in the field of genetics have the
potential to impact broad aspects of society and human culture, perhaps even humanity itself.
So, it is important that the readers of this volume who may otherwise simply be interested in
learning about the latest findings and future directions of scientific inquiry in their particular
area of focus should at least be aware of these broader implications.
Genetic testing of athletes (and potential athletes) emerged as a controversial topic in the
early days of the field, as soon as the testing technology became less expensive and Internet-​based
companies became more common. Despite the science of the field being less than conclusive
(as we’ve seen from the previous sections of the book), companies are generously employing
the existing science and marketing testing products of questionable scientific validity. Because
sports performance traits typically fall outside the stringent oversight requirements of testing
for health-​related traits, the companies have very little restriction on their ability to produce
and market such tests.While savvy consumers may have access to much of the same literature as
scientists, few are likely to delve deeply enough into that literature to understand the limitations
and consequences of genetic testing for sport performance and will instead rely on the com-
panies carefully selected (and limited) reference lists to judge the validity of a testing product.
We would do well as a field to consider the implications of our findings around specific genes

500
501

Afterword – closing the loop

and specifically mention the limitations of the use of such information for genetic testing in
our published papers. In this way, companies bent on marketing products with limited scientific
evidence will at least have to ignore our explicit cautions about such uses.
The authors in Section 6 also challenge us in several ways to consider the broad implications
of our work on society more generally. Consider the implications of studying genetic aspects
of performance in specific race groups. While such work is a natural extension of the basic
question,“Why is this group disproportionately represented on the leaderboard?”, the pursuit of
the research may inadvertently support the notion of biological determinism and racial super-
iority. Any carefully controlled research study limits the number of variables examined to better
ensure clear, valid conclusions; examination of a question as broad as that posed above would
require inclusion of social, cultural, economic, and other factors, which would clearly over-
whelm a typical research team and muddy an otherwise straightforward hypothesis and research
study. But we would be wise to tread carefully into such domains without clearly identifying for
ourselves and our audience the limits and limitations of our research.
Similarly, examination of sex and performance is fraught with challenge. As shown in Section
5’s Chapter 29, sex differences appear in a variety of sports and appear linked to biological and
physiological differences. Defining those differences and ensuring equity in the sports arena,
however, is remarkably difficult. Defining the clear biological underpinnings of sex seems as if it
should be straightforward, but the varied attempts over the years at finding such a definition (for
determining an equal playing field) have only failure in common. Scholars in biological fields
recognize the importance of “exceptions” to every rule, and the varied nature of human biology
becomes more evident with every attempt at categorizing “men” and “women.” Genetic ana-
lysis will surely shed light on this variation in our species, but will simultaneously wreak havoc
on our ability to categorize people for the purposes of parity and fairness in sport competitions.
Perhaps most important of all the challenges raised by the authors of Section 6 are the
implications that genetic engineering has for human biology and society. Mind you, the poten-
tial for genetic engineering to alter humanity is not specifically tied to the world of sport, as
much of the work and technical innovation are being driven by the study of disease treatment
and elimination (e.g., CRISPR-​Cas9, as outlined above). Nevertheless, athletes’ willingness to
pursue such technologies for performance enhancement makes the implications evident to a
broader swath of society and enables both different questions and a wider pool of discussants.
And, as Chapters 32 and 33 demonstrate, we have a long way to go before reaching any semb-
lance of consensus about the role of genetic engineering in sport, let alone society more broadly.
Do we allow somatic-​cell genetic engineering, but outlaw gamete manipulation to prevent
“permanent” changes to the human genome? What forms of genetic therapy should be allowed
for athletes, or are such athletes destined to be categorized separately? Where such manipu-
lation can even be measured in an athlete, how much testing should we require (e.g., regular
biopsies)? Where might genetic technologies be allowed for injury treatment? Many questions
remain to be considered and addressed as genetic manipulation becomes more commonplace
in our society.

Final conclusions
This volume represents an exciting look into one of the more dynamic research areas in exer-
cise and sport sciences. Within this volume we have worked to show where the field has been,
where it is currently, and maybe more importantly, where we think the field is going in the near
future. We hope that with this book as a spring-​point, you will choose to continue to journey
with us and the investigators in this book, as we all continue to work to understand the systems

501
502

J.T. Lightfoot et al.

genetics of exercise and sport. As we close this volume, we again want to thank all the authors
who have so kindly contributed chapters to this book; simply put, without their contributions,
this volume would not exist.

References
1. Amoasii L, Hildyard JCW, Li H, Sanchez-​Ortiz E, Mireault A, Caballero D, Harron R,
Stathopoulou TR, Massey C, Shelton JM, Bassel-​Duby R, Piercy RJ, and Olson EN. Gene
editing restores dystrophin expression in a canine model of Duchenne muscular dystrophy. Science
362: 86–​91, 2018.
2. Booth FW and Laye MJ. Lack of adequate appreciation of physical exercise’s complexities can pre-​
empt appropriate design and interpretation in scientific discovery. J Physiol 587: 5527–​5539, 2009.
3. Bouchard C and Rankinen T. Individual differences in response to regular physical activity. Med Sci
Sports Exerc 33: S446–​S451, 2001.
4. Doudna JA and Charpentier E. Genome editing. The new frontier of genome engineering with
CRISPR-​Cas9. Science 346: 1258096, 2014.
5. Joyner MJ and Lundby C. Concepts about VO2max and trainability are context dependent. Exerc
Sport Sci Rev 46: 138–​143, 2018.
6. Lightfoot JT, De Geus EJC, Booth FW, Bray MS, den Hoed M, Kaprio J, Kelly SA, Pomp
D, Saul MC, Thomis MA, Garland T, and Bouchard C. Biological/​genetic regulation of physical
activity level: consensus from GenBioPAC. Med Sci Sports Exerc 50: 863–​873, 2018.
7. Lundby C, Montero D, and Joyner M. Biology of VO2 max: looking under the physiology lamp.
Acta Physiol (Oxf) 220: 218–​228, 2017.
8. Narkar VA, Downes M, Yu RT, Embler E, Wang YX, Banayo E, Mihaylova MM, Nelson MC,
Zou Y, Juguilon H, Kang H, Shaw RJ, and Evans RM. AMPK and PPARdelta agonists are exer-
cise mimetics. Cell 134: 405–​415, 2008.
9. Tang L, Zeng Y, Du H, Gong M, Peng J, Zhang B, Lei M, Zhao F, Wang W, Li X, and Liu
J. CRISPR/​Cas9-​mediated gene editing in human zygotes using Cas9 protein. Mol Genet Genomics
292: 525–​533, 2017.
10. Xu S and Garland T. A mixed model approach to genome-​wide association studies for selection
signatures, with application to mice bred for voluntary exercise behavior. Genetics 207: 785–​799, 2017.

502
503

INDEX

Note: Page numbers in italics refer to figures, those in bold refer to tables.

