Complicated Grief

How can complicated grief be defined? How does it differ from normal patterns of
grief and grieving? Who among the bereaved is particularly at risk? Can clinical
intervention reduce complications?
Complicated Grief provides a balanced, up-to-date, state-of-the-art account of
the scientific foundations surrounding the topic of complicated grief. In this book,
Margaret Stroebe, Henk Schut, and Jan van den Bout address the basic questions
about the concept, manifestations, and phenomena associated with complicated
grief. They bring together researchers from different disciplines, providing a
broad range of cultural and societal perspectives, to enable the reader to access
the scientific knowledge base regarding complicated grief, on both theoretical and
empirical levels.
The book is divided into four main sections:

• an exploration of the nature of complicated grief;

• diagnostic categorizations;
• contemporary research on complicated grief;
• treatment of complicated grief.

Illuminating the foundations and new innovations in research, Complicated

Grief will be essential reading for professionals working with bereavement such
as clinical psychologists, health psychologists, psychiatrists, and researchers, as
well as graduate students of psychology and psychiatry.

Margaret Stroebe is Professor at the Department of Clinical and Health

Psychology, Utrecht University, and the Department of Clinical Psychology and
Experimental Psychopathology, University of Groningen, The Netherlands.

Henk Schut is Associate Professor at the Department of Clinical and Health

Psychology, Utrecht University, The Netherlands.

Jan van den Bout is Professor of Clinical Psychology at Utrecht University, The
Netherlands.
Complicated Grief
Scientific foundations for
health care professionals

Edited by
Margaret Stroebe,
Henk Schut and
Jan van den Bout
First published 2013
by Routledge
27 Church Road, Hove, East Sussex BN3 2FA
Simultaneously published in the USA and Canada
by Routledge
711 Third Avenue, New York NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa
business
© 2013 Margaret Stroebe, Henk Schut and Jan van den Bout
The right of the editors 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
Complicated grief : scientific foundations for health care professionals /
Margaret Stroebe, Henk Schut and Jan van den Bout (eds). – 1st ed.
p. cm.
ISBN 978-0-415-60393-5
1. Grief. 2. Grief therapy. 3. Loss (Psychology) I. Stroebe, Margaret S. II.
Schut, Henk Prof. III. Van den Bout, Jan.
BF575.G7C66 2012
155.9’37–dc23
2012003796

ISBN: 978-0-415-60393-5 (hbk)

ISBN: 978-0-415-62505-0 (pbk)
ISBN: 978-0-203-10511-5 (ebk)

Typeset in Times by Prepress Projects Ltd, Perth, UK

For Colin Murray Parkes
Contents

List of illustrations x
List of contributors xi

Part I
Introduction1

1 Introduction: outline of goals and scope of the book 3

M argaret S tr o ebe , H enk S chut, and J an van den B o ut

Part II
The nature of complicated grief: conceptual approaches 11

2 Complicated grief: philosophical perspectives 13

R achel C o o per

3 The concept of complicated grief: lessons from other cultures 27

Paul C . R o senblatt

4 On achieving clarity regarding complicated grief: lessons

from clinical practice 40
T herese A . R and o

5 On the nature and prevalence of uncomplicated and

complicated patterns of grief 55
K athrin B o erner , A nth o ny  D . M ancini , and G e o rge  B o nann o

6 Complicated grief in children 68

Atle D yregr o v and K ari D yregr o v
viii  Contents
Part III
Diagnostic categorization: scientific, clinical, and societal
implications83

7 Prolonged grief disorder as a new diagnostic category in DSM-5 85

Paul A . B o elen and H o lly G . P rigers o n

8 Is complicated/prolonged grief a disorder? Why the

proposal to add a category of complicated grief disorder to
the DSM-5 is conceptually and empirically unsound 99
J er o me C . Wakefield

9 Lessons from PTSD for complicated grief as a new DSM

mental disorder 115
J an van den B o ut and R o lf J . K leber

10 Complicated grief in the context of other psychiatric

disorders: PTSD 129
B everley R aphael , J ennifer J ac o bs , and J eff L o o i

Part IV
Contemporary research on risk factors, processes, and
mechanisms143

11 Prospective risk factors for complicated grief: a review of

the empirical literature 145
L aurie A . B urke and R o bert A . N eimeyer

12 Repetitive thought: rumination in complicated grief 162

E dward R . Watkins and M ichelle L . M o ulds

13 Autobiographical memory processes in complicated grief 176

A nn - M arie J . G o lden

14 Attachment insecurities and disordered patterns of grief 190

M ari o M ikulincer and P hillip R . S haver

15 Physiological mechanisms and the neurobiology of

complicated grief 204
M ary- F rances O ’ C o nn o r
 Contents ix
Part V
Treatment of complicated grief: principles, paradigms,
and procedures 219

16 Prolonged grief disorder: cognitive–behavioral theory and

therapy 221
Paul A . B o elen , M arcel van den H o ut, and J an van den B o ut

17 Internet-based bereavement interventions and support: an

overview 235
B irgit Wagner

18 Family therapy for complicated grief 248

D avid W. K issane , Talia I . Z aider , Y uelin L i , and
F rancesca D el G audi o

19 Brief group therapies for complicated grief: interpretive

and supportive approaches 263
W illiam E . P iper and J o hn Ogr o dniczuk

20 Complicated grief after violent death: identification and

intervention 278
E .   K . Rynears o n , H enk S chut, and M argaret S tr o ebe

Part VI
Conclusions293

21 Complicated grief: assessment of scientific knowledge

and implications for research and practice 295
M argaret S tr o ebe , H enk S chut, and J an van den B o ut

Index 312
Illustrations

Figures
13.1 OGM bias in PTSD and CG: the comparisons between
Autobiographical Memory Test and Biographical Memory Test 181
13.2 Proportions of specific memories retrieved to positive and
negative cues across the two groups 182

Tables
7.1 Criteria for prolonged grief disorder, complicated grief, and
bereavement related disorder  92
10.1 Phenomena of posttraumatic reactions and bereavement 132
10.2 Other phenomena: signs of reactive process 133
11.1 Top risk factors of common grief and complicated grief in
order of number of studies in which they were explored 149
18.1 Demographics of participants from 170 families in our current
randomized controlled trial of family-focused grief therapy 255
18.2 Mean and standard deviation scores on the Beck Depression
Inventory for 220 subjects after 6 months’ bereavement
displayed by both intervention arm and family type 256
19.1 Effect sizes for the two primary outcome variables 272
Contributors

Paul A. Boelen is Professor of Clinical Psychology at Utrecht University, cogni-

tive–behavioral psychotherapist at the Ambulatorium – the outpatient mental
health care clinic of Utrecht University – and Head of Post-Master Education
for Health Care Psychologists in the Utrecht region. He has published over
50 peer-reviewed articles, mostly about assessment and cognitive–behavioral
theory and treatment of prolonged grief disorder in children and adults.
Kathrin Boerner is Senior Research Scientist at the Research Institute on
Aging of Jewish Home Lifecare, New York, and an Associate Professor in
the Brookdale Department of Geriatrics & Palliative Medicine at Mount
Sinai School of Medicine. Her training is in developmental psychology with
particular expertise in adult development and aging. Her research, primarily
supported by the National Institutes of Health, focuses on adaptation to major
life changes related to chronic illness, end of life, and bereavement. She has
made major contributions to the literature on coping with serious illness and
bereavement for over a decade.
George A. Bonanno is Professor of Clinical Psychology and Director of the
Loss, Trauma, and Emotion Lab in the Department of Counseling and Clinical
Psychology at Teachers College, Columbia University. His research and
scholarly interests center on the question of how human beings cope with loss,
trauma, and other forms of extreme adversity, with an emphasis on resilience
and the salutary role of personality, positive emotion, and emotion regulatory
processes. His research has been funded by the National Institutes of Health
and the National Science Foundation. He co-edited the book Emotion: Current
Issues and Future Directions (Guilford).
Jan van den Bout is Professor of Clinical Psychology at Utrecht University, The
Netherlands. His main research interest is about emotional problems after loss
and psychotrauma. Being a licensed psychotherapist and cognitive therapist,
he was for many years President of the European Association for Behavioural
and Cognitive Therapies (EABCT). In the aftermath of the Chernobyl disaster,
he carried out humanitarian and research projects in Russia, Byelorussia, and
Ukraine.
xii  Contributors
Laurie A. Burke is a clinical psychology PhD candidate at the University of
Memphis, where she is a bereavement researcher, studying loss and grief. She
is also interested in how individuals uniquely bond with primary attachment
figures, forming deep relationships that are compromised and mourned as a
result of loss. Her recent publications are based on her study of complicated
grief, and the role that social support and spirituality play in helping or hin-
dering loss accommodation for individuals traumatized by loss. Additionally,
she investigates predictors of complicated grief: what makes some grievers
vulnerable to greater levels of severity in bereavement distress. Her recent
projects include an ongoing, thorough examination of the African American
grief experience, violent death bereavement, bereavement in end-of-life care,
complicated grief risk factors, and measuring and developing an intervention
for complicated spiritual grief: a spiritual crisis following loss.
Rachel Cooper is Senior Lecturer in Philosophy at Lancaster University, UK.
Her publications include Classifying Madness: A Philosophical Examination
of the Diagnostic and Statistical Manual of Mental Disorders (Springer) and
Psychiatry and Philosophy of Science (Acumen). She is currently working on
the concept of disorder and on the philosophical foundations of classification
in psychiatry.
Atle Dyregrov is a clinical and research psychologist and the director of the Center
for Crisis Psychology in Bergen, Norway. Dr. Dyregrov is the author of numer-
ous publications, journal articles, and more than 15 books. He has conducted
research on various subjects relating to bereavement, trauma, and disaster. He
is one of the founding members of the European Society for Traumatic Stress
Studies and the Children and War Foundation.
Kari Dyregrov is a researcher at the Center for Crisis Psychology, Bergen,
Norway, and a senior researcher at the Norwegian Institute of Public Health.
She is conducting research in the field of grief, traumatic bereavement, suicide,
and organization of psychosocial assistance in the community. She is the first
author of several publications. She initiated the Norwegian Association for
Suicide Survivors, for which she received the Farberow Award in 2007.
Francesca Del Gaudio is a Research Assistant in the Department of Psychiatry
and Behavioral Sciences at Memorial Sloan-Kettering Cancer Center.
Ann-Marie Jelena Golden is a cognitive scientist and a health psychologist work-
ing at the Medical Research Council (MRC) Cognition and Brain Sciences Unit
(CBU) in Cambridge, UK. Her main applied clinical/research interests are in
the area of grief/loss, depression, anxiety, and trauma. Dr. Golden works with
children, adults, and the elderly who have experienced emotional and cogni-
tive difficulties related to complicated grief, anxiety, depression, trauma, and
stress that may or may not be related to physical problems. She is specifically
interested in the interplay between cognitions and emotions in common mood
and anxiety disorders, looking at emotion regulation, resilience, and intrusive
 Contributors xiii
and avoidant tendencies, as well as autobiographical and working memory. Dr.
Golden’s professional development has benefited from a broad base of skills
related to compassionate focused therapy for disorders associated with shame
and guilt; mindfulness-based interventions; meaning-breaking, meaning-
making; and grief therapy as narrative reconstruction. She is also a co-founder
of the Cambridge Mindfulness Centre (UK). Dr. Golden recently relocated to
Canada (Toronto, Ontario) and has started working with Professor Stephen
Fleming and Dr. Leslie Balmer. She also collaborates with the Mindfulness
Clinic and Princess Margaret Hospital in Toronto.
Marcel A. van den Hout is Professor of Clinical Psychology and Experimental
Psychopathology at Utrecht University. He is a practicing clinical psycholo-
gist, teaches mainly about emotion and its disorders, and has published many
papers, mainly on the understanding and treatment of anxiety disorders.
Jennifer Jacobs is Research Officer at the University of Western Sydney and
has co-authored publications on pediatrics, mental health, and public health.
Current research interests include early childhood prevention of mental health
problems and other difficulties.
David W. Kissane is the Jimmie C. Holland Chair, Attending Psychiatrist
and Chairman of the Department of Psychiatry and Behavioral Sciences at
Memorial Sloan-Kettering Cancer Center. He is also Professor of Psychiatry at
the Weill Medical College of Cornell University.
Rolf J. Kleber is Professor of Psychotraumatology, Utrecht University, and
Head of the Research Foundation Arq, The Netherlands. Professor Kleber has
conducted and supervised research projects on the psychotherapy outcome of
posttraumatic stress disorder, work-related trauma, late sequelae of war stress,
psychosocial consequences of disasters, and cross-cultural aspects of health.
He is also a consultant in the fields of mental health care after serious life
events and cross-cultural psychology.
Yuelin Li is an Associate Attending Psychometrician and Statistician in the
Department of Psychiatry and Behavioral Sciences at Memorial Sloan-
Kettering Cancer Center and Associate Professor of Psychology in Psychiatry
at the Weill Medical College of Cornell University.
Jeffrey C. L. Looi is Associate Professor and Deputy Head of the Academic Unit
of Psychological Medicine, Australian National University Medical School,
Canberra, Australia. He is an academic neuropsychiatrist active in clinical
practice, research, and teaching. He was a Fulbright Scholar at UCLA Medical
School and is a Visiting Researcher at the Karolinska Institute, Stockholm. He
has been a co-author of more than 150 publications in neuropsychiatry and
cognitive neuroscience.
Anthony Mancini is Assistant Professor in the Department of Psychology at
Pace University, Pleasantville Campus, New York. His research focuses on the
xiv  Contributors
different patterns of adjustment people display after an acutely stressful event,
such as the death of a loved one, war, and traumatic injury, and the factors
that predict those patterns. His work also explores the clinical implications of
research findings on grief and posttraumatic disorder. His work is supported
by grants from the National Institutes of Health. He is the author of over 30
articles and book chapters.
Mario Mikulincer is Professor and Dean of the School of Psychology at the
Interdisciplinary Center (IDC), Herzliya, Israel. His main research interests
are attachment theory, terror management theory, personality processes in
interpersonal relationships, coping with stress, and grief-related processes. He
has published numerous books and over 300 journal articles and book chap-
ters, serves as a member of the editorial boards of various journals, and is the
editor of the Journal of Social and Personal Relationships. He received the
EMET Prize in Psychology and a Distinguished Mid-Career Award from the
International Association for Relationship Research.
Michelle Moulds is Associate Professor, Clinical Psychologist, and ARC
Australian Research Fellow in the School of Psychology at the University of
New South Wales, Sydney. Her research comprises experimental and clinical
studies that examine the interplay of cognitive processes and memory in clini-
cal disorders, with a focus on depression. She is a widely published author.
Robert A. Neimeyer is Professor of Psychology, University of Memphis,
where he also maintains an active clinical practice. Neimeyer is an exten-
sively published author, his works including Techniques of Grief Therapy:
Creative Practices for Counseling the Bereaved and Grief and Bereavement in
Contemporary Society: Bridging Research and Practice, and serves as editor
of the journal Death Studies. Dr. Niemeyer is a frequent workshop presenter,
and is currently working to advance a more adequate theory of grieving as a
meaning-making process. Dr. Niemeyer served as President of the Association
for Death Education and Counseling and Chair of the International Work
Group for Death, Dying, & Bereavement. In recognition of his scholarly con-
tributions, he has been granted the Eminent Faculty Award by the University
of Memphis, and made a Fellow of the American Psychological Association.
Mary-Frances O’Connor is a clinical psychologist and Assistant Professor in
the Department of Psychology at the University of Arizona. Her research has
included the neuroimaging of bereaved persons, and the immune and endocrine
aspects of their emotional functioning. She recently completed data collec-
tion for an NIA-funded K01 award, applying these methods to older widows
between 65 and 80 years of age. She has published her work in many scholarly
journals including Death Studies.
John Ogrodniczuk is Associate Professor and Director of the Psychotherapy
Program in the Department of Psychiatry, University of British Columbia.
John has written and published many articles. He serves as Associate Editor for
 Contributors xv
Psychotherapy Research and is on the editorial board for three other journals
including Journal of Personality Disorders.
William E. Piper is Professor Emeritus in Psychiatry at the University of British
Columbia. His research interests include process and outcome investigations
of both individual and group psychotherapies. He has published approximately
200 articles and book chapters as well as six books.
Holly G. Prigerson is Director of the Center for Psycho-oncology & Palliative
Care Research at the Dana-Farber Cancer Institute, and Associate Professor of
Psychiatry at Harvard Medical School. She has published many peer-reviewed
articles on the topic of bereavement. She is an advisor to the DSM-5. The
research that she and her group has conducted has provided the evidence base
for the inclusion of Prolonged Grief Disorder in DSM-5 and ICD-11.
Therese A. Rando is a clinical psychologist and the Clinical Director of The
Institute for the Study and Treatment of Loss (ISTL) in Warwick, Rhode Island,
USA. The ISTL provides mental health services through psychotherapy, train-
ing, supervision, and consultation, and specializes in: loss and grief; traumatic
stress; and the psychosocial care of persons with chronic, life-threatening, or
terminal illness, and their loved ones. Since 1970, Dr. Rando has consulted,
conducted research, provided therapy, written, and lectured internationally in
areas related to loss, grief, illness, dying, and trauma. Current professional
foci include treatment of complicated mourning, self-help after sudden and
traumatic death, loss of a child, the interface between posttraumatic stress
and grief, anticipatory mourning, and specialized intervention techniques in
the treatment of traumatic bereavement. A recipient of numerous professional
awards and a national media resource expert in dying, death, loss, and trauma
for the American Psychological Association, Dr. Rando has over 80 published
works pertaining to the clinical aspects of thanatology. Among them, she is the
author of the forthcoming Coping with the Sudden Death of Your Loved One:
Self-Help for Traumatic Bereavement. Among her existing published works
are Treatment of Complicated Mourning; Grief, Dying, and Death: Clinical
Interventions for Caregivers; and How To Go On Living When Someone You
Love Dies. Dr. Rando also serves on the editorial boards of Death Studies and
Omega. For more specifics on Dr. Rando or The ISTL, visit www.thereserando.
com.
Beverley Raphael is Professor of Population Mental Health and Disasters at the
University of Western Sydney, and Professor of Psychological and Addiction
Medicine at the Australian National University. She is an internationally
recognized expert in the field of bereavement, trauma, and disasters, and has
published extensively in these fields.
Paul C. Rosenblatt is Emeritus Professor of Family Social Science at the
University of Minnesota. His current projects include a study of know-
ing and not knowing in intimate relationships, work on alternative ways to
xvi  Contributors
conceptualize grief, and an analysis of how African American novelists depict
the impact of racism on African American families. He has published several
books on grief and bereavement.
Edward (Ted) Rynearson is Clinical Professor of Psychiatry at the University
of Washington and Medical Director of the Separation and Loss Service at
the Mason Medical Center in Seattle. He has written a wide range of articles
and book chapters on traumatic grief after violent dying and is the author of
Retelling Violent Death and Violent Death: Resilience and Intervention beyond
the Crisis.
Henk Schut is Associate Professor of Clinical and Health Psychology at Utrecht
University, The Netherlands. His research interests cover processes of coping
with loss and the efficacy of bereavement care and grief therapy. Dr. Schut also
works as a trainer for professionals (e.g., medical specialists, funeral directors)
in dealing with bereaved people and he supervises postacademic clinical psy-
chologists in their research projects. He is coauthor of a number of scientific
and professional articles and books on grief, bereavement, and death.
Phillip R. Shaver is Distinguished Professor of Psychology at the University
of California, Davis, USA. His main research interests are attachment theory,
human motivation and emotion, close relationships, personality development,
and the effects of meditation on behavior and the brain. He has published
numerous books, journal articles, and book chapters and serves as a member
of the editorial boards of various journals. He is a fellow of both the American
Psychological Association and the Association for Psychological Science,
received a Distinguished Career Award from the International Association for
Relationship Research, and was President of that organization.
Margaret Stroebe is Professor at the Department of Clinical Psychology,
Utrecht University, and Department of Clinical Psychology and Experimental
Psychopathology, University of Groningen, The Netherlands. Her research
interests cover theoretical approaches to grief and grieving, interactive patterns
of coping, and the efficacy of bereavement intervention. With Henk Schut she
developed the Dual Process Model of Coping with Bereavement. She is widely
published.
Birgit Wagner is a cognitive–behavioral therapist and working as a researcher
at the Department of Psychosomatic Medicine, University of Leipzig, with a
focus on posttraumatic stress disorder and complicated grief. She has published
numerous articles on grief-related issues such as bereavement interventions.
Current research includes Internet-based interventions after loss and trauma in
intercultural settings.
Jerome Wakefield is University Professor, Professor of Social Work, and
Professor of the Conceptual Foundations of Psychiatry, and Affiliate Faculty
in Bioethics and in the Center for Ancient Studies, at New York University.
He holds doctorates in clinical social work and in philosophy, both from the
 Contributors xvii
University of California at Berkeley, and writes on issues at the intersection of
philosophy and the mental health professions. His recent work has focused on
the concept of mental disorder and the validity of psychiatric diagnostic crite-
ria in distinguishing disorder from normal forms of suffering. He is currently
completing a two-volume study of Freud’s case history of Little Hans and its
significance in the history of psychoanalysis, to be published by Routledge.
Edward Watkins is Professor of Experimental and Applied Clinical Psychology,
University of Exeter. He is a research clinical psychologist and co-founder and
director of the Mood Disorders Centre and the Sir Henry Wellcome Building
for Mood Disorders Research. His research is focused on understanding key
cognition–emotion processes in depression, in particular negative repetitive
thought (rumination) and its underlying mechanisms, through experimental
research, and then translating these insights into more efficacious psychologi-
cal interventions such as rumination-focused cognitive–behavioral therapy.
His research has been supported by the Medical Research Council, Wellcome
Trust, and NARSAD. He has published many articles.
Talia I. Zaider is an Assistant Attending Clinical Psychologist in the Department
of Psychiatry and Behavioral Sciences at Memorial Sloan-Kettering Cancer
Center and Professor of Psychiatry at the Weill Medical College of Cornell
University.
Part I

Introduction
1 Introduction
Outline of goals and scope of the book
Margaret Stroebe, Henk Schut, and
Jan van den Bout