ACE see angiotensin-converting enzyme gene muscle metabolism regulation 334, 335; muscle
ACE model, twin studies 33–4 strength and performance 323–40; muscle
ACTN3 R577X see alpha-actinin-3 strength and performance, influences 330–7;
additive genetic variance, definition 29 recent positive selection 324–5; sarcomeric
adenosine monophosphate kinase (AMPK), alpha-actinins regulate calcineurin signaling
mitochondria 203 333–4; sprint/power elite athlete association
adherence, exercise see exercise adherence studies 326–7; therapeutic implications of
adoption studies, human systems genetic modeling ACTN3 knockout mouse findings for ACTN3
36–7 577XX patients 339–40; twin studies 376;
aging: cardiac function 190; elite athletes 360–1 Z-disk proteins are upregulated in Actn3
allele, definition 29 knockout mice 329; Z-disk stability 331–3
alpha-actinin-3 323–40; absence of, effect on elite altitude-induced erythropoietin (EPO),
sprint/power performance 325–8; ACTN3 pulmonary function 192
genotype and inherited muscle diseases 338–9; AMPK (5ʹ AMP-activated protein kinase): AMPK
ACTN3 genotype and the response to exercise pathway, endurance training adaptations 154;
training 330, 331–2; ACTN3 genotype and signaling, endurance training adaptations 155–7
training association studies 331–2; ACTN3 AMPK (adenosine monophosphate kinase),
knockout mouse model 328–30; as a modifier mitochondria 203
of Duchenne muscular dystrophy 338–9; androgens 291–6; androgen biosynthesis and
ACTN3 R577X polymorphism is common skeletal muscle steroidogenesis 293; androgen
in humans 325; alpha-actinin-3 deficiency receptor adaptations to training 296; androgen
“slows” fast-twitch muscle fiber characteristics receptor genetic variations 293–5; androgen
328–9; calcineurin signaling 333–4; calcium receptor structure 295–6; androgen signaling
handling 334–5; candidate gene approach 293; neuroendocrine responses to RET, genetic
and linkage analyses 350; Duchenne muscular contributions 291–6; testosterone 291–2
dystrophy 338–9; frequency of the ACTN3 angiotensin-converting enzyme gene (ACE)
genotype in controls and elite sprint and 216–45; ACE rs4340, endurance exercise
endurance athletes 327; genetic control of performance, and fluid electrolyte balance
muscle strength and performance 323–40; 236; ACE rs4340, endurance performance,
genetic testing 444; glycogen phosphorylase and cardiac function 223–36; ACE rs4340,
activity 334, 335; inherited muscle diseases endurance performance, and cardiorespiratory
338; linkage analyses and the candidate gene fitness as assessed by maximal or peak oxygen
approach 350; mechanistic insights from the consumption 223; ACE rs4340, endurance
ACTN3 KO mouse 337; muscle aging 337–40; performance, and muscle metabolism 236;
muscle disease 337–40; muscle disuse 337–40; ACE rs4340, plasma ACE concentration/

503
504

Index

activity, and endurance performance 222–3; Bouchard, Claude 7

ACE rs4340 and endurance exercise health- BPA see bisphenol A
related outcomes 237–44; ACE rs4340 and
endurance performance at high altitude 237; cachexia, myostatin (MSTN) 318–20
ACE rs4340 and other measures of endurance CAD see coronary artery disease
performance 223; ACE rs4340 and the blood CAIS (complete androgen insensitivity syndrome),
pressure response to endurance exercise sex testing in sport 478
237–44; ACE rs4340 and the response of other calcineurin, cardiac function 190
endurance exercise health-related outcomes calcineurin signaling, sarcomeric alpha-actinins
244; cardiac function 190; influence of ACE 333–4
on endurance exercise performance 222–37; calcium handling, alpha-actinin-3 334–5
location 216; physical performance number caloric restriction, influencing the genetic
of citations per year in PubMed 217; renin– regulation of activity 123–4
angiotensin–aldosterone (RAAS) system 220–2, calsequestrin 1 (Casq1) peripheral mechanisms
245; systematic review methods 217–19; trial regulating activity 111, 112–13
selection process 217–19; trials that examined CaMKII (calcium/calmodulin-dependent
the influence of ACE rs4340 on endurance protein kinase II) enzyme, endurance training
exercise health-related outcomes 238–43; trials adaptations 157, 158
that examined the influence of ACE rs4340 on CaMK pathway, endurance training adaptations
endurance performance 224–35 154
animal models: genetic regulation of physical cancer-associated cachexia, myostatin (MSTN)
activity 55–64; physical activity 55–6 318–19
annexin family of proteins, peripheral mechanisms candidate genes: alpha-actinin-3, candidate gene
regulating activity 110–11, 112–13 approach and linkage analyses 350; cardiac
apes and humans: comparative biology 137–8; function 189–90; hypoxia adaptations 194;
physical activity 137–8 physical activity 112–13; pulmonary function
apolipoprotein E (APOE) 46, 143; concussion 194; quantitative trait loci (QTL) 189–90; soft
389–94; genotype 46 tissue injuries 402–408
association mapping: genetic mapping 167–9; capability to be active, peripheral mechanisms
heritability of endurance traits 167–9 regulating activity 108–112
association studies, Human Genome Project cardiac function: ACE rs4340: 223–36; aging
18–19 190; angiotensin-converting enzyme gene
assortative mating, definition 29 (ACE) 190; calcineurin 190; candidate genes
athletes’ exposome, family studies 383–4 189–90; heme oxygenase 1 (HMOX1) 190;
Athlome Project Consortium: elite athletes 363, left ventricular mass 190; quantitative trait loci
366, 368; genome-wide association studies (QTL) 189–90; ventricular hypertrophy 190
(GWAS) 351; sport performance 351 cardiorespiratory fitness (CRF) 187–95; ACE
atrophy, myostatin (MSTN) 318–19 rs4340: 223; cardiac function 188–90; definition
187; definitions 194; high-altitude versus
BBP see benzyl butyl phthalate sea-level dwellers 193–4; hypoxic ventilator
BDNF signaling pathway 46–7 response (HVR) 193; pulmonary function
“behavior evolves first”, motivation for physical 191–4; respiratory function 191
activity 94–5 cardiorespiratory responses, endurance exercise 149
benzyl butyl phthalate (BBP), endocrine- Care Consortium, concussion 398
disrupting chemicals (EDCs) 119, 120, 121 Casq1 (calsequestrin 1) peripheral mechanisms
bisphenol A (BPA), endocrine-disrupting regulating activity 111
chemicals (EDCs) 119, 121 CHARGE consortium 38
black athletes 433–40 Cholesky decomposition multivariate model, twin
blood flow redistribution, endurance exercise studies 34–6
149–50 chromatin: genetic regulatory mechanisms 101;
blood pressure response to endurance exercise, neurogenetics of motivation for physical activity
ACE rs4340: 237–44 98, 101; sex testing in sport 477–8
blood volume perseveration, endurance exercise chromosome formula, sex testing in sport 477–8
150 chronic obstructive pulmonary disease (COPD),
BMI see body mass index myostatin (MSTN) 318–19
body composition, sex and performance 425–6 ciliary neurotrophic factor (CNTF), muscle
body mass index (BMI) 36, 224–34, 238–42, 489 strength cellular pathways 272

504
505

Index

clustered regularly interspaced short palindromic dropout, exercise see exercise dropout factors;
repeats (CRISPR): Cas9 systems 3; future exercise dropout risk
research 495, 496, 497; gene doping 456–7 Duchenne muscular dystrophy, alpha-actinin-3
CNTF (ciliary neurotrophic factor), muscle 338–9
strength cellular pathways 272 DZ see dizygotic twinning
Cobb, W. Montague, race and sports
performance 434 East Germany, gene doping 455–6
collagen genes, soft tissue injuries 404, 407 ecological pressures, endurance phenotype 139–40
Colombian sports families, family studies 377–83 economic costs, physical inactivity 12–13
comparative approach, endurance phenotype EDCs see endocrine-disrupting chemicals
137–8 Edwards, Harry, race and sports performance
comparative biology: apes and humans 137–8; 435–7, 438–9
physical activity 137–8 electrolyte balance, ACE rs4340: 236
complete androgen insensitivity syndrome (CAIS), elite athletes 357–69 see also sport performance;
sex testing in sport 478 aging 360–1; Athlome Project Consortium 363,
compliance, exercise see exercise adherence 366, 368; definitions 357–60; erythrocytosis
concussion 388–98 see also soft tissue injuries; 360–1; erythropoietin receptor gene (EPOR)
apolipoprotein E (APOE) 389–94; Care 360–1; Exercise at the Limit – Inherited
Consortium 398; concussion genetic Traits of Endurance (ELITE) 363–7; family
associations 397–8; future research 396–8; studies 374–5; future research 367–9; GAMES
glutamate receptor ionotropic NMDA Consortium 366, 367; GenAthlete study 363,
receptor 2A (GRIN2A) 395–6; neurofilament 375; GENESIS (Genetics of Elite Status in
392, 395; pathomechanical genetic associations Sport) 367; GeneSMART (genes and the
389–95; pathomechanics 388–9, 390; Skeletal Muscle Adaptive Response to Training)
pathophysiological genetic associations 367; global efforts to investigate elite athletes
395–6; pathophysiology 388–9, 390; proteins/ 362–7; lifestyle disease 360–1; major study
genes 397–8; tau 392, 394; vesicle glutamate cohorts 362–7, 364–6; major study consortia
transporters (VGLUT) 396 362–7, 366; as a model for research 357–69;
COPD (chronic obstructive pulmonary disease), modeling athletic ability, health and disease
myostatin (MSTN) 318–19 360; pluses and minuses of working with elite
coronary artery disease (CAD) 20 athletes 361–2; POWERGENE Consortium
cost of living, future 12–13 366, 367, 368; super athletes 360–1; thrifty
CRF see cardiorespiratory fitness genotype 360–1; twin studies 374–5
CRISPR see clustered regularly interspaced short Encyclopedia of DNA Elements Consortium
palindromic repeats (ENCODE) project 488
endocrine-disrupting chemicals (EDCs): benzyl
definitions: additive genetic variance 29; allele 29; butyl phthalate (BBP) 119, 120, 121; bisphenol
assortative mating 29; cardiorespiratory fitness A (BPA) 119, 121; influencing the genetic
(CRF) 187, 194; elite athletes 357–60; endurance regulation of activity 119–22; sex hormones
phenotype 135–7; epistatic genetic variance 29; 119–22; vinclozolin 119–22
gene doping 463–5; genetic variance due to endurance exercise: myostatin (MSTN) 315–17;
dominance 29; human systems genetic modeling physiologic responses 149–50
29; shared environmental variance 29; stress 8; endurance phenotype 135–44, 148–59;
unshared environmental exposure 29 comparative approach 137–8; definition 135–7;
deliberate practice, family studies 383–4 ecological pressures 139–40; evolution 135–44;
diabetes: type 2 diabetes, epigenetics 11; type evolutionary considerations 148–9; evolutionary
2 diabetes, myostatin (MSTN) 317; type 2 exercise physiology 143; fossil record analyses
diabetes, physical inactivity 11 138–43; genetic regulation of physical activity
diet: caloric restriction 123–4; influencing the 144; lifestyle changes 139–40; metabolic
genetic regulation of activity 122–8; overfeeding equivalents (METs) 135–6, 138; morphological
124–8 approach 138–43; physiologic responses to
dizygotic (DZ) twinning see also monozygotic endurance exercise 149–50; skeletal adaptations
(MZ) twinning: twin studies 31–6 140–2
dominance, twin studies 32–3 endurance training: adaptations 151–9; molecular
dopamine, neurogenetics of motivation for pathways 152–9; myocardium responsiveness
physical activity 97–8 150–2; skeletal muscle responsiveness 150–2;
doping, gene see gene doping whole-body adaptations 151–2