A basic motive in compiling this volume has been to try to gain understanding
of complicated grief, at a time in history when this seems particularly pertinent.
In general terms, complicated grief (CG) can be understood as something like
a “derailing” of the normal, usually painful process of adapting to the loss of
a significant person. However, it will become evident to readers of the current
volume that different definitions and criteria have been adopted to try to describe
the concept more precisely, for both scientific and clinical purposes. To provide
some basis for comparison: our earlier definition of CG has been along the lines of

a clinically-significant deviation from the (cultural) norm (i.e., that could be

expected to pertain, according to the extremity of the particular bereavement
event) in either (a) the time course or intensity of specific or general symp-
toms of grief and/or (b) the level of impairment in social, occupational, or
other important areas of functioning. (Stroebe, Hansson, Schut, & Stroebe,
2008, p. 7)

However, such an apparently detailed characterization has shortcomings (e.g., it

does not explicitly specify different types of complications that have been sug-
gested, such as absent, delayed, or chronic grief). Further difficulties in operation-
alizing and applying such a definition will become apparent through the pages of
this book.
In our view, CG is perhaps the most important contemporary topic of concern
both for the scientific community of bereavement researchers and for health care
professionals supporting bereaved people. This is reflected in a huge expansion
of research, which has resulted in broader and deeper understanding of CG in
recent decades: Earlier assumptions have been put to empirical test, new research
domains have been added, innovative techniques have been applied, novel
theoretical perspectives have been introduced, and significant developments in
intervention programming have been realized. Different academic and clinical
disciplines have been involved in this endeavor. Many of these contributions
have bearing on  – or are a direct consequence of  – the consideration whether
complicated/prolonged grief should be included as a category of mental disorder
4  Margaret Stroebe et al.
in the next edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM-5) of the American Psychiatric Association. Currently, developments point
in the direction of the inclusion of such a category in the near future.
Notwithstanding such advancements, there is often still lack of clarity or inte-
gration, and there are differences of opinion with respect to fundamental issues
surrounding CG – ones that go beyond the problems of definition, raised earlier.
Furthermore, although the body of knowledge on CG has increased consider-
ably in the past few decades, there are still many limitations in understanding
and in empirical investigation. Many fundamental questions remain unanswered.
Many theoretical claims still need to be put to empirical test. Empirical research
has improved significantly over the last decade but it still has not always been
rigorous. Along similar lines, intervention efficacy studies have sometimes lacked
adequate methodology (e.g., control groups, preferably attention-placebo control
groups, and long-term follow-up measurements). Such gaps in knowledge and
shortcomings in investigation need to be identified and directions for future
empirical study inventoried.
In our view, it seems appropriate at this point in time to take stock, to compile
and assess the contribution of the scientific knowledge base regarding CG, on not
only theoretical and conceptual but also empirical levels. Balanced inclusion of
a variety of different perspectives and approaches within one volume is timely,
to enable review of these diverse contributions, and to relate this research base,
where possible, to contemporary societal and practice issues, and to provide criti-
cal appraisal of ongoing research and societal developments relating to the topic
of CG. Thus, the objective for this volume is to provide a balanced, up-to-date,
state-of-the-art account of the scientific foundations surrounding CG. Key ques-
tions will be addressed by our authors, such as: How is CG distinct from normal
grief, or from other psychiatric disorders? Should it be included in DSM? How
can CG best be measured/assessed? Is there an absent grief subtype of CG? How
efficacious is professional intervention? Can we prevent CG, or at least iden-
tify those at most risk? Have we established determinants of CG, or pinpointed
underlying mechanisms? What are the implications of CG for health, well-being,
and daily functioning?
We hope this review will be of relevance to the bereavement research com-
munity, health care professionals, and policy makers. However, it is important to
emphasize that the volume is not designed as a sourcebook for practitioners seek-
ing practical tips or concrete guidelines for intervention with bereaved persons.
Rather, it is intended – insofar as can be done in a single volume – to provide an
overview of contemporary research on CG in relationship to practice, identifying
developments in the field, discussing current controversial issues, encouraging
debate about them, proposing research, and  – again where possible  – deriving
implications for treatment and implementation agendas for the future. In line with
this, the aim of the book is to approach the topic from diverse perspectives, allow-
ing authors the freedom to elaborate on their own scientific standpoint, to respect
different types of scholarship rather than to strive for consensus. Nevertheless, the
editors try in the final chapter to assess the contribution of the volume as a whole
to scientific knowledge about complicated grief.
 Introduction  5
The volume is divided into six parts. After outlining the scope of the volume in
the remaining part of this chapter, the focus in Part II (Chapters 2–6) is on differ-
ent conceptualizations of CG. Here the phenomena and manifestations associated
with CG are explored. These chapters provide a variety of disciplinary perspec-
tives. In Chapter 2, Rachel Cooper reviews diverse philosophical approaches that
can potentially help us address key questions relating to complicated grief, exam-
ining how common philosophical accounts of disorder can elucidate the concept
of CG. First: What is disorder? Fundamental points are raised: Might we think
of normal grief as being a mental injury analogous to physical injury? Second,
philosophical work on the role of classification in science is drawn on to discuss
the question whether CG should be considered a distinct disorder, or a variant
of another condition (ones that are already in the DSM system). At the outset,
with this contribution, awareness is gained of the complexity of the issues we are
dealing with. In Chapter 3, written by Paul Rosenblatt, the perspective shifts to
consideration of CG in different cultures from those Western ones where CG has
mostly been investigated. This raises fundamental problems: Deviant grieving
may not be seen as a “complication” in the same way as Western cultures view it,
or the loss experiences and concerns may be very different from ours, ones that
do not fit the language of a DSM category, for example. Rosenblatt emphasizes
that CG and research on CG are grounded in a particular culture and we should
be cautious about applying this in other cultures. His perspective brings to light
basic questions on a societal level: What, for example, is the place of psycho-
logical treatment in the face of economic, political, or environmental disasters?
In Chapter 4, Theresa Rando elucidates the concept of CG among adults from the
standpoint of her clinical practice, illustrating how a clinical perspective on CG
can clarify scientific understanding. She considers CG as a distinct diagnostic
entity as well as a clinical phenomenon, addressing some controversial and/or
problematic issues, and arguing the need to consider CG from both these angles.
Consideration is given to different forms and functions of CG. She suggests her
own operationalization of CG, based on a conceptual model, which she puts forth
for further discussion and research. In Chapter 5, by Kathrin Boerner, Anthony
Mancini, and George Bonanno, CG is regarded in the context of normal grief,
thus looking more from the perspective of the opposite side of the coin. Focus is
on the complex distinction between uncomplicated and complicated grief. What
distinguishes these phenomena, how prevalent are they, and what underlies the
formation of the different response patterns to loss of a loved one? Drawing on
their extensive research examining the course of grief over time, these authors
relate the patterns to different trajectories of grief and grieving and describe pre-
dictors for complicated or uncomplicated grief patterns. Importantly, they attest to
the resilience of the majority of bereaved people. Finally in this section, attention
turns to the nature of CG in a specific subgroup of bereaved individuals. Within
the scope of this book, valuable though examination of specific kinds of bereave-
ment is for our understanding of CG, we have been able to select only two special
cases. Here we focus on CG among children (in a later section we include consid-
eration of CG following violent death). With few exceptions, there has been little
research on CG among children so far. One of the exceptions has been the work
6  Margaret Stroebe et al.
of Atle and Kari Dyregrov, who contribute Chapter 6. In what ways and to what
extent does CG in children parallel or differ from that among adults? This chapter
covers several important topics, each being related specifically to children’s grief:
the phenomenology of CG; its distinctiveness from normal grief; assessment and
treatment of CG; and risk factors. The authors stress the importance of enhanc-
ing knowledge about the phenomenology, assessment, and treatment of problems
among children.
In Part III (Chapters 7–12), categorization of CG as a mental health condi-
tion is the central underlying theme. This brings the discussion regarding DSM
inclusion, which was mentioned earlier, center-stage. Chapters in this section
provide a variety of viewpoints. In the first of these (Chapter 7), Paul Boelen and
Holly Prigerson present arguments in favor of CG becoming a new psychiatric
condition, based on their extensive research. They review studies that in their
view support the case that CG (or, in their terms, also prolonged grief disorder,
abbreviated to PGD) meets the definition of a mental/psychiatric disorder and that
it should therefore be included in the DSM system. They go on to describe the
criteria that they have derived from their empirical and conceptual work for PGD/
CG. Given this, they explain why they consider it timely to include PGD/CG in
the DSM system; they outline the consensus criteria, detail the available empirical
evidence, and indicate how they think CG should be conceptualized and assessed.
By contrast, in the following chapter (Chapter 8), by Jerome Wakefield, critical
issues to do with CG’s entry in the DSM system are brought forward, so that the
reader is able to assess both the pros and cons of this (potential) major develop-
ment. Wakefield examines six arguments put forward by advocates of the proposal
in support of CG’s conceptual validity. He argues that, despite the enormous effort
to research CG over the past decades, close inspection of these aspects reveals
deficiencies that lead him to conclude that the majority of persons who would be
diagnosed under the CG proposal are suffering from lengthy but normal grief.
Next (Chapter 9), and again with DSM inclusion in mind, Jan van den Bout and
Rolf Kleber draw on the experience gained from the related area of posttraumatic
stress disorder (PTSD) to consider what consequences could ensue should CG be
included as a diagnostic category in DSM-5. PTSD was already included in DSM-
III in 1980, so there has been time to assess the scientific, clinical, and societal
consequences. These researchers identify some positive aspects (e.g., the further-
ance of research on a wide range of traumatic experiences, and development and
testing of new models on origins and maintenance). They also pinpoint matters
for concern (e.g., controversy about diagnostic criteria, clinical utility, and accu-
racy of prevalences) that could apply in the case of CG too. Whereas Chapter 9
focuses on lessons learned from PTSD for a diagnostic category of CG, in Chapter
10, by Beverley Raphael, Jennifer Jacobs, and Jeff Looi, the interest is in placing
CG in the context of other disorders. Again, traumatic experiences and PTSD
provide a useful domain for discussion of this, in the final chapter in this section.
Manifestations and phenomena following the stressors of trauma and loss through
bereavement are described, as well as overlapping versus distinct symptomatol-
ogy (and associated additional complications). Issues of comorbidity and possible
 Introduction  7
etiologies of PTSD and CG are discussed. Finally, assessment and management
strategies for CG with trauma syndromes as comorbidity are outlined.
Part IV (Chapters 11–15) covers contemporary empirical research on risk fac-
tors, processes, and mechanisms associated with CG. First, who among bereaved
persons are most vulnerable to CG; can we identify empirically supported fac-
tors that predict susceptibility to CG? In Chapter 11, Laurie Burke and Robert
Neimeyer have taken on the daunting task of reviewing the extensive body of
scientific research on risk factors, sifting the literature to find those studies that
are the most informative with respect to CG. They cover intra- and interpersonal
as well as situational features that increase an individual’s vulnerability. As far
as possible given the current state of knowledge, they link these risk factors to
the full range of responses to bereavement, and point to factors that merit further
scientific and clinical investigation. Subsequent chapters in this section cover a
range of perspectives that help understand cognitive and emotional functioning
in persons with CG. Two chapters focus on specific processes. First, Edward
Watkins and Michelle Moulds explore the role of rumination and of repetitive
thought in CG (Chapter 12). Given that rumination has to do with repeatedly
dwelling on personal concerns and feelings, it stands to reason that it may be
related to chronic grief. Does it in fact contribute to the development or mainte-
nance of CG? Does it have constructive or unconstructive consequences? These
questions are explored in this chapter, on the basis of extant evidence, considering
recent theoretical and empirical approaches to rumination and exploring their
relevance for understanding CG and its treatment. They provide an integrative
model that helps to guide both research and clinical practice. Second, Ann-Marie
Golden focuses on autobiographical memory (Chapter 13), that is, memories
specifically concerned with the recollection of previously experienced personal
events, which – importantly – contribute to a person’s sense of self. In the context
of CG, one key feature of autobiographical memory is overgenerality (the ten-
dency to recollect in terms of regularities across multiple experiences rather than
specific events), being linked as it is with psychopathology. The phenomenon of
overgeneral memory bias in relationship to CG is described, research is reviewed,
and methodologies and limitations are suggested. Future research lines are laid
out specifically for CG investigation, to further detail the psychological mecha-
nisms involved; the clinical relevance of this research line is considered. On a
different level, the next chapter in this section (Chapter 14), by Mario Mikulincer
and Phillip Shaver, examines the relationship between attachment patterns and
CG, following the attachment theory perspective. This theory has had and still
has enormous influence in the bereavement field. The authors summarize this
approach and present their own psychodynamic model of the activation and func-
tioning of the attachment behavioral system, using it to conceptualize two types
of CG: chronic grief and prolonged absence of grieving. They review relevant
research findings linking anxious attachment to chronic and avoidant attachment
to absent grief. They delineate implications of their perspective for diagnosis,
case formulation, and therapy of bereaved clients with CG. Shifting directions
again, the section closes with a chapter on physiological mechanisms and the
8  Margaret Stroebe et al.
neurobiology of CG by Mary-Frances O’Connor (Chapter 15). This represents
a new area of research into CG. The chapter reviews current studies of these
mechanisms, including functional magnetic resonance imaging (fMRI) studies
of bereavement, genetic analysis, and endocrine data. The evidence is placed in
the context of literature on the physiology of the stress response, the biological
attachment system, and theories of physiological co-regulation. The benefits of
using physiological and neurological variables in bereavement research, as part of
a multi-method approach, are described and it is shown how this line of investiga-
tion has potential to increase understanding of CG.
Part V (Chapters 16–20) spans a variety of topics relating to the treatment of
complicated grief. The first few chapters present a number of intervention prin-
ciples, paradigms, and procedures, to enable the reader to understand different
types of professional help that may be available and appropriate, and to learn the
extent to which these programs have been tested for their efficacy. In the first of
these (Chapter 16) the authors, Paul Boelen, Jan van den Bout, and Marcel van
den Hout, describe a new theoretically based treatment program for CG, based on
the cognitive–behavioral approach (CBT). They review theoretical underpinnings
and the effectiveness of treatment based on CBT. They then describe their own
theoretical approach applying CBT to PGD/CG. They explain how this framework
may be useful not only for generating hypotheses about causes and processes but
for designing effective interventions for PGD/CG. They go on to describe their
treatment program for PGD/CG based on this approach, and to review research
testing the effectiveness of CBT interventions for PGD/CG. Chapter 17, by Birgit
Wagner, introduces a different contemporary perspective to the treatment of CG,
namely, that which is Internet-based; most Internet-based approaches are also
based on a cognitive–behavioral framework. She gives an overview of computer-
ized and Internet-based interventions for (complicated) grief, describing the dif-
ferent treatment approaches that are currently available, ranging from those that
provide some level of therapist support to online bereavement self-help groups.
She describes procedures used in Internet intervention programs and identifies
key components of the therapist-supported interventions. She discusses the
effectiveness of these programs for CG and elaborates on the advantages and dis-
advantages of Internet-based therapeutic interventions. In the next two chapters
in this section we turn from individual to group perspectives in relationship to
treatment. As David Kissane, Talia Zaider, Yuelin Li, and Francesca Del Gaudio
argue in Chapter 18, bereavement is not an event that affects an individual alone,
but one that typically occurs in the family context. Thus, they explore the pos-
sibility that family therapy initiated already during the palliative care phase might
prevent the development of CG. They place their work within the literature on
family intervention in bereavement care and go on to describe their family therapy
program. As they explain, this program targets those deemed at risk; it identifies
types of families that are likely to do poorly. They present preliminary results
regarding the effectiveness of this ongoing research. A different group perspective
is adopted by William Piper and John Ogrodniczuk in Chapter 19. In today’s
cost-conscious environment, group therapies could be considered an increasingly
 Introduction  9
attractive treatment modality. The authors have conducted extensive short-term
therapy groups for clients with CG. They highlight two models on which they base
their therapy, namely interpretive and supportive therapy, detailing the different
objectives of these and comparing them with other therapies in the field. They
give details of the technical manuals produced to guide therapists and illustrate
how these may be used. The various trials that these investigators have so far con-
ducted to try to establish the effectiveness of these therapy programs are outlined
and compared with those of other treatments. Finally in this section, in Chapter
20, Edward Rynearson, Henk Schut, and Margaret Stroebe focus on the second of
the two special types of bereavement included in this volume, one which would
also seem particularly associated with CG (and thus especially relevant for treat-
ment): bereavement following violent causes of death. They first examine the
concept of violent death. Then they review studies on the prevalence and distinc-
tive features of CG following violent death. They describe remaining challenges
(e.g., who precisely should be included in the CG category among those bereaved
following a violent death?). Models of assessment are featured, and an overview
of the limited research on intervention efficacy for CG is provided.
In the last section of the book, Part VI (Chapter 21), the editors review the
contributions to this volume, and they assess the state of scientific knowledge
and the implications for research and practice. They discuss key issues and try
to provide well-balanced conclusions, based on the considerations raised and the
evidence provided by the authors.

Reference
Stroebe, M. S., Hansson, R. O., Schut, H., & Stroebe, W. (2008). Handbook of bereave-
ment research and practice: Advances in theory and intervention. Washington, DC:
American Psychological Association.
Part II

The nature of complicated

grief
Conceptual approaches
2 Complicated grief
Philosophical perspectives
Rachel Cooper

This chapter examines how work in the philosophy of medicine, philosophy of

science, and moral philosophy can help elucidate the concept of complicated
grief. The chapter addresses two key questions.
First, what is disorder? I examine various accounts from the philosophy of
medicine, ranging from the purely descriptive account proposed by Boorse (1975,
1976, 1977, 1997), who argues that disorders are biological dysfunctions, through
to accounts that claim that disorder is a value-laden concept and that disorders are
necessarily bad. In each case, I examine implications for the concept of compli-
cated grief. On some accounts, normal grief might itself be considered a disorder,
and so at points considering whether complicated grief is a disorder becomes
entwined with considering whether normal grief is a disorder. Thus, in the first part
of this chapter, I consider the following issues: Does complicated grief involve a
biological dysfunction? Is grief a bad thing, or a necessary part of the good human
life? Can normal grief and complicated grief be clearly distinguished? Might we
think of normal grief as being a “mental injury” analogous to physical injury?
Second, I will examine how work in the philosophy of science might con-
tribute to debates over whether complicated grief should be considered a dis-
tinct disorder, or whether it should be considered a mere variant of some other
condition, for example major depressive disorder (MDD), posttraumatic stress
disorder, or adjustment disorder. Here I will set out philosophical work on the role
of classification in science, and show how it can contribute to determining the
conditions under which we should conclude that complicated grief is a distinct
kind of disorder.

What is a disorder? Is complicated grief a disorder?

Philosophers have written little that explicitly addresses the issue of whether
complicated grief should be considered a disorder. However, there is a large body
of work on the concept of disorder that can be directly applied to this issue. In the
philosophy of medicine, accounts of disease, disorder, or illness (in the literature
the terms tend to be used interchangeably) can be split into two main camps. On
the one hand, there are descriptivists, who claim that whether a condition is a dis-
order is purely a matter of biological fact. On the other hand, there are normativist
14  Rachel Cooper
positions, which claim that whether a condition is a disorder depends, at least in
part, on whether it is a bad thing.
The best-known descriptivist account has been proposed by Boorse (1975,
1976, 1977, 1997). In a range of publications, Boorse has proposed that a condi-
tion is a disorder if and only if it is a biological dysfunction. His basic idea is
that the bodies and minds of human beings can be thought of as consisting of
numerous subsystems, such as organs, mental modules, and more diffuse systems,
such as the system made up of blood vessels. Each subsystem has a particular
natural function, which is whatever it normally does that contributes to the organ-
ism’s overall goals of reproduction and survival. For example, the function of the
eye is to enable sight, and the function of the heart is to pump blood round the
body. When we are healthy each of our subsystems fulfills its function. However,
when we suffer from a disorder one or more subsystems fail to function at a level
that is average for comparable organisms (i.e., the subsystem functions at a level
significantly below that which we might expect given our age and sex).
Boorse’s account is attractive insofar as it accounts for a number of common-
place intuitions. On Boorse’s account, whether there is a disorder depends simply
on natural biological facts, and medics will thus be best qualified to determine
whether a condition is a disorder.
Despite its attractions, however, Boorse’s account faces problems and has
been heavily criticized. The biggest difficulty for a Boorse-style account is that
many people have a strong intuition that disorders must necessarily be bad. As he
proposes a purely descriptive account Boorse cannot accommodate this intuition.
For Boorse, whether a condition is a disorder comes down to a purely biological
question; if there a biological dysfunction there is a disorder and whether the
condition is a bad thing is irrelevant. Tensions over this point came to a head
in discussions about homosexuality (Bayer, 1981). During the 1970s the ques-
tion of whether homosexuality is pathological was heavily debated. On Boorse’s
account the question hinges on whether homosexuality involves some biological
dysfunction. The possibility of kin-selection effects and other atypical selection
mechanisms mean that answering this question is far from straightforward. In
any case it seemed to many that questions about the biological significance of
homosexuality missed the genuine issue. To many, the key question in address-
ing whether homosexuality is pathological is whether homosexuality is harmful;
insofar as homosexuality is not a bad thing it cannot be a disorder, and questions
about biological functioning seem irrelevant.
Following these debates, Wakefield proposed his highly influential account
of disorder (Wakefield, 1992a, 1992b, 1993, 1999; Chapter 8 in this volume).
Wakefield’s account sets out to accommodate the intuition that a condition can be
a disorder only if it is bad. On his account, disorders are harmful dysfunctions,
where whether a condition is harmful is to be determined by current social norms.
For Wakefield, whether or not homosexuality is a dysfunction, insofar as it is not
harmful, it will not be a disorder. Wakefield’s account has been hugely influen-
tial. Key figures involved in the construction of the DSM (the classification of
mental disorders produced by the American Psychiatric Association) have written
 CG: philosophical perspectives  15
of their admiration for his account, and indeed suggest that explicitly adopting
Wakefield’s account might helpfully guide future revisions to the classification
system (Spitzer, 1999).1
For Boorse, if one wants to know whether complicated grief is a disorder one
should ask whether it is a biological dysfunction. For Wakefield, in order to be a
disorder, complicated grief would need to both involve some evolutionary dys-
function and also be harmful.
Boorse has not written about grief, but Wakefield considers normal grief and
complicated grief in a number of his writings. Writing together with Horwitz
in The Loss of Sadness: How Psychiatry Transformed Normal Sorrow into
Depressive Disorder (Horwitz & Wakefield, 2007), Wakefield argues at length
that forms of normal sadness, amongst which he includes normal grief, must be
distinguished from depression. Using his account of disorder, he suggests that,
although both normal sadness and depression may be harmful, normal sadness
involves no biological dysfunction, and should thus not be classed as a disorder,
whereas true depression does involve dysfunction and thus counts as a disorder.
Why do Horwitz and Wakefield think that normal sadness is normal functioning?
Because they think that there is an evolutionary explanation for those feelings of
sadness that arise in appropriate circumstances. Normal feelings of sadness can
be identified thus:

they emerge because of specific kinds of environmental triggers, especially

loss; they are roughly proportionate in intensity to the provoking loss; and
they end when the loss situation ends or gradually cease as natural coping
mechanisms allow an individual to adjust to the new circumstances and return
to psychological and social equilibrium. (Horwitz & Wakefield, 2007, p. 16)