505
506

Index

eNOS and nitric oxide signaling, endurance exercise performance, heritability 347–54
training adaptations 159 exercise prescription and intervention: enhancing
Entine, Jon, race and sports performance 438–40 46–7; genetic information 46–7
environmental factors influencing the genetic exercise systems genetics: genes identification 8–9,
regulation of activity 117–29; diet 122–8; 44–5; vs. inactivity systems genetics 8–9
endocrine-disrupting chemicals (EDCs) 119– exercise systems genomics 7–8
22; mechanisms 117–19; sex hormones 117–19 exosomes, Human Genome Project 20–1
epigenetics: exercise effects 21; Human Genome
Project 21; physical inactivity 11; type 2 FAK (focal adhesion kinase), upstream signaling to
diabetes 11 mTORC1 in response to RET 281
epistatic genetic variance, definition 29 family studies 29–31, 372–84, 487, 490 see
erythrocytosis, elite athletes 360–1 also twin studies; athletes’ exposome 383–4;
erythropoietin receptor gene (EPOR): elite Colombian sports families 377–83; deliberate
athletes 360–1; erythrocytosis 360–1 practice 383–4; elite athletes 374–5; genetic-by-
ethical challenges, predicting sport performance exposome interactions 383–4; genetic factors
442–50 and sports participation 374; heritability of
ethical perspectives, gene doping 453–61 endurance traits (human research models) 184;
ethics, sex testing in sport 475–82 history 372–4; pedigree of successful sports
evaluating voluntary physical activity 107–8 family 377; quantifying family data 30–1; sport
evolution: ecological pressures 139–40; endurance performance 374–5; structure and function of
phenotype 135–44; gene doping enhancing families 29–30
evolution 463–73; genetic regulation of FAMuSS (Functional Single Nucleotide
physical activity 144; lifestyle changes 139–40; Polymorphisms Associated with Human Muscle
phylogeny of hominidae 136–7, 136; skeletal Size and Strength) study, muscle strength
adaptations 140–2 development 266–7
evolutionary exercise physiology: endurance Finnish Twin Study on Aging (FITSA),
phenotype 143; human health 143 multivariate heritability studies for muscle mass
evolution of behavior, natural reward circuit 96–7 and strength 258
evolution of genetic mechanisms controlling fluid electrolyte balance, ACE rs4340: 236
physical activity 80–90; age of associated focal adhesion kinase (FAK), upstream signaling to
alleles 86–9; energy expenditure as a selection mTORC1 in response to RET 281
pressure 83; exercise duration as a potential fossil record analyses, endurance phenotype
selection pressure 84–5; food accessibility as a 138–43
selection pressure 83–4; humans cf. mice 81–2, FTO gene 38, 45, 47–8
83; processes and factors summary 89–90; Functional Single Nucleotide Polymorphisms
quantitative trait loci (QTL) 82; selection Associated with Human Muscle Size and
pressures, potential 82–5; single nucleotide Strength (FAMuSS) study, muscle strength
polymorphisms (SNPs) 86–9 development 266–7
exercise see also endurance training; physical future 10–13; cost of living 12–13; economic costs
activity; physical inactivity: benefits 95; of physical inactivity 12–13; era of molecular
physiologic responses to endurance exercise genetics 38; inheritance of low voluntary
149–50; population statistics 41; stress 8 running behavior 11; low voluntary running
exercise adherence 43–5 see also exercise dropout animal model to mimic increased human
factors; enhancing 43–5; genetic information inactivity 11–12; technology and physical
42–5 inactivity 10–11
Exercise at the Limit – Inherited Traits of future research 490–2; clustered regularly
Endurance (ELITE), elite athletes 363–7 interspaced short palindromic repeats
exercise dropout factors see also exercise adherence: (CRISPR) 495, 496, 497; concussion 396–8;
exercise interventions 42, 43 elite athletes 367–9; genetic architecture
exercise dropout risk, genetic information 42–5, 62–4; heritability 62–4; mitochondria 208–9;
47–8 Molecular Transducers of Physical Activity
exercise effects: epigenetics 21; lifestyle Consortium (MoTrPAC) 8–9, 490; muscle
comparisons 11; reference group 9–10; weight- strength genetic contributions 272–3; soft tissue
bearing exercise 9–10 injuries 411; sport performance 351–2
exercise genomics, translating into practice 42–3
exercise interventions: exercise dropout factors 42, GAMES Consortium, elite athletes 366, 367
43; theoretical foundations 41–2 GenAthlete study, elite athletes 363, 375

506
507

Index

gene doping 453–61; actions needed now genome-wide association studies (GWAS) 74–6;
470–3; classification 470–1; clustered regularly human studies 67–76; linkage 74; twin studies
interspaced short palindromic repeats 67–73; UK Biobank 74, 75–6
(CRISPR) 456–7; definitions 463–5; East genetic regulatory mechanisms: chromatin 101;
Germany 455–6; enhancing evolution 463–73; neurogenetics of motivation for physical
ethical perspectives 453–61; gene hacking 456– activity 101
7; health and 470–3; healthcare purpose 467–9; genetics: hypoxia 192–3; pulmonary function
health protection and longevity 460–1; Human 192–3
Genome Project 467–8; Messi, Lionel 457–60; genetic screening, talent identification 445–9
Miah, Andy 453–4, 457; Munthe, Christian 454; genetic studies in sports science, timeline 348
“spirit of sport” 466–7; terminology 463–5; genetic testing, talent identification 444–50;
Therapeutic Use Exemption (TUE) 458–9; alpha-actinin-3 444; genetic screening 445–9;
transhuman case for gene doping 463–73; potential 444–5; potential benefits 445–7;
transhumanism 469–70; types 470–1; World potential drawbacks 447–50
Anti-Doping Agency (WADA) 454–60, 461, genetic variance due to dominance, definition 29
463–7 genetic variations associated with muscular
gene-finding studies 73–6; genetic regulation of strength – methodological approaches 267–8
physical activity 73–6; genome-wide association genome-wide association studies (GWAS) 18–20,
studies (GWAS) 74–6; linkage 74 489; Athlome Project Consortium 351; gene-
gene hacking, gene doping 456–7 finding studies 74–6; heritability of endurance
genes identification, exercise systems genetics 8–9, traits 173; human systems genetic modeling
44–5 37–8; single nucleotide polymorphisms (SNPs)
gene silencing,Vivo-Morpholinos 112–13 75; sport performance 350–1
GENESIS (Genetics of Elite Status in Sport), elite genome-wide linkage studies, single nucleotide
athletes 367 polymorphisms (SNPs) 74
GeneSMART (genes and the Skeletal Muscle genomic architecture of complex traits 489–90
Adaptive Response to Training), elite Genotype-Tissue Expression (GTEx) project 488
athletes 367 glucocorticoids: acute responses and chronic
genetic architecture 55–6, 58–64; fine-mapping adaptations of glucocorticoids to RET 303–4;
approaches 61–2; future research 62–4; genetic glucocorticoid receptor 304–5; glucocorticoid
mapping 58–62 receptor gene (NR3C1) 304–5; glucocorticoid
genetic-by-exposome interactions, family studies signaling 304; neuroendocrine responses to
383–4 RET, genetic contributions 302–5
genetic contributions: mitochondrial traits 200–9; glucose transport 4 (Glut4), substrate utilization
muscle mass/strength 282; muscle size 277–83; 109–10
muscle strength 264–73; neuroendocrine glutamate receptor ionotropic NMDA receptor
responses to RET 290–305; skeletal muscle size 2A (GRIN2A), concussion 395–6
277–83 glycogen phosphorylase activity, alpha-actinin-3
genetic control of muscle strength and 334–5, 335
performance, alpha-actinin-3 323–40 glycolysis enzymes, peripheral mechanisms
genetic factors and sports participation 374 regulating activity 112
genetic factors role, muscle mass and strength GRIN2A (glutamate receptor ionotropic NMDA
responses to exercise 259–60 receptor 2A), concussion 395–6
genetic information: exercise adherence 42–5; growth factors, regulation of muscle size and
exercise dropout risk 42–5, 47–8; exercise strength by 268–70; insulin-like growth factor 1
prescription and intervention 46–7; The (IGF1) 268–70; myostatin (MSTN) 269, 270
Genetic Information Nondiscrimination Act 17; growth hormones: basic growth hormone
predicting sport performance 442–50 cybernetics 297; growth hormone (GH)
The Genetic Information Nondiscrimination therapy, Messi, Lionel 457–60; growth hormone
Act 17 in circulation 297–9; neuroendocrine responses
genetic mapping: association mapping 167–9; to RET, genetic contributions 296–9; resistance
genetic architecture 58–62; heritability of exercise responses 299, 300; somatotroph types
endurance traits 166–9; intrinsic endurance and growth hormone isoforms 297, 298
exercise capacity 166–9; linkage analysis 166–7 GWAS see genome-wide association studies
genetic regulation of physical activity: animal
models 55–64; endurance phenotype 144; handgrip strength, muscle strength heritability 256
evolution 144; gene-finding studies 73–6; health, gene doping and 470–3