Horwitz and Wakefield thus think that normal grief does not count as a disorder
because it involves no evolutionary dysfunction. If Wakefield’s claims about
functioning are correct, then a Boorsean would have to agree with him on this
matter.
Horwitz and Wakefield also write, briefly, about complicated grief. Here they
do think it reasonable to think that there is some evolutionary dysfunction, and
thus they are happy to accept that complicated grief counts as a disorder. They
write:

When grief involves extreme immobilisation, pronounced psychotic ideation,

or severe symptoms that persist despite the passage of time and changing
circumstances, then it can be presumed that an individual’s reaction to the
death of an intimate has caused a breakdown in his or her psychological func-
tioning . . . such pathological states constitute Complicated Grief. (Horwitz
& Wakefield, 2007, p. 33)

Following Wakefield, can we simply conclude that normal grief involves no dys-
function and thus is not a disorder, whereas complicated grief does involve some
16  Rachel Cooper
evolutionary dysfunction and thus is a disorder? Unfortunately not. This is what
the accounts of Boorse and Wakefield imply, but there are reasons why many
doubt that their accounts of the concept of disorder are correct. The key concern is
whether it is indeed the case that all disorders have to be biological dysfunctions.
The discipline of evolutionary psychopathology suggests that this may not be
so; there may be some mental disorders that confer an evolutionary advantage
and are thus not dysfunctions (Wilson, 1993). Psychopathy or generalized anxi-
ety disorder, for example, may have an evolutionary explanation (Akiskal, 1998;
Mealey, 1995). Of course, all evolutionary-based accounts of psychopathology
are controversial. However, it at least makes sense to think that some disorders
may have an evolutionary explanation, and this is sufficient to show that it cannot
be part of our concept of disorder that there has to be an evolutionary dysfunction.
Wakefield’s account implies that by definition all disorders would have to be evo-
lutionary dysfunctions, but evolutionarily adaptive disorder is not an oxymoron.
Given the criticisms of Wakefield’s account, some accounts of disorder have
been developed that completely separate the question of whether a condition is
a disorder from the question of whether there is an evolutionary dysfunction. A
range of such accounts are on offer. Some claim that individuals are healthy if
they have bodies and minds that will enable them to live good lives (at least if
environmental and social conditions are favorable). To a first approximation, this
is the view of Megone (1998, 2000), Nordenfelt (1995), and Richman (2004),
although these philosophers differ in the details of their accounts and on how they
characterize the good life. Other philosophers agree that disorders are necessarily
bad states, but think that additional criteria must also be met before a condition
can be considered a disorder. Reznek (1987, pp. 163–164) proposes that a disor-
der is an abnormal bodily or mental condition which requires medical interven-
tion and which harms standard members of the species in standard conditions. He
takes it that we decide what we will count as abnormal (“abnormal” functions as
a call to action stating that we consider dealing with the harmful condition to be a
priority), and that “medical interventions” can be defined enumeratively, using a
list of possible pharmacological and surgical interventions (p. 94). Along similar
lines, I have argued that by “disease” we mean a condition that it is a bad thing to
have, that is such that we consider the afflicted person to be unlucky, and that can
potentially be appropriately medically or psychologically treated (Cooper, 2002,
2005).
All these accounts hold that a condition can be a disorder only if it is a bad
thing. On such accounts a key question in determining whether complicated grief
(or normal grief) is a disorder is whether it is bad. At first glance, grief in all its
varieties looks to be a bad thing. The grieving person feels unhappy and finds
everyday tasks difficult. However, we should pause before concluding that we
would be better off without grief. Grieving for a loved one involves not simply
negative affect, but also activities such as remembering the good times one has
had. Paul Rosenblatt (1996) points out that “A bereaved parent may remember
a child’s laughter, the tender feelings of holding a sleeping infant, or a child’s
creative mischief. Thus recurrent grief is not like recurrent illness. It can be a link
 CG: philosophical perspectives  17
with the best of life” (p. 55). Not only may grief involve mixed feelings, there
are good reasons to think that grieving for a loved one is essentially tied up with
having loved him or her in the first place. The President’s Council on Bioethics
(2003) considers the possibility of medicating away grief and asks us to consider
the following thought experiment (pp. 254–255). Suppose on your death nobody
mourned. How would you feel if on your death your family and friends simply
popped some pills and forgot about you? Would you consider this a good thing?
Most people do not want their family and friends to simply move on. Those who
are ungrieved in death were unloved in life.
How is it that a capacity to love might be tied up with a capacity to grieve?
According to many accounts, to have an emotion at one point in time commits
one to other emotions in other circumstances (Helm, 2001). Thus, if I love my son
then this commits me to feel various other emotions in various circumstances. If
my son does well I will be pleased. If he is in danger I will be concerned. And if
he dies I will grieve. On such a picture, grieving for a dead loved one is ration-
ally connected with having loved him or her in the first place. We might want
to qualify this idea somewhat, as the death of a very old and ill person might,
all things considered, be a good thing, and here grief may be less apt. Plausibly,
loving someone will commit me to grieving for his or her death only if the death
is regrettable. Nonetheless there will be a conceptual link between love and grief.
Note that the link between grief and love here is supposed to be a rational con-
nection, rather than being, say, a side-effect of our evolutionary heritage (for the
idea that grief must be a side-effect of our evolved ability to form attachments
see Frances, 2010). The idea is that loving someone commits any rational being,
whether that being has the evolved vulnerabilities of humans or not, to feel grief
in appropriate contexts. On this picture grief is not some unfortunate side-effect
but is essentially tied to our capacity to love.
Following on from such thoughts, Radden and Solomon both argue that the
conceptual link between love and grief is such that grief is a moral emotion
(Radden, 2009, p. 102; Solomon, 2007, p. 75). Grief on the regrettable death of a
loved one is not only expected, or rationally appropriate, it is morally required. In
appropriate circumstances, a virtuous person will feel grief, and one who does not
grieve is condemned as callous (assuming that some pathology is not preventing
grief).
I suggest that the idea that there is a conceptual link between loving someone
and grieving when he or she dies is on the right tracks but we need to be careful
when thinking about exactly what is implied. In the normal case I will love some-
one when they are alive and then some time later, when they die, grieve for them.
However, suppose that something happens that prevents me from grieving: I die
before the loved one, or come to suffer from severe dementia, or take medication
that flattens my emotional responses. As all these things happen after the time
when I loved, my love cannot be affected by these later happenings; we should not
countenance the possibility of backwards causation here. Thus we should not say
that if we love someone this implies that we must later grieve at their death, but
rather that if we love someone this implies that we will be vulnerable to feeling
18  Rachel Cooper
grief later (if we are still in a state that makes grief possible when they die).
On such a picture, if I entered into relationships knowing that in the event of
bereavement I would use drugs to take away feelings of normal grief, this would
be problematic, as, insofar as it removed the risk involved in loving, it would alter
the nature of love. (In the same sort of way, rock climbing with ropes is a different
type of activity from free climbing.)
Suppose we accept that exposing ourselves to the risk of grief is an essential
part of loving someone. Following such reasoning we might see normal grief as
an essential risk in a good human life. Still, only grief that is proportionate to the
loss is conceptually tied to love. On such a picture we can imagine two types of
problematic case. First, there is the person who grieves too little. Such a person
either is suffering from a pathological condition – absent grief, or repressed grief,
or whatever – or has some character flaw, such as callousness, and never truly
loved in the first place. Second, there is the person who grieves too much or too
severely. Such a person suffers from complicated grief.
Can we conclude that normal grief is a part of the flourishing human life
and thus normal, whereas complicated grief is grief that is disproportionate and
thus pathological? Unfortunately, matters are not quite so clear-cut, as problems
emerge in determining whether grief is proportionate. How long should a person
grieve? When I imagine my death, it seems to me both fitting and right that my
partner would feel sad for about 6 months. However, on reflection it can be no
accident that this seems right to me! As an inhabitant of a twenty-first-century
European country this is what I have come to expect. However, we know that in
some cultures the period expected for mourning is comparatively short, whereas
in others a truly dedicated partner is expected to mourn for much longer (Stroebe,
Gergen, Gergen, & Stroebe, 1996). Given that our expectations about normal
grieving are clearly culturally shaped, and given that determining how long one
should grieve on some basis other than societal norms will be deeply problematic,
distinguishing between normal grief and grief that is too intense or lasts for too
long will be difficult. Normative accounts of disorder tell us that grief is not a
disorder so long as it plays a proper part in enabling us to lead flourishing lives,
but it is unclear how we might decide how much grief a flourishing human should
feel. At bottom the question depends on decisions about the sorts of individuals
we want to be and the sorts of societies we want to live in – and these are hard
issues indeed. To make things yet harder, not only is determining the nature of the
good life intrinsically difficult, complications arise when we reflect on the fact
that our ideas about the good life are plausibly shaped by the economic and politi-
cal structure in which we live. Contemporary Western culture idealizes people
who are independent, happy, and reliable, and it is surely no accident that these
characteristics are also those that enable an individual to be economically produc-
tive in an advanced capitalist society. Persons whose grief is “dis-ordered” are a
liability in our society, and this will create additional pressures to medicalize the
symptoms of unusually intense or long-lasting grief (Walter, 2006). Very severe
and long-lasting grief is plausibly a bad thing, but drawing the line between the
normal and the pathological is problematic.
 CG: philosophical perspectives  19
So far we have considered accounts of disorder that specify criteria that must
be met for a condition to count as a disorder. Boorse thinks that disorders are bio-
logical dysfunctions, Wakefield that disorders are harmful dysfunctions, and the
other philosophers we considered hold that disorders must be harmful (and maybe
meet some other criteria too). If we adopt one of these accounts then the way to
find out whether complicated grief (or normal grief) is a disorder is to see if it
meets the necessary and sufficient conditions. Following such reasoning, we have
considered whether complicated grief, or normal grief, might be an evolutionary
dysfunction, and whether these conditions are harmful or necessary components
of a good human life.
Apart from the philosophical accounts of the concept of disorder that we have
considered, there are also definitions that lie outside the philosophical traditions
but that have been influential and much discussed. Most importantly, the DSM
has included a definition of mental disorder since the publication of the DSM-III
in 1980:

each of the mental disorders is conceptualized as a clinically significant

behavioural or psychological syndrome or pattern that occurs in an individual
and that is typically associated with either a painful symptom (distress) or
impairment in one or more areas of functioning (disability). In addition there
is an inference that there is a behavioural, psychological, or biological dys-
function, and that the disturbance is not only in the relationship between the
individual and society. (APA, 1980, p. 6)

In the DSM-III, uncomplicated grief is distinguished from depression by fiat.

The diagnostic criteria for depression instruct clinicians that uncomplicated grief
“is not considered a mental disorder even when associated with the full depressive
syndrome” (APA, 1980, p.  213). Instead those individuals who manifest what
may be a “full depressive syndrome” in the context of bereavement are to be
given a V-code (i.e., a non-disorder code) (APA, 1980, p. 333). The DSM-III does,
however, allow that bereavement may be “complicated by the development of a
Major Depression,” which may be diagnosed in bereaved persons who display
very severe problems. By the DSM-IV, the reasoning behind excluding grief is
made explicit and the definition of mental disorder has been revised to include a
criterion specifying that the “syndrome or pattern must not be merely an expect-
able and culturally sanctioned response to a particular event, for example, the
death of a loved one” (APA, 1994, p.  xxi). Proposals for the DSM-5 currently
suggest that the grief exclusion clause will be removed from the criteria for MDD.
Proponents of this change reason that there is little to distinguish cases of depres-
sion that are caused by bereavement and those that are caused by other stressors
(Kendler, 2010; Kendler, Myers, & Zisook, 2008). Advocates have also lobbied
for the inclusion of a new category of complicated grief, and current proposals for
the DSM-5 suggest that criteria for bereavement related disorder will be included
in an appendix for further study.
In the philosophical literature, Wilkinson (2000) considers the DSM-IV
20  Rachel Cooper
definition of disorder and how it might apply to grief. He is especially critical
of the DSM-IV’s claim that an “expected and culturally sanctioned response to
external events” should not be considered a disorder. Wilkinson notes that, in
physical medicine, injuries are generally “expected responses to external events”
and yet also are considered as disorders. Following Engel (1961), Wilkinson asks
us to compare grief to a burn, which is an expected response to burning in the
same sort of way that grief is an expected response to the loss of a loved one.
Engel and Wilkinson both suggest that we might conceive of grief as being a
psychic injury. (Alternatively, Wilkinson suggests that if we are convinced that
normal grief is not a disorder then some approach other than the DSM way of
defining disorder will be needed.) The concept of “psychic injury” has been little
explored, and could usefully be considered further.
Finally, two philosophical accounts should be mentioned that both resist the
notion that whether a condition is a disorder should be determined by asking
whether some necessary and sufficient conditions obtain. Lilienfeld and Marino
(1995) suggest that it is not possible to find an adequate definition of “mental dis-
order” because mental disorder is a Roschian concept (Rosch, 1978). A Roschian
concept is one where no definition of the concept can be provided, but where we
decide whether a particular case belongs to a category on the basis of its overall
similarity to prototypical cases. In the case of “mental disorder,” prototypical
examples are schizophrenia and psychotic depression. Other conditions are clas-
sified as “mental disorders” if they seem similar enough to these prototypical
mental disorders. On this account if we want to know whether complicated grief
should be considered a disorder we should ask whether it is similar enough to
conditions like psychotic depression and schizophrenia. On such an account,
complicated grief would plausibly pass the test, and could fairly be considered a
disorder.
In What Is Mental Disorder? Bolton (2008) suggests that if we want to know
when psychiatric treatment is justifiable then setting out by asking whether the
condition is a disorder may not be the most useful approach to take. In his view,
attempts to define the concept of mental disorder have failed. However, Bolton
believes that psychiatry can justifiably continue to operate even in the absence
of a satisfactory account of mental disorder. Bolton seeks to place the distressed
patient at the heart of his account. The patients come for help, and mental
health professionals seek to help them in a way that is recognizably medical, as
opposed to, say, economic or educational. Determining when such interventions
are appropriate depends on diverse factors, such as whether the problem is most
effectively dealt with by medical or other means, and also ethical and political
decisions regarding the sort of lives we wish to lead. Thus, on Bolton’s view, the
key questions will be whether complicated grief can be helpfully treated by health
care professionals, and whether we think such treatments ethically justifiable. If
Bolton is right, philosophical debates about the nature of mental disorder will turn
out to be merely a distraction.
At this point one might be tempted to conclude that as philosophers cannot
agree amongst themselves there is little to be learnt from them. However, I
 CG: philosophical perspectives  21
suggest that some useful pointers can be taken from the philosophical literature.
We have seen that accounts of disorder are contested, but on all available accounts
complicated grief is likely to count as a disorder. Although what counts as normal
mourning varies from culture to culture, insofar as the symptoms of complicated
grief go far beyond those of normal grief in intensity and duration there are
good grounds to think that some evolutionary dysfunction might be involved.
Furthermore, such a syndrome is plausibly not conducive to living a flourishing
life. The symptoms of complicated grief are relevantly similar to those of condi-
tions that are generally accepted to be disorders. Bolton, who is skeptical of the
usefulness of the term mental disorder, will also probably accept that complicated
grief can justifiably be treated by health care professionals. We have come across
only one potential reason to hesitate before concluding that complicated grief is
a disorder: The distinction between grief that is considered appropriate and that
which is excessive is not clear-cut but is culturally shaped. Though this is a worry,
I suggest that we are on safe ground if we hold that we can at least be certain
that very severe symptoms cannot be conducive to a leading a good life, and can
rightly be considered pathological.
The case of normal grief is more problematic. Those who hold that disorders
must involve evolutionary dysfunctions may doubt whether there is any dysfunc-
tion here. Those with normative accounts of disorder may consider normal grief
to be an essential component of a good human life. Insofar as love and grief are
conceptually connected, reducing our capacity to experience normal grief would
reduce our ability to love. On the other hand, Wilkinson has suggested that we
might think of normal grief as a “psychic injury” analogous to a burn, and this
idea might be fruitfully explored further.

Is complicated grief a distinct disorder?

In much of the literature arguing that complicated grief is a valid disorder the
questions of whether complicated grief is a disorder and the question of whether
it is a distinct condition from other conditions are run together. I suggest that these
questions are best kept distinct. We can distinguish two sorts of question:

1 Is condition X genuinely a disorder – as opposed to some type of non-disorder

condition, for example a vice, or a normal variation, or a good variation? For
example, we might ask whether normal grief is pathological or a necessary
component of a flourishing human life.
2 Is condition X genuinely a different kind of condition from condition Y? For
example, we might ask whether complicated grief is distinct from MDD or
posttraumatic stress disorder (PTSD).

The distinction between these two sorts of question can be made clearer by con-
sidering an analogy. Suppose we set out to classify weeds, and define weeds as
unwanted plants. We will face many difficult questions. Are daisies weeds? What
about blackberries? A classification that sets out to list all and only weeds will
22  Rachel Cooper
become mired in controversy. However, in parallel with these questions there will
be other types of question that might prove easier to address: Are blackberries
really a distinct species from raspberries, for example? How should hybrid berries
be classified? In the same sort of way that the question of whether a blackberry
is a weed is distinct from the question of whether blackberries are of a different
species to raspberries, so too the question of whether complicated grief is a dis-
order is a different question from whether complicated grief is distinct from other
conditions (MDD, PTSD, or whatever it might be).
How might we determine whether a condition should be considered distinct
from other conditions? In their classic paper, Robins and Guze (1970) suggest that
a condition can be considered distinct if it is shown to differ from other conditions
in terms of phenomenology, etiology and correlates, outcome, clinical course,
and response to treatment. Following such reasoning, those who argue that com-
plicated grief is best considered to be distinct from other conditions have sought
to locate differences in these validators (Lichtenthal, Cruess, & Prigerson, 2004;
Prigerson, Vanderwerker, & Maciejewski, 2008). Taking a somewhat different
approach, Boelen and van den Bout (2005) use factor analysis to suggest that
complicated grief, depression, and anxiety are distinct syndromes. Meanwhile,
Stroebe and Schut (2005–2006) use a conceptual approach to argue that trauma
and grief are overlapping but distinct (insofar as some peaceful deaths will not be
traumatic, and some traumas do not result in death).
What might the philosophical literature contribute to such debates? I suggest
that an account of classification proposed by Dupré (1981, 1993) may help us
think about the relevant issues. Dupré puts forward an account that he calls pro-
miscuous realism. The key idea is that the world is a messy and complex place,
and that depending on our interests we may usefully classify in diverse ways.
Dupré asks us to imagine a multidimensional quality space in which the entities in
some domain have been plotted (he considers biological organisms, but his ideas
can be generalized). In such a space, entities that are similar will be found close
together, whereas those that are very different will be found far apart. In the space
it will be possible to find various clusters of entities that are highly similar to each
other. We can expect the patterns of similarities to be highly complex; there will
be clusters within clusters, groups of entities that cluster in certain dimensions,
but not others, and so on. Thinking in terms of such a space, Dupré notes that
there will be very many clusters that we might choose to pick out. Depending on
our interests we might focus on certain dimensions, or focus in at greater or lesser
degrees of resolution.
Dupré’s picture is compatible with the reasoning employed in debates whether
complicated grief is a distinct condition. Suppose one wants to argue that com-
plicated grief should be considered a distinct condition. How should one reason?
On Dupré’s picture the important task is to demonstrate that cases of complicated
grief differ from the other condition in some important respect: phenomenology,
treatment response, or whatever it might be. Such reasoning is indeed that adopted
by those who wish to argue that complicated grief is a distinct condition. On the
other hand, suppose one wants to argue that complicated grief should be classified
 CG: philosophical perspectives  23
alongside some other condition. Then the task is to show that complicated grief
and the other condition are alike in some important respect.
The key insight provided by Dupré’s account is that in general it is possible
to produce multiple useful but incompatible classifications of some domain. The
classification one will develop depends on the properties in which one is inter-
ested. Dupré (2001) discusses classification in biology. He notes that species can
be defined in different ways, for example by relations of ancestry or by current
characteristics. Different ways of classifying focus on different properties and
are most useful in different biological subdisciplines. Evolutionary theorists will
find it most useful to classify by patterns of ancestry; ecologists will find it more
useful to classify on the basis of current characteristics. In such a situation, Dupré
suggests that a thousand flowers should be allowed to bloom and that different
subdisciplines should be permitted to classify as they find most useful.
On Dupré’s picture, empirical data are of course relevant to decisions about
how complicated grief should be classified, but, once all the empirical data are
in, deciding whether complicated grief should be classified with other conditions
or apart may be a matter for choice. It may turn out that both those who consider
complicated grief to be a mere variant of some other disorder and those who con-
sider it to be importantly distinct have fair points to make, and that proponents of
the different positions merely concentrate on different features of the conditions.
If this turned out to be the case, on Dupré’s picture it would be permissible to
employ different classifications for different purposes. For example, those inter-
ested in developing treatments might classify in one way, whereas those exploring
the factors that make particular individuals vulnerable to developing the disorder
might classify in another. There may be no one answer to the question of whether
complicated grief should be considered a distinct disorder or a mere variant.

Conclusion
In this chapter I have explored how philosophical work might contribute to elu-
cidating the concept of complicated grief. In the first section, I set out common
philosophical accounts of disorder. Although accounts of disorder are contested,
current prominent accounts suggest that complicated grief should be considered
a disorder. In the second section, I considered whether complicated grief should
be considered a distinct condition or merely a variant of some other condition. I
suggested that insights taken from Dupré’s work on promiscuous realism suggest
that multiple answers to this question might be justified. The world is a complex
and messy place and multiple conflicting classifications might prove useful for
different purposes. For some purposes it may be helpful to consider complicated
grief alongside other conditions; for others it might best be considered separately.

Acknowledgments
I am grateful to Alison Stone and the editors of this volume, who read and com-
mented on an earlier draft of this chapter.
24  Rachel Cooper

Note
1 Despite this endorsement of Wakefield’s account, there are good reasons for thinking
that the DSM has not actually employed an evolutionary account of dysfunction. As
Bolton (2008, pp. 139–151) points out, Wakefield’s account is of limited practical
use, as in most cases of mental disorder whether or not there is an evolutionary
dysfunction remains unclear.

References
Akiskal, H. (1998). Toward a definition of generalised anxiety disorder as an anxious tem-
perament type. Acta Psychiatrica Scandinavica Suppl, 393, 66–73.
APA (American Psychiatric Association). (1980). Diagnostic and statistical manual of
mental disorders (3rd edn.). Washington, DC: American Psychiatric Association.
APA. (1994). Diagnostic and statistical manual of mental disorders (4th edn.). Washington,
DC: American Psychiatric Association.
Bayer, R. (1981). Homosexuality and American psychiatry. New York: Basic Books.
Boelen, P., & van den Bout, J. (2005). Complicated grief, depression, and anxiety as dis-
tinct postloss syndromes: A confirmatory factor analysis study. American Journal of
Psychiatry, 162, 2175–2177.
Bolton, D. (2008). What is mental disorder? Oxford: Oxford University Press.
Boorse, C. (1975). On the distinction between disease and illness. Philosophy and Public
Affairs, 5, 49–68.
Boorse, C. (1976). What a theory of mental health should be. Journal for the Theory of
Social Behaviour, 6, 61–84.
Boorse, C. (1977). Health as a theoretical concept. Philosophy of Science, 44, 542–573.
Boorse, C. (1997). A rebuttal on health. In Hunter, J., & Almeder, R. (Eds.) What is dis-
ease? (pp. 1–134). Totowa, NJ: Humana Press.
Cooper, R. (2002). Disease. Studies in History and Philosophy of Biological and Biomedical
Science, 33, 263–282.
Cooper, R. (2005). Classifying madness: A philosophical examination of the Diagnostic
and Statistical Manual of Mental Disorders. Springer: Dordrecht.
Dupré, J. (1981). Natural kinds and biological taxa. Philosophical Review, 90, 66–90.
Dupré, J. (1993). The disorder of things. Cambridge, MA: Harvard University Press.
Dupré, J. (2001). In defence of classification. Studies in History and Philosophy of
Biological and Biomedical Sciences, 32, 203–219.
Engel, G. (1961). Is grief a disease? A challenge for medical research. Psychosomatic
Medicine, 23, 18–22.
Frances, A. (2010, August 15). Good grief. New York Times.
Helm, B. (2001). Emotional reason. Cambridge: Cambridge University Press.
Horwitz, A., & Wakefield, J. (2007). The loss of sadness: How psychiatry transformed
normal sadness into depressive disorder. Oxford: Oxford University Press.
Kendler, K. (2010). Notes on the proposed the deletion of the grief exclusion criterion from
the criteria for Major Depression. Retrieved November 26, 2010, from http://www.
dsm5.org/about/Documents/grief%20exclusion_Kendler.pdf.
Kendler, K., Myers, J., & Zisook, S. (2008). Does bereavement-related major depression
differ from major depression associated with other stressful life events? American
Journal of Psychiatry, 165, 1449–1455.
 CG: philosophical perspectives  25
Lichtenthal, W., Cruess, D., & Prigerson, H. (2004). A case for establishing complicated
grief as a distinct mental disorder in DSM-V. Clinical Psychology Review, 24, 637–662.
Lilienfeld, S., & Marino, L. (1995). Mental disorder as a Roschian concept: A critique of
Wakefield’s “Harmful Dysfunction” analysis. Journal of Abnormal Psychology, 104,
411–420.
Mealey, L. (1995). The sociobiology of sociopathy: An integrated evolutionary model.
Reprinted in Baron-Cohen, S. (Ed.) (1997) The maladapted mind (pp. 133–189). Hove:
Psychology Press.
Megone, C. (1998). Aristotle’s function argument and the concept of mental illness.
Philosophy, Psychiatry and Psychology, 5, 187–201.
Megone, C. (2000). Mental illness, human function and values. Philosophy, Psychiatry and
Psychology, 7, 45–65.
Nordenfelt, L. (1995) On the nature of health: An action-theoretic approach (2nd edn.).
Dordrecht: Kluwer.
President’s Council on Bioethics. (2003) Beyond therapy: Biotechnology and the pursuit of
happiness. Washington, DC: President’s Council on Bioethics.
Prigerson, H., Vanderwerker, L., & Maciejewski, P. (2008). A case for inclusion of pro-
longed grief disorder in DSM-V. Grief Matters, Autumn, 23–32.
Radden, J. (2009). Moody minds distempered: Essays on melancholy and depression.
Oxford: Oxford University Press.
Reznek, L. (1987). The nature of disease. London: Routledge and Kegan Paul.
Richman, K. (2004). Ethics and the metaphysics of medicine. Cambridge, MA: MIT Press.
Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric ill-
ness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987.
Rosch, E. (1978). Principles of categorization. In Rosch, E., & Lloyd, B. (Eds.), Cognition
and categorization (pp. 27–48.) Hillsdale, NJ: Lawrence Erlbaum Associates.
Rosenblatt, P. (1996) Grief that does not end. In Klass, D., Silverman, P., & Nickman,
S. (Eds.), Continuing bonds: New understandings of grief (pp. 45–59). Philadelphia:
Taylor & Francis.
Solomon, R. (2007). True to our feelings. Oxford: Oxford University Press.
Spitzer, R. (1999). Harmful dysfunction and the D.S.M. definition of mental disorder.
Journal of Abnormal Psychology, 108, 430–432.
Stroebe, M. & Schut, H. (2005–2006). Complicated grief: A conceptual analysis of the
field. Omega, 52, 53–70.
Stroebe, M., Gergen, M., Gergen, K., & Stroebe, W. (1996). Broken hearts or broken
bonds? In Klass D., Silverman P., & Nickman S. (Eds.), Continuing bonds: New under-
standings of grief (pp. 31–44). Philadelphia: Taylor & Francis.
Wakefield, J. (1992a.). The concept of mental disorder: On the boundary between biologi-
cal facts and social value. American Psychologist, 47, 373–388.
Wakefield, J. (1992b.). Disorder as harmful dysfunction: A conceptual critique of
D. S. M.-III-R’s definition of mental disorder. Psychological Review, 99, 232–247.
Wakefield, J. (1993). Limits of operationalization: A critique of Spitzer and Endicott’s
(1978) proposed operational criteria for mental disorder. Journal of Abnormal
Psychology, 102, 160–172.
Wakefield, J. (1999). Evolutionary versus prototype analyses of the concept of disorder.
Journal of Abnormal Psychology, 108, 374–399.
Walter, T. (2006). What is complicated grief? A social constructionist answer. Omega: The
Journal of Death and Dying, 52, 71–79.
26  Rachel Cooper
Wilkinson, S. (2000). Is “normal grief” a mental disorder? Philosophical Quarterly, 50,
289–304.
Wilson, D. (1993). Evolutionary epidemiology: Darwinian theory in the service of medi-
cine and psychiatry. Reprinted in Baron-Cohen, S. (Ed.) (1997) The maladapted mind
(pp. 39–56). Hove: Psychology Press.
3 The concept of complicated grief
Lessons from other cultures
Paul C. Rosenblatt

Psychiatric diagnostic categories and psychiatric standards for what is normal

and healthy and what is not are saturated with the standards of Western culture
(Caplan, 1985; Charmaz & Milligan, 2006; Fabrega, 1987). Psychiatry created
its standards, perspectives, and vocabulary out of the language, ideas, values, and
social forms of Western culture. Psychiatry focuses on emotions, beliefs, inten-
tions, impulses, and actions that have no objective reality independent of culture
(Fabrega, 1987).
Using the standards, perspectives, and vocabulary of one culture in order to
make sense of the emotions, beliefs, intentions, impulses, and actions of people
in other cultures is risky. Can we respectfully evaluate the healthiness and appro-
priateness of what people think, feel, say, and do without understanding what is
healthy and appropriate in their own culture?
Grief scholars are not in full agreement about what constitutes complicated
grief (Stroebe & Schut, 2005–2006). However, I believe the analysis offered here
applies equally well to all conceptualizations of complicated grief as involving
grief that goes on too long and too intensely or that is absent or very muted.
Typically writings about complicated grief do not say explicitly that what they
say about complicated grief applies to people in all cultures. However, it seems
to me that this standard pattern of psychological writing in which statements are
not made about qualifications or limitations with regard to culture implies that
the analysis offered applies universally. It may well be true that there are aspects
of human psychology that are invariant across cultures, but with regard to emo-
tions and concepts of normality versus pathology in expressing emotions there are
reasons to think that a psychological language that makes it seem that all humans
are the same can be misleading. There seems to me to be a claim of applicability
to all humans inherent in writings that say “complicated grief is such-and-such
and comes with these symptoms,” when those writings offer no sense that there
are limits coming from the culture and language of the people who were and were
not studied. The perspective in this chapter is that the concept of complicated grief
and research on complicated grief are grounded in a particular culture, and so we
should be cautious about applying the work on complicated grief to people of
other cultures. Also, to the extent that we all live in pluralistic societies, the cau-
tion would include a sense that, with regard to the people in our own society, the
concept and clinical implications of complicated grief may apply only to some.
28  Paul C. Rosenblatt

In some cultures the symptoms of complicated grief are not

problematic
Wikan (1988) reported that in Cairo, Egypt, a bereaved mother whose grieving
went on for years, with palpable suffering, muted depression, withdrawal, inactiv-
ity, and self-absorption, was seen as sane and making cultural sense. The parents
of Israeli soldiers who died while serving in the military may grieve for decades,
and in Israel that grief is honored and understandable (Malkinson & Bar-Tur,
2000). Charmaz and Milligan (2006) claimed that historically in the United States
and Europe what is now seen as a problem, grief that is too intense over a very
long time, was at one time seen as normal. Similarly, Stearns (1994) wrote that
there was a time when middle-class people in the United States saw intense ongo-
ing grief as more or less inevitable and laudable. Possibly the normality of what
Charmaz and Milligan (2006) and Stearns (1994) wrote about was the normal-
ity of mourning, rather than of grief, if one can distinguish between the rituals
of loss (mourning) and emotional expressions that are not part of rituals (grief).
However, if they were writing mainly about grief, it is only in recent times that
strong and enduring grief has been seen as a problem by middle-class people in
the United States and Europe.

Cross-cultural variation in ideas about deviant grieving

There are indigenous notions of deviant grieving in many cultures. Sometimes
the standards of other cultures resemble the standards implied in the diagnosis of
complicated grief. For example, the Toraja of Indonesia are concerned about any
bereaved person who does not adequately express feelings of grief (Hollan, 1992;
Wellenkamp, 1988). The Ifaluk of Micronesia worry about any newly bereaved
person who does not quickly get back to ordinary daily life after a good cry (Lutz,
1985). However, in some cultures, the standards for grief are at odds with the idea
of complicated grief. The Balinese, for example, worry about a grieving person
who does not present a happy and smooth outer appearance at all times following
a loss (Wikan, 1990). In a number of cultures people might most commonly be
concerned about a deviant grieving that might offend the spirit of the deceased or
fail to help the spirit of the deceased appropriately.
In many cultures, people can be said to try to police or stop the deviant grieving
of a person in their community. In Bali, grieving people who do not seem happy
are cheered up by friends and relatives, teased, and told they should be happy
(Wikan, 1990). From culture to culture the reasons for others to pressure a person
who is grieving deviantly differ. In Bali, the pressure comes because people fear
that a person who is overtly grieving is vulnerable to sorcery (Wikan, 1990). In
another example of community reactions to deviant grieving, Maschio (1992)
wrote about a widow on the island of New Britain who fought to continue her
grieving for her husband when those who had the role of ending her public display
of grieving tried to quiet her. She wanted to continue avoiding pork and wearing
soot marks on her face and a loin cloth of her dead husband around her neck, but
 CG: lessons from other cultures  29
her resistance to ending her public display of grieving provoked anger in others,
and she was scolded.
But then the concept of “deviance” is problematic. The concept of “deviance”
presumes a clear set of standards for grieving. Yet in many societies there are
multiple standards; for example, in Tana Toraja, Indonesia, Christian beliefs and
standards of grieving coexist with indigenous beliefs and standards (Adams,
1993). In fact, a pattern of contradiction between earlier ways of grieving and
more recent ones brought by outsiders is common around the world. Among the
consequences of the multiplicity of standards is that there are many people in
the world who know that however they grieve they will be judged deviant by
the standards of some in their social environment. Consider the Maisin people
of Papua New Guinea, living with the conflict between more recently adopted
Christian standards for grieving and standards that antedate the Christian stand-
ards (Barker, 1985). By older standards, mourning for years after the death is
proper (Barker, 1985), but by the newer, Christian standards, those activities may
be inappropriate. One might look at the older pattern and say that it is in some
sense more basically human; for example, it fits evidence that normal grief can
be long term, even lifelong (Klass, Silverman, & Nickman, 1996). However, the
traditional pattern, according to Barker, was not only about grief but about com-
petition among widows, the relationships among kin groups, beliefs about regen-
eration following a loss, and other matters that are unrelated to continuing grief as
described in the Klass, Silverman, and Nickman book. However, the oddity of the
Maisin by Western standards still reinforces the idea that, in pluralistic societies
where there are multiple and diverse standards by which to evaluate grieving,
there are reasons to question labeling some pattern of grieving deviant and of
concern.

Other problematic assumptions underlying the concept of

complicated grief
From a cross-cultural perspective, there are a number of assumptions that I
suspect are commonly made in writings about complicated grief that are worth
making explicit. One that has already been discussed is the assumption that grief
that goes on too long and too intensely is a problem in need of treatment. What
other assumptions are there that may underlie the concept of complicated grief?

Assumption that someone’s grief arises from a discrete, time-limited

loss
It seems to me that one assumption about complicated grief, when there is a judg-
ment that grief has gone on too long, is that there is a discrete point in time at
which the loss can be said to occur. That point of time serves as the marker from
which the duration of grieving can be measured, and duration matters in what I
have seen of DSM-5 draft texts about grief. However, from a cross-cultural per-
spective, thinking in terms of grief arising at a discrete point in time is challenged
30  Paul C. Rosenblatt
by the experiences of people for whom a specific loss goes on continuously, occurs
again and again, or is part of an ongoing series of losses. There are, for example, a
number of North American native cultures that have experienced severe historical
traumas, including near-genocide, mass sexual abuse, and the destruction of cul-
ture and the social and physical environment, and these losses have been spread
over many years and continue (Brave Heart & DeBruyn, 1998; Tafoya & Del
Vecchio, 2005). A diagnosis of complicated grief and treatment of complicated
grief may well be usefully carried out without assessing that appropriateness of
grief duration, but to the extent that grief duration is a factor in assessment, there
could be situations in which, by the standards of some culture, it is a mistake to
assume grief can be assessed, based on a discrete starting point.

Assumption that grieving in all cultures is rather the same

As was said above, the language of much that is written about complicated grief
could be understood to imply that grieving in all cultures is more or less the same.
If that is assumed, is it a reasonable assumption to make? There is considerable
evidence that people deal with and talk about losses quite differently from one
culture to another (Charmaz & Milligan, 2006; Currer, 2001; Rosenblatt, 2001;
Wierzbicka, 2003).
Wierzbicka (2003) challenged the universality of grief, beginning with the
observation that in her native Polish there is no term for grief. From her perspec-
tive, the English language gives a privileged position to grief, but it is a mistake
to claim universality for grief on the basis of that. For her, the interesting ques-
tion was not something like: When is grief too prolonged? It was: “Why should
English have singled out the experience that it calls grief from the great ocean of
human emotions as a subject of special attention and given it a distinct name?”
(p. 582). Tracing the history of the English term grief, which used to mean other
things than sorrow and pain over a death, she came to the idea that the English
term grief developed as a culturally based idea that grief should be short term
and should not interfere too much with one’s pursuit of happiness (p. 584). This
is in contrast to her native Poland, where she felt that there is a widespread belief
that it is good and appropriate to mourn for a lifetime (Polish has a word for
mourning) for great loss (p. 588). Implied in this is a sense she offered that, in
English, “grief” disrupts normal life, but in some societies normal life is to feel
sorrow and great pain, to mark that something terrible has happened (e.g., p. 592).
So from her perspective not only does culture shape, limit, influence, define, and
give meaning to grief, it also creates it or does not create it as a human emotion
following loss. That I as an observer of Polish people who have experienced a loss
may think I observe grief seems to me, if I understand Wierzbicka correctly, not
to be relevant, since it is not what I observe but what they feel and how they put
words to it that matters. That there is a word in Polish for mourning does not mean
that a person who thinks and feels in Polish does not grieve, because I assume that
Wierzbicka draws the standard distinction between grief as a matter of feeling
and mourning as a culturally called for way of behaving after a loss. So Poles may
 CG: lessons from other cultures  31
mourn in ways that could look to an observer from a culture in which “grief” is
part of the vocabulary as though they were grieving, but what they are doing and
feeling may be rather different from grief.
Related to this, Klass (1999) indicated that there are a number of cultures in
which people’s concerns following a death seem to be about separating from and
being protected from the deceased. Klass then argued that cultures differ in the
meta-interpretative schemes that are engaged when a death occurs, and paying
attention to those schemes we will find that there is not a single grief or mourning
scheme that is dominant in all cultures.
Moreover, implied in the notion of complicated grief is the notion of something
like recovery from grief. Recovery is a concept in Western culture; but in other
cultures there may be no sense of recovery from grief and no sense that something
like recovery is normal or desirable (Rosenblatt, 2008). From this perspective, is
it a mistake to claim that “recovery” in bereavement is or should be a universal
process? And then is it a mistake to impose the concept on people whose cultural
background leads them to other values and processes than recovery from grief
after a loss? And if complicated grief treatment with people from diverse cultures
must be carried out with a diversity of desired outcomes, and not just “recovery”
as a desired outcome, how are those diverse outcomes to be defined and assessed?

Assumption that the individual is what needs healing

Looking at prolonged or intense grief as a psychological problem may lead to
ignoring or discounting what grieving people would say about their economic,
political, or environmental challenges. Imagine that a Guatemalan widow whose
husband has been assassinated by the military because he spoke out against injus-
tices continues years after his death to show symptoms of complicated grief; and
imagine that she says that her pain is in large part about the economic, political,
and environmental system that oppresses poor Guatemalans. If we take her words
seriously, she may not appreciate efforts to help her to move to a state in which
her grief is less intense and dominates her life less. For her, the grief may be about
ongoing injustices, and to lessen or turn away from the grieving may be a betrayal
of important values. In cases like hers, focusing only on a psychological problem
and not on the context and stated realities and desires of those we are trying to
help might not be welcomed or respectful.

Cultural critique of the concept of complicated grief

The foregoing suggests a cultural critique of the concept of complicated grief, and
along with it the concept of diagnostic categories, the psychologizing of problems
that reflect problems in political and other larger systems, and the disenfranchise-
ment of indigenous ways of understanding and dealing with loss. If we want to
understand and help people from various cultures in ways that are most respect-
ful and helpful, we need to examine the fit between their grieving and what is
appropriate in their culture (Shapiro, 1996). If we know what the person’s culture
32  Paul C. Rosenblatt
expects, reinforces, models, and offers as meaning we will understand not only
what is appropriate but also reasons why it is appropriate; for example, what a
death such as the one for which the person is grieving means in that person’s
culture and what various ways of expressing emotion mean.
Arguably there are ways in which, by applying standards of a culture that is
not that of the client or not the only one of the client, diagnosis and treatment can
be harmful. Some critics of contemporary mental health diagnosis and treatment
that is not culturally attuned have raised the question of whether there is policing
linked to the concept of complicated grief and treatment for it (Foote & Frank,
1999; Walter, 1999, 2005). The critics who assert that policing goes on in the
area of bereavement have written that one reason for such policing may be that
bereavement that goes on too long or too intensely is not what employers want.
They also raised questions about the possibility that focus on the bereaved indi-
vidual takes the focus of potential critics away from larger processes that may be
responsible for a loss. For example, treating the widow of a U.S. soldier killed in
Iraq in terms of complicated grief may take the focus away from the circumstances
in U.S. politics that led to and maintained military involvement in Iraq. Inherent
in this line of criticism is the idea that when diagnosis and treatment “police” the
cultural and situational contexts of the person’s grieving are ignored (Foote &
Frank, 1999) and, to the extent that the policing is effective, the person accepts
being defined in that way. Of course, from a culturally attuned perspective, one
would also want to respect the realities of clinicians who provide good faith help
to the bereaved. However, we must keep in mind the Foucauldian perspective,
that therapy polices, in trying to understand the larger contextual forces that may
push psychological treatment in some directions more than others.
A related line of criticism regarding possible cultural and situational insensitiv-
ity in mental health work can be found in Gone’s (2008) remarks about “cultural
proselytizing.” Gone, writing about mental health work with American Indians,
argued that there can be a kind of cultural imperialism in saying mental health
language should be adopted by others. He asserted that promoting mental health
discourse in work with American Indians could be seen as serving to silence crit-
ics of what the culture from which that discourse comes has done to American
Indians. Related to that, Adelson (2008) pointed out how using a mental health
language about psychological stress undermines, smothers, and diminishes dis-
course about the inequities and oppressions that led to the psychological stress. It
becomes a mental health problem, not a problem of historical and current injustice.
From that perspective, imposing a mental health view on people in cultures that
arguably have been oppressed in some way by a dominant Western culture might
be seen as a step toward silencing client critiques of that oppression. As Charmaz
and Milligan (2006) asserted, ongoing, intense grief over injustices often provides
strong motivation to people who are involved in protests and political action that
aim to end political violence and injustices. We are legitimately in the business
of helping people to deal with great psychological pain, but should not be in the
business of suppressing efforts to change oppressive situations. So the clinician
will have to be sensitive to issues of oppression while helping people with what
 CG: lessons from other cultures  33
seems to be complicated grief when that grief can be seen as stemming from
oppressive situations.