507
508

Index

healthcare purpose: gene doping 467–9; Human human health, evolutionary exercise
Genome Project 467–9 physiology 143
health-related outcomes, ACE rs4340: 237–44 human studies see also heritability of endurance
heart rate physical activity score (HRPAS) 45 traits (human research models): genetic
heme oxygenase 1 (HMOX1), cardiac regulation of physical activity 67–76
function 190 human systems genetic modeling 28–38; adoption
heritability: exercise performance 347–54; future studies 36–7; definitions 29; family studies
research 62–4; meaning of 349; muscle mass 29–31; future in the era of molecular genetics
253–5; muscle mass and strength responses to 38; genome-wide association studies (GWAS)
exercise 259–60; muscle size and strength traits 37–8; purpose 28; twin family designs 37; twin
253–60; physical activity 56–8; pulmonary studies 31–6
function 191–2 human variation in endurance training responses,
heritability of endurance traits (animal research heritability of endurance traits (human research
models) 164–73; association mapping 167–9; models) 182–3
complex nature of the phenotype 173; genetic HVR (hypoxic ventilator response),
mapping 166–9; genome-wide association cardiorespiratory fitness (CRF) 193
studies (GWAS) 173; inbred strain comparisons hyperandrogenism, sex testing in sport 480–1
166, 169–70; intrinsic endurance exercise hypomethylation of DNA, myostatin (MSTN)
capacity 165–9; linkage analysis 166–7; mouse– 315–16
rat–human comparative genomics 171–2; hypoxia: genetics 192–3; pulmonary function
quantitative trait loci (QTL) 164–73; responses 192–3
to exercise training 169–71; selected strains/ hypoxia adaptations: candidate genes 194;
artificial selection 165–6, 169 pulmonary function 194
heritability of endurance traits (human research hypoxia-inducible factor 1α (HIF-1α), pulmonary
models) 178–85; family studies 184; heritability function 192
estimates for endurance-related traits in the hypoxic ventilator response (HVR),
intrinsic/untrained state 180; heritability of the cardiorespiratory fitness (CRF) 193
response of endurance traits to exercise training
183–4; human variation in endurance training IGF1 (insulin-like growth factor 1): acute
responses 182–3; intrinsic levels of endurance responses and chronic adaptations of IGFs to
traits 178–81; meta-analyses of heritability RET 301–2; IGF-1 receptor and intracellular
estimates 180, 181; response of endurance traits signaling 302, 303; IGF and IGFBP genes 300–
to exercise training 183–4; twin studies 183 01; isoforms 300–01; neuroendocrine responses
HIF-1α (hypoxia-inducible factor 1α), pulmonary to RET, genetic contributions 299–302;
function 192 regulation of muscle size and strength 268–70
high-altitude endurance performance, ACE IL-6 (interleukin-6), muscle strength cellular
rs4340 237 pathways 271–2
high-altitude versus sea-level dwellers, IL-15 (interleukin-15), muscle strength cellular
cardiorespiratory fitness (CRF) 193–4 pathways 272
hippocampus, neurogenetics of motivation for IL-15Rα 1775AA genotype, peripheral
physical activity 100 mechanisms regulating activity 110
homeostasis 8 imaging techniques, muscle mass heritability 254
hormones see also growth hormones; sex inactivity systems genetics, vs. exercise systems
hormones; sex testing in sport; testosterone: sex genetics 8–9
and performance 425–6 inbred strain comparisons, intrinsic endurance
HRPAS see heart rate physical activity score exercise capacity 166
human cell initiative 488 inflammatory factors, regulation of muscle
human genome in numbers 488–9 strength and size by 271–2; ciliary neurotrophic
Human Genome Project 16–22; association factor (CNTF) 272; interleukin-6 (IL-6) 271–2;
studies 18–19; epigenetics 21; exosomes interleukin-15 (IL-15) 272; tumor necrosis
20–1; gene doping 467–8; genome-wide factor-alpha (TNFα) 271
association studies (GWAS) 18–20; healthcare inheritance: low voluntary running behavior 11;
purpose 467–9; microRNAs 20; minION Pima Indians 11
sequencer 17; noncoding RNA and antisense inherited muscle diseases, alpha-actinin-3 338
sequences 20; noncoding sequences 20–1; insulin-like growth factor 1 (IGF1): acute
sequencing advances 17–18; single nucleotide responses and chronic adaptations of IGFs to
polymorphisms (SNPs) 18–20 RET 301–2; IGF-1 receptor and intracellular