Clients from many different cultures

In pluralistic societies (most countries on the planet), it is risky and seen by some
as culturally disrespectful and unhelpful to impose the standards of one culture
on people from other cultures. However, that is what seems to be going on if one
takes the standards one culture has for emotional expression and pathologizes
what people do who are from cultures with different standards. Fabrega (1987)
made that point with regard to illness, and, to the extent that making compli-
cated grief a DSM category makes complicated grief an illness, one can join
Fabrega in raising questions about the appropriateness of applying one culture’s
standards to people from other cultures. Psychiatric diagnosis imposes a set of
cultural meanings, and so psychiatric diagnosis across cultures always raises
questions of validity (Good & DelVechio Good, 1986; Stroebe, Gergen, Gergen,
& Stroebe, 1996; Watters, 2010). How can we be sure we are not reifying the
diagnosing culture’s ways of understanding, making them universally real when
we know only that in their culture of origin they have reality standing? Claims
of universal biological or psychological processes may be made to support the
universalistic application of psychiatric diagnoses, but those claims can also be
understood as culturally based. It is risky to use theories grounded primarily in
a single culture as a basis for prescriptive assertions from and claims by experts
rooted in that culture that there is sure, universal knowledge about what is
normal and good bereavement and what is a problem in need of help (Small &
Hockey, 2001).
One might argue that the complicated grief framework is well supported by
evidence, but the evidence and its interpretation are culturally saturated. As some-
one who has tried to carry out research in a number of cultures, I believe that there
is no psychological research method that is not saturated with the assumptions,
ways of thinking, beliefs, and values of a specific culture. From that perspective
it can be argued that there is not validity in an objective sense for cross-cultural
applicability of research findings, just a claim of validity. For example, the pri-
mary way of assessing complicated grief involves a questionnaire, not a culturally
sensitive interview. That the questionnaire can be translated into the languages of
other cultures and produce similar results does not mean the questions have valid-
ity across cultures and does not mean the questions are attuned at all to cultural
realities in these other cultures. In fact, the very act of translating the questions
is fraught with cultural complexities. In areas of emotion and psychology, things
do not translate well. There is not a standard cross-cultural way for experienc-
ing feelings or putting feelings into words (Kirmayer, 2005). So it is difficult to
know what people’s answers to translated questionnaires mean to them. Writing
as though there are human universals in experience, expression, thought, feeling,
and communication is a rhetorical form and a statement of belief, but ignoring
other realities does not make them go away. Possibly any psychological research
34  Paul C. Rosenblatt
on humans must rest on assumptions that can be questioned from some perspec-
tives, and projects assessing grief with paper and pencil measures certainly make
such assumptions (e.g., Small, 2001). If there is a cultural component to reading
and answering questions and writing questions and interpreting answers, I have
not seen research on complicated grief that accounts for those cultural processes.
If it is assumed that the answers to questions are true and can be interpreted with-
out problems, that, I think, is a mistake.
Everyone I know who works to help grieving people seems to me to be well
intentioned, but there can be dark sides to efforts to curtail mourning. Redmond
(2008), for example, wrote about the pressure from the Australian Ministry of
Health on indigenous Australians to abandon or shorten their mourning periods,
with the explicit goal of making them more ready for work in the mainstream
economy. So the message may be that “we care about your pain,” but it also is
that “we want you to be available to work and to be off the dole.” Related to this,
Pratt (1994) showed how intolerant businesses in the United States are of griev-
ing, giving an average of 3 days of bereavement leave to grieving employees.
Pratt argued that such intolerance of grieving does not reflect culture so much as
it shapes and drives it. From the perspective of the work of Redmond and Pratt,
the form of complicated grief treatment that reflects these larger economic issues
would seem to be treatment of grief that goes on too long and too intensely.
As a number of social critics who write about grief have pointed out, there
has been a movement in the United States, and other societies where Western
medicine has a strong influence, toward constructing grief as a disease instead of
an emotion (Averill & Nunley, 1988; Charmaz & Milligan, 2006; Foote & Frank,
1999; Walter, 2005), and that pushes grief into the domain of the medical model.
Among the many critics of the medical model, the model that makes distress and
social difficulties into illnesses to be treated, are critics who see the model as a tool
of social control, as defining people’s reactions to injustice as illness, as defining
sane reactions to insane situations as insane, as imposing one culture’s realities on
people whose cultures give them quite different realities, and as silencing people
who protest against government and corporate harm-doing by labeling them as
needing therapy. From that perspective we must be very careful not to become
part of the problem in using a medical model of individual distress.
People in one culture can be changed by people from another culture, especially
if the people in the latter culture have economic and political power. Consider the
power in the mental health area of U.S. and European-based drug companies and
the U.S. and European medical establishment, a power that comes with domi-
nance in research funding and publication, considerable power in medical school
and postgraduate medical education, and power to shape public policy. That
power affects the research that is published and becomes a knowledge base for
people of all cultures, affects what is taught to medical students in all cultures, and
affects the power of U.S. and European drug companies to market their products
worldwide. So there are many people around the world who now seem to accept
the U.S. medical model of how people express a wide range of distress (Watters,
2010). The changes are both in expert opinion and practice and in how ordinary
 CG: lessons from other cultures  35
people in the culture think, feel, and act (Watters, 2010). Efforts to change a cul-
ture can work, and we can see that in the history of changing cultural standards
and practices in the United States for grief (Stearns, 1994). The medical model for
dealing with bereavement, like the medical model for other forms of distress, has
become central to the ways that many people around the world think about grief
and how they make meanings and channel their concerns. Thus, these people may
define personal grief that goes on too long or too intensely as a problem in need
of treatment, and thus they may accept a diagnosis of complicated grief and the
treatment that follows such a diagnosis. That does not mean the medical model
is universally valid or that the concept of complicated grief has meaning in all
cultures. However, it does mean that, within certain cultural contexts, complicated
grief can be said to be real. From that perspective, an important task in research on
complicated grief is to be able to distinguish people for whom a complicated grief
diagnosis is meaningful and respectful from people for whom such a diagnosis is
inappropriate.
Therapeutic treatment for bereavement has an important and valuable place
for people whose cultural perspectives make such treatment meaningful and
desirable. However, for clinicians who want to work with bereaved people, the
argument in this chapter is that a checklist of symptoms and the definitiveness
of a universal categorization of some forms of grieving as normal and others as
complicated and in need of therapy must be used with caution. There must be
a genuine openness to the realities, perspectives, cultures, social situations, and
understandings of each person. Two people with the same checklist score may
well be in need of very different attention because of the realities, meanings, and
so on in which their responses are embedded. Clinical work must then rely on the
hallowed clinical traditions of listening and understanding. The clinician must
learn and work with the realities of the client, because it is the client’s realities,
not the therapist’s realities, that are key to understanding what is going on and
providing help.

Cultural formulation
DSM-IV included what has been labeled as “cultural formulation” (American
Psychiatric Association, 1994), which includes the injunction to attend to culture.
For example, there is this in the section on anxiety disorders:

Culturally prescribed ritual behavior is not in itself indicative of Obsessive–

Compulsive Disorder unless it exceeds cultural norms, occurs at times and
places judged inappropriate by others of the same culture, and interferes with
social role functioning . . . Mourning may lead to an intensification of ritual
behavior that may appear to be an obsession to a clinician who is not familiar
with the cultural context. (American Psychiatric Association, 1994, p. 420)

The passages like that in DSM-IV do not say how to assess a person’s cultural
background, how a “norm” is to be assessed, whether the concept of “norm” makes
36  Paul C. Rosenblatt
sense for all clients, and what to do clinically once the cultural background is
known. Moreover the DSM-IV cultural formulation gives mixed messages about
what to do with cultural information. For example, there is this caution about
being too attuned to culture in a discussion of major depressive episodes: “It is
. . . imperative that the clinician not routinely dismiss a symptom merely because
it is viewed as the ‘norm’ for a culture” (p. 324). Also, the cultural material in
DSM-IV that enjoins sensitivity to culture is typically a few sentences surrounded
by many paragraphs written as though culture were irrelevant and there were an
unambiguous and clearly known truth regarding the disorder that would apply in
all cultures. For example, following a statement that points out that the norms for
duration of bereavement vary across cultures the next sentence is “The diagnosis
of Major Depressive Disorder is generally not given unless the symptoms are
still present 2 months after the loss” (p. 684), with no sense of how the previous
sentence might relate to this one. Appendix I (pp. 843–844) of DSM-IV offers
a brief outline of what to assess culturally, but I do not see in it enough help to
assess well or to translate what is learned into clinical decisions and actions. In
fact the cultural formulation was added rather late in the creation of DSM-IV and
without much work at integrating it into the rest of the volume (Mezzich, 2008).
Since the publication of DSM-IV in 1994, there has been a great deal of
development in mental health specialties (including those working with bereave-
ment) of a literature base, workshops, and training programs to produce greater
cultural sensitivity. However, arguably there is still a tension in mental health
fields between the idea that diagnosis and treatment must be tailored to culture
and the idea that the same diagnostic and treatment research and language apply
to everyone.
I imagine that the editors of DSM-5 will go further than those for DSM-IV
in acknowledging the situational and cultural relativity of the DSM diagnostic
classification system. First steps have been taken to develop the field beyond
the mere injunction to be culturally sensitive, for example a guide to cultural
formulation assessments (e.g., Rohlof, 2008). The dual-process model (Stroebe
& Schut, 1998) is one way to account flexibly for cultural differences in grieving.
However, based on my own admittedly unrepresentative and subjective experi-
ence, the DSM cultural formulation cautions are yet to become a universal and
well-integrated part of practice and scholarship. I hope that this chapter and this
volume will help us to move toward the day when the leading edge of scholarship
and clinical practice on complicated grief will provide rich examples of how to
work with the complexities of people’s cultural situations. One must also expect
that efforts to move toward greater cultural sensitivity will have a rational and
deliberative quality but will also involve struggle. As one example of such strug-
gle, DSM-5 apparently will remove the bereavement exclusion from the section
on major depressive disorders, and a number of reputable grief scholars have
written in opposition to that removal (Balk, Noppe, Sandler, & Werth, 2011) on
grounds that could be understood to say that removal of the bereavement exclu-
sion is harmful for all people. One can see in the clash of views differences in
ideas of what is culturally sensitive, and one can also see an area in which there
 CG: lessons from other cultures  37
will be contentious debate. However, we should not shy away from such debate.
Ideally it is in such debate that we learn from each other, ideas are clarified, and
in the end the people we serve benefit.

References
Adams, K.  M. (1993). The discourse of souls in Tana Toraja (Indonesia): Indigenous
notions and Christian conceptions. Ethnology, 32, 55–68.
Adelson, N. (2008). Discourses of stress, social inequities and the everyday worlds of First
Nations women in a remote northern Canadian community. Ethos, 36, 316–333.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4th edn.). Washington, DC: American Psychiatric Association.
Averill, J.  R., & Nunley, E.  P. (1988). Grief as an emotion and as a disease: A social-
constructionist perspective. Journal of Social Issues, 44(3), 79–95.
Balk, D.  E., Noppe, I., Sandler, I., & Werth, J., Jr. (2011). Removing the exclusionary
criterion about depression in cases of bereavement: Executive summary of a report to
the ADEC board of directors. ADEC Forum, 37(2), 24–26.
Barker, J. (1985). Missionaries and mourning: Continuity and change in the death ceremo-
nies of a Melanesian people. Studies in Third World Cultures, 25, 263–294.
Brave Heart, M.  Y.  H., & DeBruyn, L.  M. (1998). The American Indian holocaust:
Healing historical unresolved grief. American Indian and Alaska Native Mental Health
Research, 8(2), 60–82.
Caplan, P. J. (1985). They say you’re crazy: How the world’s most powerful psychiatrists
decide who’s normal. Reading, MA: Addison-Wesley.
Charmaz, K., & Milligan, M.  J. (2006). Grief. In Stets, J.  E., & Turner, J.  H. (Eds.),
Handbook of the sociology of emotion (pp. 516–543). New York: Springer.
Currer, C. (2001). Is grief an illness? Issues of theory in relation to cultural diversity and
the grieving process. In Hockey, J., Katz J., & Small, N. (Eds.), Grief, mourning and
death ritual (pp. 49–60). Philadelphia: Open University Press.
Fabrega, H., Jr. (1987). Psychiatric diagnosis: A cultural perspective. Journal of Nervous
and Mental Disease, 175, 383–394.
Foote, C. E., & Frank, A. W. (1999). Foucault and therapy: The disciplining of grief. In
Chambon, A.  S., Irving, A., & Epstein, L. (Eds.), Reading Foucault for social work
(pp. 157–176). New York: Columbia University Press.
Gone, J. P. (2008). Introduction: Mental health discourse as western cultural proselytiza-
tion. Ethos, 36, 310–315.
Good, B. J., & DelVechio Good, M.-J. (1986). Cultural context of diagnosis and therapy: A
view from medical anthropology. In Miranda, M. R., & Kitano, H. H. L. (Eds.), Mental
health research and practice in minority communities: Development of culturally sensi-
tive training programs (pp. 1–27). Washington DC: U.S. Dept. of Health and Human
Services, Pub. No. (ADM) 86-1466.
Hollan, D. W. (1992). Emotion, work and value of emotional equanimity among the Toraja.
Ethnology, 31, 45–56.
Kirmayer, L. (2005). Culture, context and experience in psychiatric diagnosis.
Psychopathology, 34, 192–196.
Klass, D. (1999). Developing a cross-cultural model of grief: The state of the field. Omega,
39, 153–178.
Klass, D., Silverman, P.  R., & Nickman, S.  L. (Eds.). (1996). Continuing bonds: New
understandings of grief. Washington, DC: Taylor & Francis.
38  Paul C. Rosenblatt
Lutz, C. (1985). Depression and the translation of emotional worlds. In Kleinman, A.,
& Good, B. (Eds.), Culture and depression (pp.  63–100). Berkeley: University of
California Press.
Malkinson, R., & Bar-Tur, L. (2000). The agony of grief: Parents’ grieving of Israeli sol-
diers. Journal of Personal and Interpersonal Loss, 5, 247–261.
Maschio, T. (1992). To remember the faces of the dead: Mourning and the full sadness of
memory in southwestern New Britain. Ethos, 20, 387–420.
Mezzich, J. E. (2008). Cultural formulation: Development and critical review. In Mezzich,
J. E., & Caracci, G. (Eds.), Cultural formulation: A reader for psychiatric diagnosis
(pp. 87–92). Lanham, MD: Jason Aronson.
Pratt, L. (1994). Business temporal norms and bereavement behavior. In Fulton, R., &
Bendiksen, R. (Eds.), Death and identity (3rd edn., pp. 263–287). Philadelphia: Charles
Press.
Redmond, A. (2008). Time wounds: Death, grieving and grievance in the Northern
Kimberly. In Glaskin, K., Tonkinson, M., Musharbash, Y., & Burbank, V. (Eds.),
Mortality, mourning and mortuary practices in indigenous Australia (pp.  69–86).
Burlington, VT: Ashgate.
Rohlof, H. (2008). The cultural interview in the Netherlands: The cultural formulation in
your pocket. In Mezzich, J. E., & Caracci, G. (Eds.), Cultural formulation: A reader for
psychiatric diagnosis (pp. 203–213). Lanham, MD: Jason Aronson.
Rosenblatt, P.  C. (2001). A social constructionist perspective on cultural differences in
grief. In Stroebe, M. S., Hansson, R. O., Stroebe, W., & Schut, H. (Eds.), Handbook of
bereavement research: Consequences, coping, and care (pp.  285–300). Washington,
DC: American Psychological Association Press.
Rosenblatt, P.  C. (2008). Recovery following bereavement: Metaphor, phenomenology,
and culture. Death Studies, 32, 6–16.
Shapiro, E. (1996). Family bereavement and cultural diversity: A social developmental
perspective. Family Process, 35, 313–332.
Small, N. (2001). Theories of grief: A critical review. In Hockey, J., Katz, J., & Small, N.
(Eds.), Grief, mourning and death ritual (pp. 19–48). Philadelphia: Open University
Press.
Small, N., & Hockey, J. (2001). Discourse into practice: The production of bereavement
care. In Hockey, J., Katz, J., & Small, N. (Eds.), Grief, mourning and death ritual
(pp. 97–124). Philadelphia: Open University Press.
Stearns, P. N. (1994). American cool: Constructing a twentieth century emotional style.
New York: New York University Press.
Stroebe, M., Gergen, M., Gergen, K., & Stroebe, W. (1996). Broken hearts or broken
bonds? In Klass, D., Silverman, P. R., & Nickman, S. L. (Eds.), Continuing bonds: New
understandings of grief (pp. 31–44). Washington, DC: Taylor & Francis.
Stroebe, M., & Schut, H. (1998). Culture and grief. Bereavement Care, 17(1), 7–11.
Stroebe, M., & Schut, H. (2005–2006). Complicated grief: A conceptual analysis of the
field. Omega, 52, 53–70.
Tafoya, N., & Del Vecchio, A. (2005). Back to the future: An examination of the Native
American holocaust experience. In McGoldrick, M., Giordano, J., & Garcia-Preto, N.
(Eds.), Ethnicity and family therapy (3rd edn., pp. 55–63). New York: Guilford.
Walter, T. (1999). On bereavement: The culture of grief. Philadelphia: Open University
Press.
Walter, T. (2005). What is complicated grief? A social constructionist perspective. Omega,
52, 71–79.
 CG: lessons from other cultures  39
Watters, E. (2010). Crazy like us: The globalization of the American psyche. New York:
Free Press.
Wellenkamp, J.  C. (1988). Notions of grief and catharsis among the Toraja. American
Ethnologist, 15, 486–500.
Wierzbicka, A. (2003). Emotion and culture: Arguing with Martha Nussbaum. Ethos, 31,
577–600.
Wikan, U. (1988). Bereavement and loss in two Muslim communities: Egypt and Bali
compared. Social Science and Medicine, 27, 451–460.
Wikan, U. (1990). Managing turbulent hearts: A Balinese formula for living. Chicago:
University of Chicago Press.
4 On achieving clarity regarding
complicated grief
Lessons from clinical practice
Therese A. Rando

Without trying to be humorous, complicated grief is quite complicated. Its causes,

forms, risk factors, comorbid conditions, associated elements, and treatment
requirements can differ markedly among individuals.
The purpose of this chapter is to elucidate complicated grief in adults through
lessons imparted from clinical practice. Focusing upon complicated grief as both
a distinct diagnostic entity as well as a clinical phenomenon, it is apparent that
some confusion and lack of clarity exists. After addressing several controversial
issues and the diversity of its forms, it is suggested that the field move toward
operationalizing a definition of complicated grief and establishing a generic com-
prehensive conceptual model for it.
Before proceeding, a small but important clarification is necessary. Elsewhere,
I argue strongly for the benefits of distinguishing grief from mourning and iden-
tify problems resulting when this is not done (Rando, 1993, 2012). Although I
wholeheartedly maintain these assertions regarding the importance of discrimi-
nating between these two terms, out of respect for this volume’s editors and for
easier communication within this particular book, I have temporarily adopted its
terminology. Consequently, for this chapter, I use grief to mean the reactions one
has to loss and grieve to refer to what one does over time to cope with it and its
consequences.

Where are we now regarding complicated grief?

Historically, professional understanding of complicated grief has suffered from
a lack of consensus regarding definition, different theoretical conceptualizations,
obscurity of criteria differentiating normal from complicated grief, and confu-
sion regarding issues of comorbidity (Stroebe et al., 2000). Further, imprecise
terminology, methodologically deficient studies, and missing operationalizations
held sway. Yet, despite these handicaps, there is still much that can be said about
complicated grief today.
In the last few decades, there has been a significant increase in research and
clinical investigation into complicated grief. A great amount of attention to it
swirls around the debate over including a complicated grief disorder in the next
edition of the Diagnostic and Statistical Manual of Mental Disorders (5th edn.;
 CG: lessons from clinical practice  41
DSM-5; American Psychiatric Association, forthcoming; currently due in 2013)
and, if so, what the diagnostic criteria should be. Two major proposals have been
offered by Horowitz and associates (Horowitz et al., 1997) and Prigerson and
colleagues (Prigerson et al., 1999).
The latter group has amassed the most empirical support for its proposed dis-
order, now termed prolonged grief disorder (PGD). As described by Prigerson,
Vanderwerker, and Maciejewski (2008), those with PGD are essentially stuck in a
state of chronic mourning characterized by significant separation distress marked
by intense yearning for the deceased and ruminative or intrusive thoughts about
the loss. Other associated symptoms may include difficulty accepting the death,
anger or bitterness over it, and personal traumatization and diminishment because
of it. In its wake, there is dysfunction in one or more important areas of life,
trouble trusting others, and difficulty moving on with life, which is perceived to
be empty and meaningless now and in the future.
Although there is impressive research documenting the usefulness of PGD (see
Prigerson et al., 2008, for a review), arguments for it are not without their share of
criticism (e.g., Hogan, Worden, & Schmidt, 2004; Rubin, Malkinson, & Witztum,
2008; Stroebe et al., 2000). It takes only a cursory review of the literature or a
superficial discussion with certain clinicians to see clearly that great passion and
extremely spirited debate are often generated by its proposed inclusion. I contend
that much of this comes from a lack of clarity about both PGD and complicated
grief.

Complicated grief as a distinct diagnostic entity

Reviewing the professional literature and listening to the self-reports of practition-
ers, it appears that the vast majority of clinicians in the United States, and many
others elsewhere, would heartily endorse a diagnostic category for complicated
grief. There have been many public calls for development of such a distinction
(e.g., Hartz, 1986; Horowitz, Bonanno, & Holen, 1993; Jacobs, 1993; Prigerson
& Jacobs, 2001; Rando, 1993), as well as countless private ones. Nevertheless,
concerns have been raised as to potential negative ramifications of such a category
(e.g., Rubin, Malkinson, & Witztum, 2008; Stroebe et al., 2000).
Despite some measure of agreement regarding the need for a DSM category for
complicated grief, there is significantly less consensus regarding what specifically
should compose it. Arguably, the main controversy within contemporary grief
literature arises around PGD, given that it is the predominant set of diagnostic cri-
teria proposed. It is also here where some significant mischaracterization occurs.

Regarding PGD as a diagnostic entity for complicated grief

Despite desires for a complicated grief disorder diagnosis, many are reluctant to
embrace PGD. They perceive the syndrome to be valid as far as it goes in describ-
ing a select group of mourners with attachment-related issues, but view it as being
too restrictive to be as useful as needed for a complicated grief disorder category.
42  Therese A. Rando
They see it as unsuitable for encompassing the myriad ways they know compli-
cated grief is manifested. In other words, they believe the syndrome accurately
pertains, but only for a subset of complicated grievers.
In the fray of the debate over the appropriateness of PGD for inclusion into the
DSM, it appears two errors are made repeatedly. First, there is misinterpretation
of what PGD actually represents. Second, there is overfocus upon it to the detri-
ment of fully appreciating the complexity of complicated grief and developing a
comprehensive conceptual model for it.

The fallacy of PGD as “equivalent” to complicated grief

In examining empirical and clinical literature regarding PGD, one more often than
not finds a grave mistake: There is an equation of PGD with complicated mourn-
ing. Yet equate means make equal and the reality is that complicated grief is more
than just PGD, thus not equivalent to it. This great equation myth, as I term it,
is not only inaccurate, but harmful. It is harmful because some people assume
that PGD is the sole way a mourner exhibits complicated grief. Should he or she
manifest something different, it may not be recognized as the complication in
bereavement that it, in fact, signals. Appropriate intervention may not take place
and inaccurate pathologization can occur. Rejection of this myth is the basis for
many people’s inability to accept PGD.
The fact is that the developers of PGD have put it forth as just one form of
complicated grief disorder. These originators have explicitly specified that “focus
. . . on PGD is not [italics added] intended to imply that this disorder is the only
[italics added], or even the primary [italics added], complication that may follow
from bereavement” (Prigerson et al., 2008, p. 173). However, far too many pro-
fessionals have mistaken PGD as equal to complicated grief as opposed to being
one representation of it.
Even experienced researchers and clinicians have equated the two. For exam-
ple, in a 2009 special article for World Psychiatry, “Grief and Bereavement: What
Psychiatrists Need to Know,” the respected researchers Zisook and Shear write
about complicated grief as “the current designation for a syndrome of prolonged
and intense grief . . . [in which] bereaved individuals . . . find themselves in a repeti-
tive loop of intense yearning and longing” (Zisook & Shear, 2009, pp. 67–69).
Here and in the rest of the article, it is clear the authors are referring to PGD. They
make statements about complicated grief, yet describe PGD specifically. And they
do this even though they include in that same article a discussion of grief-related
major depressions, with such depressions identified as a complication of bereave-
ment. In other words, although Zisook and Shear equate complicated grief with
PGD, they actually disprove that very point once they discuss grief-related major
depression, since that indicates ipso facto that complicated grief has other forms
as well. PGD simply cannot be equivalent to complicated grief since there is obvi-
ously more to complicated grief than just PGD.
Zisook and Shear are by no means the only ones to make this equation and to
fail to mention that there are other forms besides it. A brief survey of empirical
 CG: lessons from clinical practice  43
studies addressing PGD reveals that this happens fairly often. The findings of
such researchers are not being questioned here. The sole issue is their not distanc-
ing themselves from the great equation. Researchers and writers who permit PGD
to represent all of complicated grief – as opposed to just a single form of it – are
perpetuating an untruth and doing a disservice. Simple phrases such as “PGD
is one form of complicated grief” or “As one example of complicated grief, we
examined PGD” would clearly convey that complicated grief can be manifested
in other ways. This could go a long way in defusing the source of much conflict
regarding PGD.
The bottom line is that we must accurately appreciate what PGD is and what
it is not. What it is is a distilled set of grief symptoms that reliably and accurately
predict enduring distress and dysfunction (H. Prigerson, personal communication,
February 4, 2011). It is just one of a number of forms of complicated grief, not the
only one. (Others will be discussed below.) It is also an excellent example of what
today’s research can reveal. However, it does not represent all of complicated
grief, and cannot legitimately be equated to it. The plain truth – recognized by
the proposers of PGD but not necessarily by those who have supported them – is
that plenty of people with complicated grief do not have PGD. To only look at
complicated grief in that one fashion is to eliminate a huge proportion of those
with complications in their bereavement. The subtype must not be mistaken for
the overarching category.

Complicated grief as a clinical phenomenon

We can learn much about complicated grief from clinical practice, which helps
illustrate that it is more complex than could be encompassed by just one presenta-
tion (e.g., PGD). Below, I discuss what clinical practice reveals about its forms
and functions, after first addressing a controversial issue.

What about discrepancies between clinical observation and empirical

investigation?
In several areas, discrepancies have been identified between what has been reported
in the clinical literature about complicated grief and what has been demonstrated
empirically (e.g., Bonanno, Boerner, & Wortman, 2008). This seems particularly
relevant regarding treatment efficacy, the existence of certain syndromes, and the
predictive validity of various risk factors. In each domain, clinical observation
has sometimes come to conclusions that have not been supported empirically (at
least not yet). What does this mean? Should we cast aside clinical observations
made repeatedly through the years by solid clinicians because it conflicts with or
is unsupported by empirical studies? Yet, is it not true that many clinicians had
been operating according to some “myths” we now know empirically to be untrue
(see Wortman & Silver, 2001)?
In this age of emphasis on empiricism, we ought not dismiss what clinical
practice has taught, but must not accept it uncritically either. For the benefit of the
44  Therese A. Rando
bereaved, we must strive to find ways to integrate the best of both ways of know-
ing. To do so, it helps to understand why such differences exist. I would suggest
that one explanation stems from the sometimes discrepant foci of clinicians and
researchers. These not only give each group different viewpoints, but can yield
apparently differing conclusions. (Although I speak of differences here, it is with
awareness that some are both clinician and researcher, as well as that both groups
share a common goal to understand and assist in bereavement.)
Let us look at treatment efficacy as an example. In general, clinicians writing
about their observations of complicated grief contend with individuals who are
self-referred, seen relatively late in bereavement, and present for tertiary preven-
tion intervention – all of which factors are associated with more effective treat-
ment outcome (Schut & Stroebe, 2005). Further, this population is often seen in
individual treatment over a period of time. Clinicians have seen treatment work
with this population they serve, and have written about complicated grief from
that perspective.
In contrast, bereavement researchers typically deal with another group, which
includes not only the aforementioned grievers clinicians typically see, but others
as well. I submit that these others, outside the subset seen by clinicians, are the
source for most of the discrepancies. As reported in a meta-analysis of bereave-
ment outcome studies by Neimeyer and Currier (2009), these other grievers were
not self-referred (they were solicited or outreached); were not specifically indi-
cated for therapy (they were not complicated grievers, but merely people who
had suffered loss or were at risk for problems); and received interventions often
quite different from those delivered by clinicians, which transpired in relatively
time-limited formats, after a variety of recruitments (i.e., aggressive outreach
and media and community advertising were included), and within group sessions
lacking a therapy focus.
It is not surprising that researchers’ conclusions vary at times from their clini-
cal colleagues, given that they have essentially been dealing with two different
areas of focus within the same population; clinicians tend in general to deal with a
subset of the grief spectrum (i.e., those with complications), whereas researchers
in general deal with the entire spectrum (i.e., those with and without complica-
tions). This is why research is specifically mandated with complicated grievers,
in order to see whether the clinical observations derived from work with them are
empirically supported.
Three other factors might explain some of the inconsistencies sometimes
found between clinical practice and research findings: equating measures of dis-
tress with complicated grief, failure to get beyond merely looking at symptoms to
evaluate underlying grief processes, and overgeneralization of clinicians’ beliefs.
Amount of distress does not always portend complications in grief processes.
Although there certainly is a correlation between distress and complicated grief,
one can be in great pain on some measures – especially after the loss of a healthy,
close, and warm relationship – but still move on to grieve healthily. Currently, we
tend to measure symptoms of distress, but not underlying grief processes. Thus,
in the exemplary Changing Lives of Older Couples Study, Carr (2008) found
 CG: lessons from clinical practice  45
increased yearning and anxiety after the death of a spouse who had been confi-
dante and helpmate. She felt this challenged the longstanding clinical observa-
tion that those losing spouses after conflicted marriages had more complications.
However, this could well be an artifact of looking at symptoms only and not grief
processes as well. In other words, it is only natural that there would be more
distress when losing a close and warm relationship as opposed to losing a strained
one. The more relevant issue is whether that actually translated into complicated
grief processes because of it.
To my knowledge, Carr inferred complicated grief by the symptoms that were
present, and not by any measurement of the griever’s engagement in grieving
processes per se. I would contend that one can be enormously distressed by the
loss of a spouse yet, because of not having the conflictual issues a person with a
strained marriage would have, could move forward to grieve without complica-
tions. This is because grieving is not just about having symptomatic reactions,
but about what one does to cope with and adapt to the loss that stimulated those
reactions, as well as managing those reactions themselves.
Unfortunately, this was unable to be empirically tested because we currently
lack sufficient ability to measure grief processes per se – in fact, as observed below,
we have not even agreed upon what they might be – and only have measures of
symptoms. This speaks to the need for our field to take steps to operationalize
grieving, develop instruments to assess it, and not rely so heavily on assessment
of symptoms exclusively. Further, it underscores the necessity to better differenti-
ate conceptually between acute grief reactions to loss and processes undertaken
to cope with and adapt to that loss (see Rando, 1993, 2012). Too often they are
confounded.
Finally, we sometimes see discrepancies because some people have mischar-
acterized clinicians’ assumptions. They overlook clinicians’ focus on the idiosyn-
cratic situation of the griever, and mistakenly portray them as holding absolute and
universal expectations (e.g., “Everyone has to express great emotion”), which, of
course, is easily disproved and taken as clinical error. In fact, wise clinicians do
not maintain such universal views. Rather, the issue is: Are particular grievers,
with their own idiosyncratic factors operating, expressing whatever there is they
need to express, how they need to express it, and when? Their perspective flows
from meeting the needs of the griever, not from a universal expectation that eve-
ryone has to express great emotion. Those criticizing clinical wisdom often fail to
appreciate this, overstate what many clinicians believe, and then claim to debunk
“myths” that many do not hold.

Four forms of complicated grief

In clinical practice, we find that complicated grief manifests in a variety of forms.
This coincides with the conclusion that “complicated grief is not a single syndrome”
(Stroebe, Hansson, Schut, & Stroebe, 2008, p. 6) and echoes the diversity of clini-
cal presentations found in the literature (for reviews, see Bowlby, 1980; Jacobs,
1993; Middleton, Raphael, Martinek, & Misso, 1993; Parkes & Weiss, 1983;
46  Therese A. Rando
Rando, 1993; Raphael, 1983). Synthesizing the literature regarding outcomes of
loss, Rando (1993) identified four different types of presentation for complicated
grief: symptoms, syndromes, diagnosable mental or physical disorders, and death.
Recognizing that not all adversity subsequent to a death stems directly from
it, and to avoid false positives, Rando (2003) specified two criteria for presenta-
tions to be indicative of complicated grief: They must have developed or signifi-
cantly worsened since the death in question, and must be associated with some
compromise, distortion, or failure of some normal grief process(es). When one
or both criteria are absent, the manifestations cannot be construed as a type of
complicated grief (e.g., they may be comorbid with but independent of it, such as
depression from unrelated problems).

Complicated grief symptoms

John’s wife died of breast cancer at 33. It took John 8 years, one divorce,
and two therapists before he realized that he was furious at his God for
taking his wife and at himself for being unable to protect her. In the interim,
John had complicated grief that manifested through extreme death anxiety
and a pattern of overreaction to experiences entailing separation and loss.

In this form of complicated grief, the mourner experiences some psychologi-

cal, behavioral, social, or physical symptoms of distress, disability, dysfunction,
pathology, or loss of freedom. Although inadequate in terms of number, intensity,
duration, or type to meet criteria for any of the other three forms of complicated
grief, these symptoms accompany a compromise, distortion, or failure in one or
more of the normal processes of grief. For a list of symptoms constituting clinical
indicators of complicated grief, see Rando (1993). To my knowledge, complicated
grief symptoms as a form of complicated grief per se have not been subjected to
empirical study yet.

Complicated grief syndromes

Jacqueline’s estranged husband and three friends were on a fishing trip

when their boat capsized. No bodies were found. After a period of shock,
she commenced 9 years of “floating” through life, feeling disconnected,
chronically bewildered and confused, going from job to job and relationship
to relationship, anxious and always awaiting another loss to befall her.

Complicated grief symptoms can coalesce into any of eight complicated grief
syndromes. Seven of them are identified in the clinical literature: absent grief
 CG: lessons from clinical practice  47
(Deutsch, 1937; Bowlby, 1980); delayed grief (e.g., Raphael, 1983); inhibited
grief (e.g., Raphael, 1983); distorted grief (of the extremely angry and guilty
types; Raphael, 1983); conflicted grief (Parkes & Weiss, 1983); unanticipated
grief (Parkes & Weiss, 1983); and chronic grief (Bowlby, 1980; Parkes & Weiss,
1983). The eighth syndrome is PGD, which has been empirically derived as cited
above and is described as a type of chronic grief.
Empirical investigation of PGD has been mentioned earlier. Only some of the
other syndromes have been researched, and even then minimally. The literature
“indicates that minimal or absent grief reactions are very prevalent, whereas
delayed grief reactions are quite rare” (Bonanno et al., 2008, p. 290). According
to Stroebe et al. (2008), “chronic (or prolonged) grief has been generally well
accepted as a pathological category, whereas delayed, inhibited, and absent grief
remain the subject of much debate” (p. 8). Despite this, chronic and absent grief
have been well explained by the dual process model of coping with bereave-
ment (DPM; Stroebe & Schut, 2010). To the extent that the unanticipated grief
syndrome is associated with traumatic bereavement, there has been more research
(for reviews, see Armour, 2006; Pearlman, Wortman, Feuer, Farber, & Rando, in
press).

Diagnosable mental or physical disorder

At 25 years old, Carl came to therapy at the suggestion of family concerned

by his increasing sadness. He had dropped out of a stellar sophomore year
in college, following the death of his girlfriend. Afterwards, he held a series
of low-level, dead-end jobs before moving back with his parents. Now, Carl
meets diagnostic criteria for a major depressive disorder.

Without warning, Gisele’s husband informed her he had been having an

affair. Furious because she did not “understand,” he drove straight into a
tree, killing himself instantly. Gisele grieved little and demonstrated few
changes in behavior. Eleven months later, she was referred for psycho-
therapy by a physician alarmed by her development of severe colitis.

Investigations into morbidity after loss consistently show that bereavement can
cause much suffering, associated with severe consequences to health and well-
being (Stroebe, Schut, & Stroebe, 2007). It has been axiomatic in the field since
the classic Institute of Medicine Study that bereaved individuals are at increased
risk for physical and mental illness (Osterweis, Solomon, & Green, 1984). Recent
long-term studies corroborate this (e.g., Jones, Bartrop, Forcier, & Penny, 2010).
In terms of complicated grief specifically, research into PGD (and forerunners)
found it associated with heightened risk of both mental and physical impairments
48  Therese A. Rando
(Prigerson et al., 2008). When complicated grief is examined in terms of mental
health, there have been remarkably consistent findings that it can result in psy-
chiatric disorders, particularly along the depressive and anxiety spectrums (e.g.,
Middleton et al., 1993; Raphael, Minkov, & Dobson, 2001; Stroebe, Schut, &
Stroebe, 2007).

Death

Giorgio’s father impulsively killed himself when it was discovered he had

embezzled his client’s money to cover gambling debts. At 34, Giorgio
immediately moved back to his hometown with his own family to care for
his bereaved mother. Wracked with shame and rage, he began to drink.
Fifteen years later, Giorgio died in an alcohol-related one-car crash.

Complicated grief can be manifested in death that is consciously chosen (i.e.,

suicide). This coincides with research identifying it as a risk factor for completed
suicide (Luoma & Pearson, 2002) and suicidality (Latham & Prigerson, 2004).
Further, death resulting from complicated grief-related behavior can be subin-
tended or unintended (Rando, 1993), stemming from the griever’s behavior (e.g.,
car crash from drunk driving, as illustrated above, or neglect of health care).

Two functions of complicated grief

Clinical practice suggests two functions served by complicated grief: avoidance
and connection (Rando, 1993). Given the variability among grievers and their
situations, these may be the only elements consistent across all types of compli-
cated grief.

In all forms of complicated [grief], the mourner attempts to do two things: (a)
deny, repress, or avoid aspects of the loss, its pain, and the full realization of
its implications for the mourner and (b) hold on to and avoid relinquishing the
lost loved one. These attempts, or some variation thereof, are what underlie
and cause complications in the . . . processes of [grief]. (Rando, 1993, p. 149)

There is much in the literature corroborating this. Numerous researchers

have documented the critical role of avoidance in complicated grief, including
Parkes (1987), who identified it as being evident in all kinds of “atypical” grief
he encountered, as well as numerous others who pinpoint avoidance as a central
feature associated with complicated grief (e.g., Boelen & van den Bout, 2010;
Shear, 2010; Stroebe, Boelen, et al., 2007).
The griever’s holding on to the deceased (also known as maintaining con-
tinuing bonds) has been the focus of much discussion and research (e.g., Boelen,
Stroebe, Schut, & Zijerveld, 2006; Field & Filanosky, 2010; Rubin, 1999). There
 CG: lessons from clinical practice  49
have been enough findings confirming associations with complicated grief to jus-
tify further investigations to identify under which conditions continuing bonds is
adaptive and under which conditions it is not.

Next steps
Current knowledge regarding complicated grief suggests two next steps to clarify
it and further its development: operationalizing its definition and developing a
comprehensive conceptual model for it.

An operationalized definition
A very surprising fact about complicated grief is how many definitions of it are
remarkably non-specific. A well-regarded definition illustrates this below, pre-
sented with one for “grief,” upon which it rests:

Grief is the term applied to the primarily emotional (affective) reaction to the
loss of a loved one through death . . . it also incorporates diverse psychologi-
cal . . . and physical . . . manifestations. (Stroebe et al., 2008, p. 5)

Complicated grief . . . [is] a clinically significant deviation from the (cultural)
norm . . . in either (a) the time course or intensity of specific or general symp-
toms of grief and/or (b) the level of impairment in social, occupational, or
other important areas of functioning. (p. 7)

Both definitions describe, but do not operationalize. It is clear that complicated

grief constitutes a derailing of normal grief, but exactly what is being derailed is
unclear. First and foremost, this leaves the field of grief relatively confused about
the overly abstract object of our attention. Without explicitly delineated ways
hypothesized to actualize its functioning and delineate its processes, how can we
as a field accurately communicate, describe, compare, contrast, treat, and research
grieving that is complicated? How do we know what it entails or if we are refer-
ring to the same thing? It seems unscientific to be defining the phenomenon in
such vague and nebulous terms. Complicated grief begs to be concretized not only
so we can ensure we are all talking about the same situation, but so it can be put
to empirical investigation. Understanding of complicated grief and, ultimately,
assistance to the bereaved are fundamentally best served by an accurate definition
identifying its hypothesized underlying processes in testable terms.
There are other reasons to move toward operationalizing complicated grief.
Missing an operationalized delineation of what is hypothesized as necessary to
grieve healthily, clinicians lack a framework for specifically what to promote and/
or intervene in to transform complicated into uncomplicated grief. Researchers
lack definitive elements to investigate.
Furthermore, abstract definitions mean clinicians operate without clear-cut
criteria against which to evaluate a person’s grieving status. This knowledge of
where a person is relative to hypothesized processes underlying healthy grieving
50  Therese A. Rando
is one of three basic areas to consider to appreciate any griever’s responses
properly. The others are the 43 sets of factors influencing grief reactions (psy-
chological, social, and physical; Rando, 2012) and current levels of functioning.
Such multidimensional determinants are in line with Middleton et al.’s (1993)
assessment that “Grief is being viewed increasingly as a complex and evolving
process, requiring the use of a multidimensional model” (p. 60) and support their
prognostication: “It seems unlikely that pathological grief will become a unitary
concept. Instead, future research will likely adopt a multidimensional framework”
(ibid.).
Precisely because reactions are so relative, multiply determined, and idiosyn-
cratic, complicated grief cannot be determined merely by presence or absence
of specific symptoms alone. We know that symptoms indicate complicated grief
only when we take them into account given the situation of this person, dealing
with this loss, at this time, under these circumstances, and then evaluate them
in light of where that person is in relation to hypothesized processes of healthy
grieving. A particular symptom present in concert with healthy grieving is not
necessarily indicative of complicated grief, but the same one in the absence of
healthy grieving could be. The clinical question should be: Is the griever, given
the context of his or her own unique grief situation, doing what is necessary to
grieve healthily?
Obviously, to answer this question it is necessary to operationalize what is
hypothesized as required to grieve healthily. There have been notable attempts
to do this, including Stroebe and Schut’s (1999) DPM, Worden’s (2009) four
tasks, and Rando’s (1993, 2012) six “R” processes. These last are: Recognize the
loss; React to the separation; Recollect and re-experience the deceased and the
relationship; Relinquish the old attachments to the deceased and the old assump-
tive world; Readjust to move adaptively into the new world without forgetting
the old; and Reinvest. They are employed here to exemplify how a complicated
grief definition potentially could be operationalized: Complicated grief is present
whenever, considering the idiosyncratic situation of the griever and his or her
current functioning, there is some compromise, distortion, or failure of one or
more of the six “R” processes of grieving.
Such a definition permits consideration of the multidimensional elements in
the griever’s idiosyncratic situation, plus evaluation of that person in relation to
a standard set of hypothesized operationalized grief processes that can be recog-
nized, measured, and manipulated.

A comprehensive conceptual model

A subsequent step in developing complicated grief is to construct a generic com-
prehensive conceptual model for it. Rubin, Malkinson, and Witztum (2008) have
argued that the multidimensional complexity of complicated grief demands recog-
nition of an array of responses, not the restricted set associated with PGD. Further,
any model must be robust enough to encompass the varieties of complicated grief
 CG: lessons from clinical practice  51
presentations. Despite insufficient empirical backing at this point, there is simply
too much clinical history and support to dismiss non-PGD forms of complicated
grief (e.g., delayed or inhibited), which “make considerable theoretical and clini-
cal sense . . . and would seem to merit inclusion [within a spectrum of complicated
grief]” (Stroebe & Schut, 2005–2006, p. 66).
It appears that a generic comprehensive conceptual model of complicated grief
should account for eight elements: its idiosyncratic nature; explanatory theories;
the coping driving it; integral issues and current functioning associated with it;
its functions; its forms; and hypothesized operationalized processes. Although it
would be premature to consider proposing other DSM diagnoses of complicated
grief to expand its scope beyond PGD, it is not premature to start discussing models
incorporating such elements. They could serve as springboards for research that
ultimately would bring such diagnoses into existence. Along the way, they could
aid thanatologists in appreciating relationships between all types of bereavement
complications, their etiologies, and their presentations, while providing research-
ers with specific variables to investigate.
I believe that numerous advantages would accrue from such a model. Primarily,
it would avoid the reductionism that has obfuscated the complexity of compli-
cated grief. Multiple influencing factors would be considered (e.g., attachment
style, culture, death characteristics); bereavements with and without trauma could
be incorporated; complicated grief displaying as mental and/or physical disorders
would be contained; subclinical manifestations would not be overlooked; diverse
theoretical conceptualizations could offer explanations (e.g., psychoanalytic,
attachment, trauma/stress, constructivist); and associated integral issues would be
included (e.g., relationship with the deceased, meaning making, resource deficien-
cies). Importantly, it would recognize different routes to the common pathway of
grieving process complications. So, for example, complicated grief signaled by
inability to adapt could be due to any number of things (not solely attachment
problems per PGD), such as personal traumatization secondary to the death cir-
cumstances, social disenfranchisement, excessive guilt, dynamics of child loss,
or failure to make meaning, among other possibilities. Such crucial stimulants of
complicated grief would be given their due and could be targeted for treatment
and research.