508
509

Index

signaling 302, 303; IGF and IGFBP genes 300–1; Messi, Lionel: gene doping 457–60; growth
isoforms 300–1; neuroendocrine responses to hormone (GH) therapy 457–60
RET, genetic contributions 299–302; regulation meta-analyses of heritability estimates, heritability
of muscle size and strength 268–70 of endurance traits (human research models)
insulin resistance, myostatin (MSTN) 317–18 180, 181
interleukin (IL)-6 7–8 metabolic equivalents (METs), endurance
interleukin-6 (IL-6), muscle strength cellular phenotype 135–6, 138
pathways 271–2 metabolic flexibility, endurance exercise 149, 150
interleukin-15 (IL-15), muscle strength cellular Miah, Andy, gene doping 453–4, 457
pathways 272 microRNAs, Human Genome Project 20
intramuscular sensing and signaling regulating minION sequencer 17
resistance exercise training-induced muscle mitochondria 200–9; adenosine monophosphate
hypertrophy 279–82; PGC1 and myostatin as kinase (AMPK) 203; disorders 205, 206;
regulators of RET-induced muscle hypertrophy exercise adaptations 205–8; functions 204–5;
281–2; upstream signaling to mTORC1 in future research 208–9; genetic contributions
response to RET 280–1 to mitochondrial traits 200–9; healthy
intrinsic endurance exercise capacity 165–9; individuals 205–8; mitochondrial biogenesis
genetic mapping 166–9; inbred strain 202–4; mitochondrial biogenesis genes 206–8;
comparisons 166; quantitative trait loci (QTL) mitochondrial biology 201–5; mitochondrial
168; selected strains/artificial selection 165–6 DNA 201–2, 203, 205–6, 208–9; mitochondrial
intrinsic levels of endurance traits, heritability regulation 202–4; mitochondrial respiratory
of endurance traits (human research models) chain 204–5; mutations 205; NuGEMP
178–81 variants 206–8; origin 201–9; single nucleotide
isokinetic torque, muscle strength heritability 256–7 polymorphisms (SNPs) 206–8; stimuli responses
isometric strength tests, muscle strength 200; structure 201–2
heritability 256 mitochondrial DNA 203, 208–9; origin and
structure 201–2; variants 205–6
Johnson, Jack 433 modeling see also heritability of endurance traits
Johnson, Michael 437 (human research models); human systems
Joint Genome Institute 16 genetic modeling: modeling athletic ability,
“junk DNA” 20 health and disease, elite athletes 360
molecular pathways, endurance training
Kane, Martin, race and sports performance 435–7 adaptations 151–9; AMPK (5ʹ AMP-activated
Kenyan distance runners, race and sports protein kinase) 154, 155–7; CaMK 154;
performance 440 CaMKII (calcium/calmodulin-dependent
Krebs’s cycle, peripheral mechanisms regulating protein kinase II) enzyme 157, 158; eNOS
activity 112 and nitric oxide signaling 159; nitric oxide
(NO) 159; NRF2 (nuclear respiratory
Lancaster CS 81 factor 2) 158–9; p38 MAPK 154, 157, 158;
lifestyle changes: endurance phenotype 139–40; PGC-1α (peroxisome proliferator-activated
human evolution 139–40 receptor gamma coactivator 1 alpha) 153–5;
lifestyle comparisons, exercise effects 11 SIRT (sirtuin) proteins 158;VEGF (vascular
lifestyle disease, elite athletes 360–1 endothelial growth factor) 153–5
linkage: gene-finding studies 74 Molecular Transducers of Physical Activity
linkage analysis: genetic mapping 166–7; Consortium (MoTrPAC): future research 8–9,
genome-wide linkage studies, single nucleotide 490; muscle strength genetic contributions
polymorphisms (SNPs) 74; heritability of 267–8
endurance traits 166–7; sport performance, monozygotic (MZ) twinning see also dizygotic
linkage analyses and the candidate gene (DZ) twinning: MZ discordant pairs to study
approach 349–50 causal associations 36; twin studies 31–6
liver disease, myostatin (MSTN) 318 morphological approach, endurance phenotype
longitudinal studies, muscle mass and strength 138–43
heritability 259 motivation for physical activity 94–102; “behavior
lower-triangle model, twin studies 34–6 evolves first” 94–5
MoTrPAC (Molecular Transducers of Physical
major study cohorts, elite athletes 362–7, 364–6 Activity Consortium): future research 490;
major study consortia, elite athletes 362–7, 366 muscle strength genetic contributions 267–8

509
510

Index

mouse–rat–human comparative genomics, of muscle strength and size by inflammatory

heritability of endurance traits 171–2 factors 271–2; tumor necrosis factor-alpha
MSTN see myostatin (TNFα) 271
multivariate heritability studies for muscle mass muscle strength development: Functional Single
and strength 257–9 Nucleotide Polymorphisms Associated with
multivariate twin models, twin studies 34–6 Human Muscle Size and Strength (FAMuSS)
Munthe, Christian, gene doping 454 study 266–7; muscle strength genetic
muscle, skeletal see skeletal muscle contributions 266–7; nervous system adaptations
muscle aging, alpha-actinin-3 337–40 266–7; strength training programs 265
muscle disease, alpha-actinin-3 337–40 muscle strength genetic contributions 264–73;
muscle disuse, alpha-actinin-3 337–40 future research 272–3; genetic variations
muscle factors, peripheral mechanisms regulating associated with muscular strength –
activity 110–12 methodological approaches 267–8; limitations
muscle fiber type distribution, muscle mass 272–3; Molecular Transducers of Physical
heritability 255 Activity (MoTrPAC) 267–8; muscle strength
muscle functional tests, muscle strength cellular pathways 268–72; muscle strength
heritability 257 development 266–7; muscular strength
muscle hypertrophy: resistance exercise training primer 264–5; practical applications 272–3;
(RET) 277, 279–82; ventricular hypertrophy, relationships between muscle mass and strength
cardiac function 190 gains in response to RET 282–3; skeletal muscle
muscle mass and strength responses to exercise, composition 265–6
genetic factors role 259–60 muscle strength heritability 255–9; handgrip
muscle mass heritability 253–5; circumferences – strength 256; isokinetic torque 256–7; isometric
anthropometric measures 254; imaging strength tests 256; longitudinal studies 259;
techniques 254; longitudinal studies 259; multivariate heritability studies for muscle mass
multivariate heritability studies for muscle and strength 257–9; muscle functional tests 257;
mass and strength 257–9; muscle fiber type power and jump test performance 257
distribution 255 muscular strength primer, muscle strength genetic
muscle metabolism, ACE rs4340: 236 contributions 264–5
muscle metabolism regulation, alpha-actinin-3 myocardium responsiveness: endurance training
334–5, 335 150–2; whole-body adaptations 151–2
muscle protein synthesis/breakdown, resistance myokines 7–8
exercise training (RET) effects 277–8 myostatin (MSTN): activation/inhibition 311–14;
muscle size and strength: myostatin (MSTN) 312, atrophy 317–18; cachexia 317–20; cancer-
314–16; regulation by growth factors 268–70; associated cachexia 317–19; chronic obstructive
traits heritability 253–60 pulmonary disease (COPD) 318–19; endurance
muscle size genetic contributions 277–83; exercise 315–16; gene expression 311–14;
genetic contributors to muscle mass/strength genetic control of muscle size and strength
282; intramuscular sensing and signaling 310–20; hypomethylation of DNA 316; insulin
regulating RET-induced muscle hypertrophy resistance 316–17; liver disease 318; muscle size
279–82; nutrient interactions with resistance and strength 312, 314–16; myostatin inhibitors
exercise training for regulation of muscle 318–19; obesity 316–18; PGC1 and myostatin as
size 278–9; relationships between muscle regulators of RET- induced muscle hypertrophy
mass and strength gains in response to RET 281–2; regulation 311–14; regulation of muscle
282–3; resistance exercise training (RET) and size and strength 269, 270; resistance exercise
muscle hypertrophy 277, 279–82; resistance training (RET) 312, 314–16; secretion 311–14;
exercise training (RET) effects, muscle protein signaling 311–14; SMAD2/3 phosphorylation
synthesis/breakdown 277–8 313; SMAD7 313; type 2 diabetes 317–18
muscle strength and performance: alpha- myostatin inhibitors 318–19
actinin-3 323–40; alpha-actinin-3 influences MZ see monozygotic twinning
330–7; alpha-actinins and skeletal muscle
323–4 National Human Genome Research Institute
muscle strength cellular pathways 268–72; (NHGRI) 16
ciliary neurotrophic factor (CNTF) 272; National Institutes of Health (NIH):
growth factors, regulation of muscle size and Common Fund Program 8–9; Molecular
strength by 268–70; interleukin-6 (IL-6) Transducers of Physical Activity Consortium
271–2; interleukin-15 (IL-15) 272; regulation (MoTrPAC) 8–9