Conclusion
To achieve currently needed clarity on complicated grief, it appears necessary to
appreciate it as both a distinct diagnostic category and a clinical phenomenon;
grasp the realities of PGD; understand complicated grief’s forms and functions;
operationalize its definition; and develop a comprehensive conceptual model for
it. Elements of such a model that might provide more lucidity are put forth here for
discussion and research. Others are invited to do similarly. Hopefully, such actions
stimulate meaningful discourse and research in the field and add to the understand-
ing and treatment of the richly complex experience of complicated grief.
52  Therese A. Rando

Bibliography
American Psychiatric Association. (forthcoming). Diagnostic and statistical manual of
mental disorders (5th edn.). Washington, DC: Author.
Armour, M. (2006). Violent death: Understanding the context of traumatic and stigmatized
grief. Journal of Human Behavior in the Social Environment, 14, 53–90.
Boelen, P., & van den Bout, J. (2010). Anxious and depressive avoidance and symptoms of
prolonged grief, depression, and post-traumatic stress disorder. Psychologica Belgica,
50, 49–67.
Boelen, P., Stroebe, M., Schut, H., & Zijerveld, A. (2006). Continuing bonds and grief: A
prospective analysis. Death Studies, 30, 767–776.
Bonanno, G., Boerner, K., & Wortman, C. (2008). Trajectories of grieving. In Stroebe,
M., Hansson, R., Schut, H., & Stroebe, W. (Eds.), Handbook of bereavement research
and practice: Advances in theory and intervention (pp.  287–307). Washington, DC:
American Psychological Association.
Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and depression. New York:
Basic Books.
Carr, D. (2008). Factors that influence late-life bereavement: Considering data from the
Changing Lives of Older Couples Study. In Stroebe, M., Hansson, R., Schut, H., &
Stroebe, W. (Eds.), Handbook of bereavement research: Advances in theory and inter-
vention (pp. 417–440). Washington, DC: American Psychological Association.
Deutsch, H. (1937). Absence of grief. Psychoanalytic Quarterly, 6, 12–22.
Field, N., & Filanosky, C. (2010). Continuing bonds, risk factors for complicated grief, and
adjustment to bereavement. Death Studies, 34, 1–29.
Hartz, G. (1986). Adult grief and its interface with mood disorder: Proposal of a new diag-
nosis of complicated bereavement. Comprehensive Psychiatry, 27, 60–64.
Hogan, N., Worden, J., & Schmidt, L. (2004). An empirical study of the proposed compli-
cated grief disorder criteria. Omega, 48, 263–277.
Horowitz, M., Bonanno, G., & Holen, A. (1993). Pathological grief: Diagnosis and expla-
nation. Psychosomatic Medicine, 55, 260–273.
Horowitz, M., Siegel, B., Holen, A., Bonanno, G., Milbrath, C., & Stinson, C. (1997).
Diagnostic criteria for complicated grief disorder. American Journal of Psychiatry, 154,
904–910.
Jacobs, S. (1993). Pathologic grief: Maladaptation to loss. Washington, DC: American
Psychiatric Press.
Jones, M., Bartrop, R., Forcier, L., & Penny, R. (2010). The long-term impact of bereave-
ment upon spouse health: A 10-year follow-up. Acta Neuropsychiatrica, 22, 212–217.
Latham, A., & Prigerson, H. (2004). Suicidality and bereavement: Complicated grief as
psychiatric disorder presenting greatest risk for suicidality. Suicide and Life-Threatening
Behavior, 34, 350–362.
Luoma, J., & Pearson, J. (2002). Suicide and marital status in the United States, 1991–
1996: Is widowhood a risk factor? American Journal of Public Health, 92, 1518–1522.
Middleton, W., Raphael, B., Martinek, N., & Misso, V. (1993). Pathological grief reactions.
In Stroebe, M., Stroebe, W., & Hansson, R. (Eds.), Handbook of bereavement: Theory,
research, and intervention (pp. 44–61). New York: Cambridge University Press.
Neimeyer, R., & Currier, J. (2009). Grief therapy: Evidence of efficacy and emerging direc-
tions. Current Directions in Psychological Science, 18, 352–356.
Osterweis, M., Solomon, F., & Green, M. (Eds.). (1984). Bereavement: Reactions, conse-
quences, and care. Washington, DC: National Academy Press.
 CG: lessons from clinical practice  53
Parkes, C. (1987). Bereavement: Studies of grief in adult life (2nd edn.). Madison, CT:
International Universities Press.
Parkes, C., & Weiss, R. (1983). Recovery from bereavement. New York: Basic Books.
Pearlman, L., Wortman, C., Feuer, C., Farber, C., & Rando, T. (in press). Treating traumatic
bereavement: Intervening with survivors of sudden death. New York: Guilford Press.
Prigerson, H., & Jacobs, S. (2001). Traumatic grief as a distinct disorder: A rationale, con-
sensus criteria, and a preliminary empirical test. In Stroebe, M., Hansson, R., Stroebe,
W., & Schut, H. (Eds.), Handbook of bereavement research: Consequences, coping,
and care (pp. 613–645). Washington, DC: American Psychological Association.
Prigerson, H., Shear, M., Jacobs, S., Reynolds, C., Maciejewski, P., Davidson, J., et al.
(1999). Consensus criteria for traumatic grief: A preliminary empirical test. British
Journal of Psychiatry, 174, 67–73.
Prigerson, H., Vanderwerker, L., & Maciejewski, P. (2008). A case for inclusion of pro-
longed grief disorder in DSM-V. In Stroebe, M., Hansson, R., Schut, H., & Stroebe,
W. (Eds.), Handbook of bereavement research and practice: Advances in theory and
intervention (pp. 165–186). Washington, DC: American Psychological Association.
Rando, T. (1993). Treatment of complicated mourning. Champaign, IL: Research Press.
Rando, T. (2003). Public tragedy and complicated mourning. In Lattanzi-Licht, M., & Doka,
K. (Eds.), Living with grief: Coping with public tragedy (pp. 263–274). Washington,
DC: Hospice Foundation of America.
Rando, T. (2012). Coping with the sudden death of your loved one: Self-help for traumatic
bereavement. Indianapolis, IN: Dog Ear Publishing.
Raphael, B. (1983). The anatomy of bereavement. New York: Basic Books.
Raphael, B., Minkov, C., & Dobson, M. (2001). Psychotherapeutic and pharmacologi-
cal intervention for bereaved persons. In Stroebe, M., Hansson, R., Stroebe, W., &
Schut, H. (Eds.), Handbook of bereavement research: Consequences, coping, and care
(pp. 587–612). Washington, DC: American Psychological Association.
Rubin, S. (1999). The two-track model of bereavement: Overview, retrospect and prospect.
Death Studies, 23, 681–714.
Rubin, S., Malkinson, R., & Witztum, E. (2008). Clinical aspects of a DSM complicated
grief diagnosis: Challenges, dilemmas, and opportunities. In Stroebe, M., Hansson,
R., Schut, H., & Stroebe, W. (Eds.), Handbook of bereavement research and prac-
tice: Advances in theory and intervention (pp. 187–206). Washington, DC: American
Psychological Association.
Schut, M., & Stroebe, M. (2005). Interventions to enhance adaptation to bereavement.
Journal of Palliative Medicine, 8, S140–S147.
Shear, M. (2010). Exploring the role of experiential avoidance from the perspective of
attachment theory and the dual process model. Omega, 61, 357–369.
Stroebe, M., Boelen, P., van den Hout, M., Stroebe, W., Salemink, E., & van den Bout, J.
(2007). Ruminative coping as avoidance: A reinterpretation of its function in adjust-
ment to bereavement. European Archives of Psychiatry and Clinical Neuroscience,
257, 462–472.
Stroebe, M., Hansson, R., Schut, H., & Stroebe, W. (2008). Bereavement research:
Contemporary perspectives. In Stroebe, M., Hansson, R., Schut, H., & Stroebe, W.
(Eds.), Handbook of bereavement research and practice: Advances in theory and inter-
vention (pp. 3–25). Washington, DC: American Psychological Association.
Stroebe, M., & Schut, H. (1999). The dual process model of coping with bereavement:
Rationale and description. Death Studies, 23, 1–28.
Stroebe, M., & Schut, H. (2005–2006). Complicated grief: A conceptual analysis of the
field. Omega, 52, 53–70.
54  Therese A. Rando
Stroebe, M., & Schut, H. (2010). The dual process model of coping with bereavement: A
decade on. Omega, 61, 273–289.
Stroebe, M., Schut, H., & Stroebe, W. (2007). Health outcomes of bereavement. The
Lancet, 370, 1960–1973.
Stroebe, M., van Son, M., Stroebe, W., Kleber, R., Schut, H., & van den Bout, J. (2000).
On the classification and diagnosis of pathological grief. Clinical Psychology Review,
20, 57–75.
Worden, J. (2009). Grief counseling and grief therapy (4th edn.). New York: Springer.
Wortman, C., & Silver, R. (2001). The myths of coping with loss revisited. In Stroebe,
M., Hansson, R., Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement
research: Consequences, coping, and care (pp. 405–429). Washington, DC: American
Psychological Association.
Zisook, S., & Shear, K. (2009) Grief and bereavement: What psychiatrists need to know.
World Psychiatry, 8, 67–74.
5 On the nature and prevalence of
uncomplicated and complicated
patterns of grief
Kathrin Boerner, Anthony D. Mancini, and
George Bonanno

Complicated grief has been variously defined, but most theorists understand it as
a form of grief characterized by persistent, intense longing and yearning for the
deceased (separation distress), intrusive thoughts or images, emotional numbness,
anger or guilt related to the loss, a sense of emptiness, and reactivity in response
to cues (e.g., Horowitz et al., 1997; Prigerson, 2004; Shear, Frank, Houck, &
Reynolds, 2005). At the same time, persons with complicated grief often avoid
people and places they associate with the loss, because of the intense distress those
reminders evoke. Thus, complicated grief tends to involve a vacillation between
an anxious preoccupation with and an avoidance of memories of the deceased. In
addition, complicated grievers commonly have difficulty redefining themselves
(Mancini & Bonanno, 2006), and often experience difficulties forming satisfying
new relationships or engaging in potentially rewarding activities.
To understand complicated grief, it is first necessary to be familiar with the
distinction between complicated and uncomplicated forms of grief. Although
there are at this point some defined criteria for what constitutes complicated grief,
such definitions are largely lacking for grief patterns that are normative and do
not warrant a complicated grief diagnosis. Only one pattern has traditionally been
considered as normal grief (a period of distress followed by recovery). Variations
from this response pattern, such as consistent minimal distress following bereave-
ment, were suspected to be pathological in nature (i.e., inhibited or absent grief)
and most likely to result in another version of pathological grief: delayed grief.
These traditional ways of thinking about grief were heavily influenced by
clinical observation and a very limited empirical database. Our understanding of
grief has increased tremendously following large-scale studies with prospective
longitudinal data that include data from before the death to years after the loss,
allowing the full range of possible grief patterns to emerge. The purpose of this
chapter is to delineate different pathways of grieving that to date have been well
established empirically, as well as describing how these pathways have revolu-
tionized our understanding of different types of grief response.

Historical background
The earliest taxonomies of individual differences in grief reaction were based
primarily on clinical observation or on data sets based on psychiatric samples. Not
56  Kathrin Boerner et al.
surprisingly, the earliest models of bereavement outcome focused primarily on
the distinction between normal and abnormal or pathological forms of grieving.
Using these models, bereavement scholars pondered the question of what consti-
tutes a normal grief course. In addition, they focused attention on the possible role
played by avoidant or defensive processes in delaying the onset of grief.
One of the earliest comparative descriptions of normal and pathological forms
of grieving came from Parkes’s (1965) ground-breaking study of bereaved psy-
chiatric patients. Parkes distinguished three types of pathological grief reaction:
chronic grief, inhibited grief, and delayed grief. Bowlby (1980) later echoed this
taxonomy to propose disordered forms of mourning that could be arrayed along
a single conceptual dimension. Anchoring one end of the continuum was chronic
mourning. At the other end, Bowlby placed the prolonged absence of conscious
grieving (p.  138). He maintained that those showing an absence of conscious
grieving “may appear to be coping splendidly” (p. 153) but are often tense and
short-tempered, with tears just below the surface. Bowlby believed that physi-
cal symptoms (e.g., headaches and heart palpitations) were also common in this
group, and that, sooner or later, many of those who consciously avoid grieving
become depressed, often in response to a subsequent, more minor loss.
Based on the evidence available at the time, Raphael (1983) also proposed
a number of “morbid or pathological patterns of grief” (p. 59). These included
chronic, unresolved grief reactions as well as the absence of grief in which “the
grieving affects or mourning process may be totally absent, partially suppressed,
or inhibited” (p. 60). Like Bowlby, she noted that some bereaved people seem to
cope remarkably well, and often carry on “as if nothing had happened.” Although
she acknowledged that such responses “may be seen as evidence of strength and
coping by many” (p. 205), she too argued that in most cases they were actually
markers of psychopathology.
In 1984, the Institute of Medicine released a report summarizing the state-of-
the-art knowledge about bereavement. The report concluded that the death of a
loved one produced a “near universal occurrence of intense emotional distress
.  .  . with features similar in nature and intensity to those of clinical depres-
sion” (Osterweis, Solomon, & Green, 1984, p.  18). The report also concluded
that “absent grief” was a “pathological” form of mourning that “represents
some form of personality pathology” and that “persons who show no evidence
of having begun grieving” should receive “professional help” (p.  65). Several
years later, Middleton, Moylan, Raphael, Burnett, and Martinek (1993) surveyed
an international sample of researchers, theorists, and clinicians working in the
field of bereavement. A compelling majority of these experts endorsed the idea
that “absent grief” was a pathological grief reaction that usually stemmed from
denial or inhibition of the normal grief reaction. This response was almost always
viewed as maladaptive in the long run.
But is this really the case? When people experience relatively mild or short-
lived grief reactions, should this be considered atypical or pathological? Wortman
and Silver (1989) noted that there was no convincing empirical evidence to sup-
port this assertion. More recently, Bonanno and colleagues (e.g., Bonanno, 2004,
 Uncomplicated and complicated patterns of grief  57
2005) have argued that many bereaved people show a clear resilience in the face
of loss. At the other end of the spectrum, it might also be questioned whether
psychopathology observed during bereavement should always be interpreted as
an abnormal grief reaction. Might not at least some of the chronic dysfunction
be attributed to an enduring emotional disturbance that pre-dates the loss? In the
following sections, we review the evidence regarding the prevalence of different
patterns of grief, with a particular focus on evidence for the experience of intense
distress following the death of a loved one, as well as for delayed grief and chro-
nicity in poor adjustment to bereavement.

Grief trajectories
Among people who have faced the loss of a loved one, is it true that distress is
commonly experienced? Will distress or depression emerge at a later date among
those who fail to exhibit distress in the first several weeks or months follow-
ing the loss? We identified several studies that provide information bearing on
these questions. Most of these studies focused on the loss of a spouse (Boerner,
Wortman, & Bonanno, 2005; Bonanno & Field, 2001; Bonanno, Keltner, Holen,
& Horowitz, 1995; Bonanno, Moskowitz, Papa, & Folkman, 2005; Bonanno et
al., 2002; Lund et al., 1985–1986; Vachon et al., 1982; Zisook & Shuchter, 1986),
with several of these examining response to loss following a time of caregiv-
ing for a chronically ill loved one (Aneshensel, Botticello, & Yamamoto-Mitani,
2004; Bonanno et al., 2005; Chentsova-Dutton et al., 2002; Li, 2005; Schulz,
Mandelson, & Haley, 2003; Zhang, Mitchell, Bambauer, Jones, & Prigerson,
2008). A few studies examined reactions to the death of a child (Bonanno et al.,
2005; Wijngaards-de Meij et al., 2008; Wortman & Silver, 1993). These stud-
ies assessed depression or other forms of distress in the early months following
the death, and then again anywhere from 13 to 60 months after the loss. The
construct of depression/distress was operationalized differently in the different
studies. For example, some studies utilized the Symptom Checklist-90 (SCL-90)
depression subscale and/or Diagnostic and Statistical Manual of Mental Disorders
(DSM)-based Structured Clinical Interview (SCID; e.g., Bonanno et al., 2005);
other studies such as the Changing Lives of Older Couples Study (CLOC) used
the Center for Epidemiologic Studies Depression Scale (CES-D; e.g., Bonanno et
al., 2002). For each study, the investigators determined a cut-off score to classify
respondents as high or low in distress or depression.
The longitudinal studies identified here provide evidence regarding the preva-
lence of different patterns of grief. The pattern that was traditionally considered
to be “normal” grief (moving from high distress to low distress over time) was
found among 41% of participants in a study on loss of a child from sudden infant
death syndrome (SIDS; Wortman & Silver, 1987), and anywhere between 9% and
41% in studies on conjugal loss (see Wortman & Boerner, 2011, for a review).
Furthermore, in these studies, evidence for “minimal” or “absent” grief (scoring
low in distress consistently over time) was found for 26% in the SIDS study, and
for anywhere between 41% and 78% in the studies on loss of a spouse (Wortman
58  Kathrin Boerner et al.
& Boerner, 2011). Taken together, these studies provided evidence that different
patterns of grief can typically be found in response to different types of losses
(e.g., death of a spouse or child); that what was traditionally viewed as a “normal”
pattern of grief is never experienced by a majority, in some cases even only by a
minority; and that the absence of intense distress is not at all an uncommon phe-
nomenon. However, none of these studies had a prospective design spanning from
the time before the loss to the time after. Thus, the ability to capture trajectories
of distress was rather limited.
This limitation was overcome in more recent prospective studies. In the CLOC
study on conjugal loss, which assessed older adults 3 years before and up to 18
months after the loss (Bonanno et al., 2002; Bonanno, Wortman, & Nesse, 2004),
nearly half of the participants (46%) experienced low levels of distress consistently
over time and were labeled “resilient.” Only 11% showed so-called “common”
grief. Another trajectory in this study referred to as “depressed–improved”
reflected elevated distress before the loss and improvement after the loss (10%).
A similar pattern of reduced distress levels following the loss was detected in
prospective studies that included both pre- and post-loss data on caregivers of
dementia patients (Aneshensel et al., 2004; Schulz et al., 2003; Zhang et al., 2008),
as well as on caregiver samples that included a variety of illnesses (Li, 2005).
In two of these studies (Aneshensel et al., 2004; Zhang et al., 2008), only about
17% showed a pattern of distress reflecting “common” grief following the death.
Moreover, Aneshensel and colleagues observed a pattern of stable but low distress
(64%) and absent distress (11%) in a majority of their participants, and Zhang and
colleagues found persistently absent depression in about half of their sample.
Taken together, in all studies, less than half of the sample showed what tra-
ditionally was considered normal grief. In the methodologically more advanced
prospective studies described above, such a reaction was even shown by only
a small minority of respondents. In fact, in the CLOC study on conjugal loss
(Bonanno et al., 2002), the relatively small proportion of those who showed the
pattern of moving from high distress to low distress over time was almost equal to
those who showed a depressed–improved pattern of being more distressed before
the loss, followed by improvement after the loss. Most importantly, however, the
available evidence shows that “minimal” grief is very common. The number of
respondents failing to show elevated distress or depression at the initial or final
time point was sizable, ranging from one quarter of the sample to more than three
quarters of the sample. In fact, in the available prospective studies that included
data from before and after the death, the resilient trajectory consistently emerged
for at least half of the sample. Similarly, a comparison of non-bereaved and
bereaved individuals (who lost either a child or a spouse; Bonanno et al., 2005)
showed that, in terms of distress levels, slightly more than half of the bereaved
did not significantly differ from the matched sample of non-bereaved individuals
when assessed at 4 and 18 months post loss.
It should be noted that labels such as “minimal” grief mean not an absence of
distress after the loss, but rather that, despite brief spikes in distress (Bonanno et
al., 2005) or a short period of daily variability in well-being (Bisconti, Bergeman,
 Uncomplicated and complicated patterns of grief  59
& Boker, 2006), people with these patterns had generally managed to function
at or near their normal levels (Bonanno, 2005). The prevalence of the “minimal”
grief reaction alone calls into question the assumption that failure to show distress
following a loss is pathological. In fact, it suggests that understanding why so
many people do not exhibit significant distress following a loss should become an
important research priority.

Evidence for delayed grief

Does the failure to display overt grief and sadness predispose a person to a
“delayed” grief reaction or later health problems? The data from the longitudi-
nal studies we identified fail to support this view. In two studies, there were no
respondents showing a delayed grief reaction (Bonanno & Field, 2001; Zisook &
Shuchter, 1986). In the remaining studies, the percentage of respondents showing
delayed grief ranged from 0.02% to 5% (Wortman & Boerner, 2011, for a review),
indicating that if delayed grief does occur it is exceedingly rare. Nor do physical
symptoms appear to emerge among those who fail to experience distress soon
after the loss. Both the Boerner et al. (2005) and Bonanno and Field (2001) stud-
ies are convincing on this point, because conjugally bereaved individuals were
assessed over a 4- and 5-year period, respectively, using multiple outcome meas-
ures. Similar findings emerged in a study by Middleton and colleagues. Based on
cluster analyses of several bereaved samples, the authors concluded that there was
no evidence for delayed grief (Middleton, Burnett, Raphael, & Martinek, 1996).
Nonetheless, in the previously described survey conducted by Middleton et al.
(1993), a substantial majority of researchers and clinicians (76.6%) indicated that
delayed grief does occur.