510
511

Index

natural reward circuit: evolution of behavior 96–7; performance, sport see race and sports
neurogenetics of motivation for physical activity performance; sex and performance; sport
96–7 performance
nature vs. nurture: predicting sport performance gap: nurture influences 418–19;
performance 442–50; race and sports quantifying the performance gap 422–3, 424;
performance 442–50 sex and performance 418–23, 424; sex testing
nervous system adaptations, muscle strength in sport 480, 482
development 266–7 peripheral mechanisms regulating activity 107–13;
neuroendocrine influence on metabolic flexibility, annexin family of proteins 110–11, 112–13;
endurance exercise 150 calsequestrin 1 (Casq1) 111, 112–13; capability
neuroendocrine responses to RET, genetic to be active 108–12; evaluating voluntary
contributions 290–305; acute testosterone physical activity 107–8; glycolysis enzymes 112;
responses and chronic adaptations to RET IL-15Rα 1775AA genotype 110; Krebs’s cycle
292; androgen biosynthesis and skeletal 112; muscle fatigue 110–12; skeletal muscle
muscle steroidogenesis 293; androgen receptor factors 110–12; substrate utilization 109–10, 109
adaptations to training 296; androgen receptor peroxisome proliferator-activated receptor gamma
genetic variations 293–5; androgen receptor coactivator 1 alpha (PGC-1α), endurance
structure 295–6; androgens 291–6; androgen training adaptations 153–5
signaling 293; factors influencing the acute peroxisome proliferator-activated receptor gamma
hormonal responses to RET 290–1, 291; coactivator 1-alpha (PGC1α: α1, α2, α3, and
glucocorticoids 302–5; growth hormones 296– α4,): intramuscular sensing and signaling
9; insulin-like growth factor 1 (IGF1) 299–302; regulating resistance exercise training-induced
testosterone 291–2 muscle hypertrophy 281–2; PGC1 and
neurofilament, concussion 392, 395 myostatin as regulators of RET-induced muscle
neurogenetics of motivation for physical hypertrophy 281–2
activity 94–102; “behavior evolves first” 94–5; phylogeny of hominidae 136–7, 136
chromatin 98, 101; dopamine 97–8; genetic physical activity see also exercise; physical inactivity:
regulatory mechanisms 101; hippocampus 100; animal models 55–6; apes and humans 137–8;
natural reward circuit 96–7; neurotransmitter candidate genes 112–13; comparative biology
systems 97–8; reward circuitry 98–100 137–8; genetic regulation, animal models
neurotransmitter systems, neurogenetics of 55–64; genetic regulation, human studies 67–76;
motivation for physical activity 97–8 heritability 56–8; hippocampus 100; motivation
NIH see National Institutes of Health for 94–102; neurogenetics of motivation 94–102
nitric oxide (NO), endurance training physical inactivity 9–13 see also exercise; physical
adaptations 159 activity; economic costs 12–13; epigenetics
noncoding RNA and antisense sequences, Human 11; future, look to the 10–13; low voluntary
Genome Project 20 running animal model 11–12; population
noncoding sequences, Human Genome Project statistics 41; reference group 9–10; technology
20–1 10–11; type 2 diabetes 11
NRF2 (nuclear respiratory factor 2), endurance physiologic responses to endurance exercise
training adaptations 158–9 149–50; blood flow redistribution
nurture influences: performance gap 418–19; sex 149–50; blood volume perseveration 150;
and performance 418–19 cardiorespiratory responses 149; metabolic
nurture vs. nature: predicting sport flexibility 149, 150; neuroendocrine influence
performance 442–50; race and sports on metabolic flexibility 150; thermoregulatory
performance 442–50 and blood volume perseveration 150
nutrient interactions with resistance exercise Pima Indians, inheritance of low voluntary
training for regulation of muscle size 278–9 running behavior 11
plasma ACE concentration/activity, and endurance
obesity, myostatin (MSTN) 317–18 performance, ACE rs4340: 222–3
overfeeding, influencing the genetic regulation of population statistics, exercise/physical inactivity 41
activity 124–8 power and jump test performance, muscle strength
heritability 257
p38 MAPK pathway, endurance training POWERGENE Consortium, elite athletes 366,
adaptations 154, 157, 158 367, 368
paradigms for exercise training in rodents 165 practical applications, muscle strength genetic
peptides release 7–8 contributions 272–3

511
512

Index

predicting sport performance 442–50; ethical responses to exercise training 169–71; inbred strain
challenges 442–50; genetic information 442–50; comparisons 169–70; selected strains/artificial
genetic testing 444–50; scientific challenges selection 169
442–50; talent identification 442–50 RET see resistance exercise training
prohibited substances/techniques, World Anti- reward circuitry: neurogenetics of motivation for
Doping Agency (WADA) 457 physical activity 98–100; specificity 98–100
pulmonary function: altitude-induced
erythropoietin (EPO) 192; candidate genes 194; sarcomeric alpha-actinins regulate calcineurin
cardiorespiratory fitness (CRF) 191–4; genetics signaling 333–4
192–3; heritability 191–2; hypoxia 192–3; screening, genetic, talent identification 445–9
hypoxia adaptations 194; hypoxia-inducible selected strains/artificial selection: intrinsic
factor 1α (HIF-1α) 192; quantitative trait loci endurance exercise capacity 165–6; responses to
(QTL) 191–2 exercise training 169
sequencing advances, Human Genome Project
quantitative trait loci (QTL) 58–64, 82; candidate 17–18
genes 189–90; cardiac function 189–90; sex and performance 416–26 see also race and
heritability of endurance traits 164–73; intrinsic sports performance; sport performance; body
endurance exercise capacity 168; pulmonary composition 425–6; historical perspectives
function 191–2 416–18; hormones 425–6; nurture influences
418–19; performance gap 418–23, 424; sex
RAAS see renin–angiotensin–aldosterone system differences due to biology 425–6; sports
race and sports performance 433–40 see also sex participation and opportunity as a nurture
and performance; sport performance; black factor 421–2
athletes 433–40; Cobb, W. Montague 434; sex hormones: endocrine-disrupting chemicals
Edwards, Harry 435–7, 438–9; Entine, Jon (EDCs) 119–22; influencing the genetic
438–40; Johnson, Jack 433; Johnson, Michael regulation of activity 117–19
437; Kane, Martin 435–7; Kenyan distance sex testing in sport 475–82; 1936–1966: sex
runners 440; Snyder, Jimmy “The Greek” verification 475–6; 1967–1988: chromosome
438–9; watershed episode 435–6 formula 477–8; 1991–1999: search for SRY
reference group, exercise effects 9–10 478–9; 2000–2011: suspicion-based testing
renin–angiotensin–aldosterone (RAAS) 479; complete androgen insensitivity syndrome
system: angiotensin-converting enzyme gene (CAIS) 478; ethics 475–82; history 475–82;
(ACE) 220–2, 245; classical/non-classical view hyperandrogenism 480–1; impossibility of fair
220–2 481–2; performance gap 480, 482; testosterone
research see also future research: needs 9; NIH question 480–1; Therapeutic Use Exemption
Common Fund Program 8–9 (TUE) 481; World Anti-Doping Agency
resistance exercise training (RET): acute (WADA) 480–1
responses and chronic adaptations of shared environmental variance, definition 29
glucocorticoids to RET 303–4; acute single nucleotide polymorphisms (SNPs) 18–20,
responses and chronic adaptations of IGFs 37; evolution of genetic mechanisms controlling
to RET 301–2; genetic contributions physical activity 86–9; genes containing activity-
to neuroendocrine responses 290–305; related SNPs 86–9; genome-wide association
growth hormones responses 299, 300; studies (GWAS) 75; genome-wide linkage
intramuscular sensing and signaling regulating studies 74; mitochondria 206–8
RET-induced muscle hypertrophy 279–82; SIRT (sirtuin) proteins, endurance training
muscle hypertrophy 277, 279–82; muscle adaptations 158
protein synthesis/breakdown, effects on skeletal adaptations: endurance phenotype 140–2;
277–8; muscle size and strength 312, 314–16; energy economy 140; evolution 140–2; high
myostatin (MSTN) 312, 314–16; relationships forces 140–1; physiological adaptations for
between muscle mass and strength gains in endurance 141–2
response to RET 282–3; upstream signaling skeletal muscle, alpha-actinins and 323–4
to mTORC1 in response to RET 280–1 skeletal muscle composition, muscle strength
respiratory function, cardiorespiratory fitness genetic contributions 265–6
(CRF) 191 skeletal muscle factors, peripheral mechanisms
response of endurance traits to exercise training, regulating activity 110–12
heritability of endurance traits (human research skeletal muscle responsiveness: endurance training
models) 183–4 150–2; whole-body adaptations 151–2

512
513

Index

skeletal muscle size, genetic contributions 277–83 thermoregulatory and blood volume perseveration,
SNPs see single nucleotide polymorphisms endurance exercise 150
Snyder, Jimmy “The Greek”, race and sports thrifty genotype, elite athletes 360–1
performance 438–9 TIGER (Training Interventions and Genetics of
soft tissue injuries 402–11 see also concussion; Exercise Response) study 44–5, 47
candidate gene studies 402–48; clinical timeline of genetic studies in sports science 348
relevance 410–11; collagen genes 404, 407; TNFα (tumor necrosis factor-alpha), muscle
future research 411; genetic risk factors 402–10; strength cellular pathways 271
hypothesis-free approach 408–10 Training Interventions and Genetics of Exercise
somatotroph types and growth hormone isoforms Response (TIGER) study 44–5, 47
297, 298 transhuman case for gene doping 463–73
“spirit of sport”: gene doping 466–7; World Anti- transhumanism, gene doping 469–70
Doping Agency (WADA) 466–7 TUE (Therapeutic Use Exemption): gene doping
sport performance 347–52 see also elite athletes; 458–9; sex testing in sport 481
race and sports performance; sex and tumor necrosis factor-alpha (TNFα), muscle
performance; Athlome Project Consortium strength cellular pathways 271
351; contributing factors 416–17, 417; exercise twin family designs 37
performance, heritability 347–54; family studies twin studies 31–6, 67–73, 372–84 see also family
374–5; future research 351–2; genome-wide studies; ACE model 33–4; alpha-actinin-3 376;
association studies (GWAS) 350–1; linkage assumptions 32; basic concepts of the twin
analyses and the candidate gene approach model 32–3; biology of twins and twinning 31;
349–50; predicting sport performance 442–50; Cholesky decomposition multivariate model
race and sports performance 433–40; summary 34–6; classical twin model 31–2; dizygotic (DZ)
findings on genetics and 347–52; twin studies twinning 31–6; dominance 32–3; elite athletes
374–5 374–5; Finnish Twin Study on Aging (FITSA)
sports participation and genetic factors 374 258; genetic factors and sports participation 374;
sports participation and opportunity as a nurture genetic regulation of physical activity 67–73;
factor, sex and performance 421–2 heritability of endurance traits (human research
sports science, timeline of genetic studies 348 models) 183; history 372–4; lower-triangle
sprint/power elite athlete association studies, model 34–6; meta-analyses 70; monozygotic
alpha-actinin-3 326–7 (MZ) twinning 31–6; multivariate heritability
SRY gene, sex testing in sport 478–9 studies for muscle mass and strength 258;
strength training programs, muscle strength multivariate twin models 34–6; MZ discordant
development 265 pairs to study causal associations 36; physical
stress: definition 8; exercise 8; survival of the activity phenotypes 69–70; results for leisure-
fittest/survival of the fitness 8 time physical activity 70–3; results for moderate
substrate utilization: glucose transport 4 (Glut4) to vigorous physical activity 70–3; results for
109–10; peripheral mechanisms regulating total physical activityresults for total physical
activity 109–10, 109 activity 70–3; results for voluntary exercise
super athletes, elite athletes 360–1 behavior 70–3; sport performance 374–5;
survival of the fittest/survival of the fitness 8 univariate twin model 33–4
systems genetics: potential 13; terminology 1–2 type 2 diabetes: epigenetics 11; myostatin (MSTN)
317–18; physical inactivity 11
talent identification: genetic testing 444–50;
predicting sport performance 442–50 UK Biobank 38; genetic regulation of physical
tau, concussion 392, 394 activity 74, 75–6
technology, physical inactivity 10–11 univariate twin model, twin studies 33–4
terminology: gene doping 463–5; systems genetics unshared environmental exposure, definition 29
1–2; World Anti-Doping Agency (WADA)
463–5 VEGF (vascular endothelial growth factor),
testosterone: biosynthesis 292; neuroendocrine endurance training adaptations 153–5
responses to RET, genetic contributions 291–2; vesicle glutamate transporters (VGLUT),
sex testing in sport 480–1 concussion 396
theoretical foundations, exercise interventions vinclozolin, endocrine-disrupting chemicals
41–2 (EDCs) 119–22
Therapeutic Use Exemption (TUE): gene doping in vivo gene silencing techniques 3
458–9; sex testing in sport 481 Vivo-Morpholinos, gene silencing 112–13

513
514

Index

Washburn SL 81 substances/techniques 457; sex testing in sport

weight-bearing exercise 9–10 480–1; “spirit of sport” 466–7; terminology
whole-body adaptations, endurance training 463–5
151–2
Wilson, E.O. 384, 499 Z-disk proteins are upregulated in Actn3 knockout
World Anti-Doping Agency (WADA): gene mice, alpha-actinin-3 329
doping 454–60, 461, 463–7; prohibited Z-disk stability, alpha-actinin-3 330–3

514