Evidence for chronicity

Empirical evidence suggests that, whereas most bereaved individuals do not seem
to experience intense distress for extended periods of time, a significant minority
of people develop long-term difficulties. In each of the longitudinal bereavement
studies mentioned previously, for example, a pattern of persistent distress or
“chronic grief” emerged. This pattern of chronic grief was found among 30% of
participants who lost a child from SIDS (Wortman & Silver, 1987), and anywhere
between 8% and 26% of the conjugally bereaved (Wortman & Boerner, 2011, for
a detailed review). Recent caregiving studies have found similar percentages for
the chronic grief trajectory among bereaved caregivers (ranging from 8% to 17%;
e.g., Aneshensel et al., 2004; Zhang et al., 2008).
In our prospective work on conjugal loss (Bonanno et al., 2002), the avail-
ability of pre-loss data made it possible to further distinguish two types of chronic
distress. Although chronically elevated depressive symptoms during bereave-
ment were evidenced by 23% of the sample, these participants formed two
distinct trajectories. One trajectory, constituting 16% of the sample, suggested
an unambiguous chronic grief reaction. These participants manifested low levels
60  Kathrin Boerner et al.
of depression prior to the loss but then showed elevated depression at 6 and 18
months of bereavement. A second smaller group, constituting 8% of the sample,
had markedly elevated depression prior to bereavement and then showed only
a slight increase and remained depressed during bereavement. Both groups had
higher levels of grief-specific symptoms (e.g., yearning) measured at 6 and 18
months of bereavement than all other participants, and did not differ from each
other in their level of grief symptoms. However, as we discuss below, additional
data suggested that one of these trajectories represented a relatively pure chronic
grief reaction whereas the other pattern was more representative of a pre-existing
chronic depression.
Most classic grief theorists (e.g., Jacobs, 1993) discuss the notion of chronic
grief but fail to indicate how long it typically lasts and whether it abates at some
point. To address this issue, we conducted a follow-up analysis investigating
whether the chronic grievers and the chronically depressed would remain dis-
tressed up to 48 months post loss (Boerner et al., 2005). Overall, the chronic grief
group experienced an intense and prolonged period of distress but, by the 48–
month time point, the grief of this group did resolve. In contrast, the chronically
depressed group clearly demonstrated long-term problems, with little indication
of improvement between 18 and 48 months. This group not only showed the poor-
est adjustment 4 years after the loss but also struggled the most with questions
about meaning. These differential findings for the chronic grief and depression
group underscore the need to further refine the criteria used to identify those who
are at risk for long-term problems.
Are these grief trajectories veridical, or are they merely arbitrary distinctions?
More recently, Mancini, Bonanno, and Clark (2011) examined grief trajectories
using latent growth mixture modeling. This approach allows bereavement trajec-
tory patterns to emerge in a non-arbitrary and purely empirical manner, addressing
critical limitations of prior approaches. The authors used a panel data set of over
16,000 German citizens followed for over 20 years that included pre-loss meas-
ures of functioning. This analysis revealed a four-trajectory solution that mirrored
the prototypical outcome trajectories to a surprising degree. What emerged were
resilient (58.7%), chronic post-loss distress (21.3%), chronic pre- and post-loss
distress (14.6%), and improved (5.4%) trajectories. These analyses offered impor-
tant confirmatory support for the trajectory patterns described earlier, indicating
that there are complicated and uncomplicated (or resilient) patterns of response
following a loss.

Predictors of grief trajectories

A variety of contextual, situational, and person-centered factors can potentially
contribute to whether a person displays a complicated or uncomplicated grief
trajectory (Bonanno, 2004; Mancini & Bonanno, 2009). However, because rela-
tively few studies have identified trajectories of bereavement and even fewer have
used pre-loss data, our understanding of these factors is still somewhat limited.
Nevertheless, extant research has identified factors that are associated with each
of the trajectory patterns discussed earlier. Next we discuss these predictors.
 Uncomplicated and complicated patterns of grief  61

Resilience
There is consistent evidence across studies that resilient individuals are gener-
ally unlikely to have a history of prior mental health problems or to show any
other signs of dysfunction in their lives (e.g., Bonanno, Boerner, & Wortman,
2008; Wortman & Boerner, 2011, for a review). For example, an examination of
pre-bereavement measures in the CLOC study (Bonanno et al., 2002) showed
that, prior to the spouse’s death, this group did not have conflicted or low-quality
marital relations with the spouse, nor were they ambivalent about or excessively
dependent on the spouse. They also did not evidence extreme scores on any of the
personality measures included in the study such as extraversion or emotional sta-
bility. Rather, participants in the resilient group scored higher than other partici-
pants on several pre-loss measures suggestive of resilience-promoting factors that
would better prepare them for coping with the impeding loss. For example, they
reported relatively high levels of instrumental support, and scored higher than
other participants on questionnaire measures of belief in a just world and accept-
ance of death. Overall, it seemed that participants in the resilient group were able
to cope with their loss in a very positive way (Bonanno et al., 2004). They were
better able than other participants to gain comfort from talking about or thinking
about the spouse. For example, they were more likely than other bereaved people
to report that thinking about and talking about their deceased spouse made them
feel happy or at peace. They had low scores on avoidance/distraction, suggesting
that their lack of distress is indicative of good adjustment rather than defensive
denial. They also reported the fewest regrets about their behavior with the spouse,
or about things they may have done or failed to do when he or she was still alive.
Finally, participants in the resilient group were less likely to try to make sense of
or find meaning in the spouse’s death. Thus, they did not engage in any type of
ruminative thought about the loss.

Improvement during bereavement

By definition, and in contrast to the stable low depression and resilient groups,
the improved participants were highly depressed prior to the loss. However, it
is not necessarily clear that this reflects a history of prior mental health prob-
lems. Rather, evidence from different studies suggests that behind the scenes of a
depressed–improved trajectory one is likely to find serious illness and a demand-
ing caregiving situation. As mentioned above, this trajectory has been found to
be the most dominant pattern in caregiving and bereavement studies (e.g., Schulz
et al., 2003). Similarly, Bonanno and colleagues (Bonanno et al., 2002) found
that virtually all the participants in the depressed–improved group had been con-
tending with a seriously ill spouse at the time of the pre-bereavement interviews.
This was an elderly sample and illness among the spouses was not uncommon.
However, no other group was so clearly characterized by spousal illness as the
improved group.
Given the demands and burden of providing care to a seriously ill person, it is
unsurprising that this group also showed a relatively unfavorable profile on other
62  Kathrin Boerner et al.
pre-bereavement measures of the CLOC study. They had the poorest-quality mar-
riages of all participants, making the least positive and most negative evaluations
of their spouse and marriage, and scoring higher on a measure of ambivalence
toward the spouse in the pre-bereavement interviews. They also scored high on
measures of emotional instability (neuroticism), introspection, and perceived
personal injustice. However, despite this conspicuously unfavorable pre-bereave-
ment profile, the improved group dropped to relatively low levels of depression
and reported relatively low levels of grief symptoms during bereavement. Like
the resilient group, the improved participants were also relatively less likely to
search for meaning in the loss. Given the difficulties they had experienced prior to
the death of their spouse, it is tempting to assume that people could improve this
much only by relying heavily on denial or distraction during bereavement. Yet,
again like the resilient group, the improved participants had relatively low scores
on a set of questionnaire items that tapped the use of avoidance or distraction
(Bonanno et al., 2004). Although the improved participants reported thinking and
talking about the loss less frequently than the resilient group, and reported the
lowest levels of comfort from memories of the spouse, it is noteworthy that they
exhibited marked increases in the ability to find comfort from thinking or talking
about the spouse (the only group of participants in the CLOC study to do so). By
18 months of bereavement they had increased so much in this regard that they
were no longer distinct from the resilient group on this variable. Finally, and of
particular note, those in the improved group were fully aware of the remarkable
progress they had made. This group scored higher than any other group on a
scale designed to measure the perception of pride in coping ability. The improved
group was also more likely than other participants in the CLOC study to report
becoming more confident and becoming a stronger person as a result of dealing
with the loss of their spouse.

Chronic grief and distress

Both response patterns, chronic grief and depression, were characterized by high
levels of distress during bereavement, but only the latter pattern also involved
high distress prior to the loss. To further elucidate the nature of these patterns,
Bonanno et al. (2002) identified their pre-loss predictors. Chronic grievers were
likely to have had healthy spouses, to rate their marriage positively, and to show
high levels of pre-loss dependency on the spouse (e.g., stating that no one could
take the spouse’s place). The chronically depressed group was less positive about
their marriage than chronic grievers, but equally dependent on their spouse.
Further analyses examined the context and processing of the loss at 6 and 18
months post loss (Bonanno et al., 2004). Results indicated that chronic grief stems
from an enduring struggle with cognitive and emotional distress related to the
loss, whereas chronic depression results more from enduring emotional difficul-
ties that are exacerbated by the loss. For example, at 6 months post loss, chronic
grievers were more likely to report current yearning and emotional pangs, and
they reported thinking and talking about the deceased more often than did chroni-
cally depressed individuals.
 Uncomplicated and complicated patterns of grief  63
Together, these findings suggest that during bereavement the chronic grief
group was struggling primarily with the loss of a beloved and vital spouse on
whom they were also dependent. In addition, because the spouses for this group
were typically healthy in the years prior to their death, this struggle was most
likely exacerbated by a lack of anticipation or psychological preparation for the
loss. In contrast, the pre-bereavement characteristics of people showing a chronic
depression trajectory suggested that whatever negative reactions they might have
had to the spouse’s death were layered upon an already considerable number of
ongoing psychological difficulties.
Although the distinction between chronic grief and depression has greatly
informed the ways in which we think about high distress patterns in response
to loss, we would also like to highlight some of the work that has furthered our
understanding of long-term adjustment difficulties in a more general way. For
example, a trajectory of chronic distress has also been found in the population of
bereaved caregivers. As described above, although for many caregivers symptoms
of depression and grief decline rapidly after the death and return to near normal
levels within a year of the death (e.g., Schulz et al., 2003), a sizable minority
continues to experience high levels of stress and psychiatric problems after death.
High levels of caregiving burden, feeling exhausted and overloaded, and lack of
support have consistently been found to be associated with negative bereavement
outcomes (Gross, 2007; Hebert, Dang, & Schulz, 2006). In addition, a study with
dementia caregivers found that among the caregivers with clinical levels of com-
plicated grief were not only those who were in difficult caregiving situations (e.g.,
caring for a more cognitively impaired patient), but also some who reported very
positive features of the caregiving experience (Schulz, Boerner, Shear, Zhang, &
Gitlin, 2006). This intriguing finding suggests first of all that positive and negative
aspects of caregiving can coexist, and that there may be some positive caregiving
experiences that can also put a person at risk for subsequent difficulties. This
may be the result of two related factors: Losing their loved one deprives these
individuals of a meaningful and important role; and a positive view of caregiving
may be a reflection of a very close relationship between caregiver and the person
they cared for that might have been the center of the caregiver’s life (Boerner,
Schulz, & Horowitz, 2004).
As noted above, another aspect that is likely to contribute to the described
patterns of grief is preparedness for the death. For example, in the CLOC study,
those with a trajectory of chronic grief tended to have healthy spouses and were
probably rather unprepared for their spouse’s death. However, even in the context
of illness and caregiving, it seems clear that, despite providing high-intensity care,
often for years, many bereaved caregivers perceive themselves as unprepared for
the death. There is also evidence that those who do feel unprepared typically
report more depressive, anxiety, and complicated grief symptoms (Hebert et
al., 2006). However, we need to gain a better understanding of what it means
to be prepared. Based on a series of focus groups with caregivers, Hebert and
colleagues proposed that preparedness has emotional (e.g., being at peace with
prospect of death), pragmatic (e.g., having funeral arrangements planned), and
informational (e.g., medical aspects of end of life) components (Hebert, Schulz,
64  Kathrin Boerner et al.
Copeland, & Arnold, 2009). This study also showed that, for example, a person
could feel prepared with respect to the informational and pragmatic components,
but yet feel entirely unprepared emotionally. Overall, this work suggests that even
the relatively certain prospect of death does not necessarily translate into being
prepared for what lies ahead, and that this may be an important area for profes-
sionals to address in their encounters with caregivers, before and after the loss.
Finally, not only circumstances around the death but also cause of death are gen-
erally considered as important predictors of bereavement outcomes. Specifically,
complications in grieving are more likely if the death was sudden and violent
(Kaltman & Bonanno, 2003; Mancini, Prati, & Black, 2011; Murphy, Johnson,
Chung, & Beaton, 2003). Adjustment difficulties are more likely in response to
the death of a child, particularly if the death occurred under sudden or violent
circumstances. Rather, experiencing intense distress is considered as normative.
In fact, comparing the different studies with data on grief trajectories described
above, the highest percentage of respondents showing a pattern of consistently
high levels of distress following the loss came from the study on the death of a
child from SIDS (Wortman & Silver, 1987). Another important consideration is
the striking difference among the studies on conjugal loss in the percentage of
respondents evidencing chronic grief (ranging from 8% to 26%). This may be
related to differences in the age of the respondents, and hence the untimeliness
of the loss in some of these studies (e.g., Vachon et al., 1982; loss of a spouse in
mid- rather than late life).

Concluding thoughts
Research has clearly demonstrated that the pattern of “normal grief” is not nearly
as common as was long assumed, and that variations from this response pattern,
such as consistent minimal distress following the death of a loved one, not only are
rather common but can also constitute a very adaptive response to loss. Moreover,
research studies have shown that under certain circumstances, for example when
death occurs after a long period of illness and caregiving, distress levels may be
heightened during the time leading up to the death, and then subsequently decline.
A better understanding of the prevalence and variety of grief patterns is critical
for health care professionals because it allows them to meet bereaved individuals
with realistic expectations and an empirically grounded understanding of what
represents complicated versus uncomplicated patterns of grief. As a result, the
health care professional is less likely to impose an expectation of how one should
grieve. Such expectations may undermine a person’s coping and result in unnec-
essary distress (Wortman & Boerner, 2011).
At the same time, it is important for health care professionals to be aware
that complicated grief reactions do occur in a significant minority of bereaved
individuals, and to know the symptoms of complicated grief, as well as charac-
teristics or circumstances (e.g., having a history of prior mental health problems,
or experiencing extremely high levels of caregiving burden) that may increase
the likelihood of a complicated grief reaction. Health care professionals may be
 Uncomplicated and complicated patterns of grief  65
in a position to help address some of these issues during the time prior to death
(e.g., help find additional support to reduce burden when caregiving needs begin
to become overwhelming), or they may have a key role in connecting bereaved
individuals who seem more severely distressed to the appropriate support source,
by making a referral to a clinician who can diagnose complicated grief and pro-
vide or recommend a specific intervention strategy geared toward the individual’s
needs.

References
Aneshensel, C.  S., Botticello, A.  L., & Yamamoto-Mitani, N. (2004). When caregiving
ends: The course of depressive symptoms after bereavement. Journal of Health and
Social Behavior, 45, 422–440.
Bisconti, T. L., Bergeman, C. S., & Boker, S. M. (2006). Social support as a predictor of
variability: An examination of the adjustment trajectories of recent widows. Psychology
and Aging, 21, 590–599.
Boerner, K., Schulz, R., & Horowitz, A. (2004). Positive aspects of caregiving and adapta-
tion to bereavement. Psychology and Aging, 19, 668–675.
Boerner, K., Wortman, C. B., & Bonanno, G. A. (2005). Resilient or at risk? A four-year
study of older adults who initially showed high or low distress following conjugal loss.
Journal of Gerontology: Psychological Science, 60B, P67–P73.
Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the
human capacity to thrive after extremely aversive events. American Psychologist, 59,
20–28.
Bonanno, G. A. (2005). Resilience in the face of potential trauma. Current Directions in
Psychological Science, 14, 135–138.
Bonanno, G. A., Boerner, K., & Wortman, C. B. (2008). Trajectories of grieving. In Stroebe,
M., Hansson, R., Schut, H., & Stroebe, W. (Eds.), Handbook of bereavement research
and practice: 21st century perspectives (pp.  287–307). Washington, DC: American
Psychological Association Press.
Bonanno, G. A., & Field, N. P. (2001). Evaluating the delayed grief hypothesis across 5
years of bereavement. American Behavioral Scientist, 44, 798–816.
Bonanno, G. A., Keltner, D., Holen, A., & Horowitz, M. J. (1995). When avoiding unpleas-
ant emotion might not be such a bad thing: Verbal-autonomic response dissociation
and midlife conjugal bereavement. Journal of Personality and Social Psychology, 46,
975–985.
Bonanno, G. A., Moskowitz, J. T., Papa, A., & Folkman, S. (2005). Resilience to loss in
bereaved spouses, bereaved parents, and bereaved gay men. Journal of Personality and
Social Psychology, 88, 827–843.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring, M., Sonnega, J.,
et al. (2002). Resilience to loss and chronic grief: A prospective study from pre-loss to
18 months post-loss. Journal of Personality and Social Psychology, 83, 1150–1164.
Bonanno, G. A., Wortman, C. B., & Nesse, R. M. (2004). Prospective patterns of resilience
and maladjustment during widowhood. Psychology and Aging, 19, 260–271.
Bowlby, J. (1980). Attachment and loss, vol. 3: Loss: Sadness and depression. New York:
Basic Books.
Chentsova-Dutton, Y., Shuchter, S., Hutchin, S., Strause, L., Burns, K., Dunn, L., et al.
(2002). Depression and grief reactions in hospice caregivers: From pre-death to 1 year
afterwards. Journal of Affective Disorders, 69, 53–60.
66  Kathrin Boerner et al.
Gross, J. (2007, November 19). Study finds higher outlays for caregivers of older relatives.
New York Times, p. A18.
Hebert, R. S., Dang, Q., & Schulz, R. (2006). Preparedness for the death of a loved one
and mental health in bereaved caregivers of patients with dementia: Findings from the
REACH study. Journal of Palliative Medicine, 9, 683–693.
Hebert, R. S., Schulz, R., Copeland, V. C., & Arnold, R. M. (2009). Preparing family care-
givers for death and bereavement: Insights from caregivers of terminally ill patients.
Journal of Pain Symptom Management, 37, 3−12.
Horowitz, M.  J., Siegel, B., Holen, A., Bonanno, G. A., Milbrath, C., & Stinson, C.  H.
(1997) Diagnostic criteria for complicated grief disorder. American Journal of
Psychiatry, 154, 904–910.
Jacobs, S. (1993). Pathological grief: Maladaptation to loss. Washington, DC: American
Psychiatric Press.
Kaltman, S., & Bonanno, G. A. (2003). Trauma and bereavement: Examining the impact of
sudden and violent death. Journal of Anxiety Disorders, 17, 131–147.
Li, L. W. (2005). From caregiving to bereavement: Trajectories of depressive symptoms
among wife and daughter caregivers. Journal of Gerontology, 60B, P190–P198.
Lund, D. A., Dimond, M. F., Caserta, M. S., Johnson, R. J., Poulton, J. L., & Connelly, J. R.
(1985–1986). Identifying elderly with coping difficulties after two years of bereave-
ment. Omega, 16, 213–224.
Mancini, A.  D., & Bonanno, G. A. (2006). Resilience in the face of potential trauma:
Clinical practices and illustrations. Journal of Clinical Psychology: In Session, 62,
971–985.
Mancini, A. D., & Bonanno, G. A. (2009). Predictors and parameters of resilience to loss:
Toward an individual differences model. Journal of Personality, 77, 1805–1832.
Mancini, A.  D., Bonanno, G.  A., & Clark, A (2011). Stepping off the hedonic tread-
mill: Individual differences in response to major life events. Journal of Individual
Differences, 32, 144–152.
Mancini, A. D., Prati, G., & Black, S. (2011). Self-worth mediates the effects of violent
loss on PTSD symptoms. Journal of Traumatic Stress, 24, 116–120.
Middleton, W., Burnett, P., Raphael, B., & Martinek, N. (1996). The bereavement response:
A cluster analysis. British Journal of Psychiatry, 169, 167–171.
Middleton, W., Moylan, A., Raphael, B., Burnett, P., & Martinek, N. (1993). An interna-
tional perspective on bereavement related concepts. Australian & New Zealand Journal
of Psychiatry, 27, 457–463.
Murphy, S. A., Johnson, L. C., Chung, I., & Beaton, R. D. (2003). The prevalence of PTSD
following the violent death of a child and predictors of change 5 years later. Journal of
Traumatic Stress, 16, 17–25.
Osterweis, M., Solomon, F., & Green, F. (Eds.) (1984). Bereavement: Reactions, conse-
quences, and care. Washington, DC: National Academy Press.
Parkes, C.  M. (1965). Bereavement and mental illness. British Journal of Medical
Psychology, 38, 1–26.
Prigerson, H. (2004). Complicated grief: When the path of adjustment leads to a dead-end.
Bereavement Care, 23, 38–40.
Raphael, B. (1983). The anatomy of bereavement. New York: Basic Books.
Schulz, R., Boerner, K., Shear, K., Zhang, S., & Gitlin, L. N. (2006). Predictors of compli-
cated grief among dementia caregivers: A prospective study of bereavement. American
Journal of Geriatric Psychiatry, 14, 650–658.
 Uncomplicated and complicated patterns of grief  67
Schulz, R., Mendelson, A. B., & Haley, W. E. (2003). End-of-life care and the effects of
bereavement on family caregivers of persons with dementia. New England Journal of
Medicine, 349, 1936–1942.
Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. III. (2005). Treatment of complicated
grief: A randomized controlled trial. Journal of the American Medical Association, 293,
2601–2608.
Vachon, M. L. S., Rogers, J., Lyall, W. A., Lancee, W. J., Sheldon, A. R., & Freeman, S. J. J.
(1982). Predictors and correlates of adaptation to conjugal bereavement. American
Journal of Psychiatry, 139, 998–1002.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., van
der Heijden, P.  G., & Dijkstra, I. (2008). Parents grieving the loss of their child:
Interdependence in coping. British Journal of Clinical Psychology, 47, 31–42.
Wortman, C. B., & Boerner, K. (2011). Reactions to the death of a loved one: Myths of
coping versus scientific evidence. In Friedman, H. S. (Ed.), Oxford handbook of health
psychology (pp. 441–479). New York: Oxford University Press.
Wortman, C. B., & Silver, R. C. (1987). Coping with irrevocable loss. In VandenBos, G. R.,
& Bryant, B.  K. (Eds.), Cataclysms, crises, and catastrophes: Psychology in action
(pp. 189–235). Washington, DC: American Psychological Association.
Wortman, C.  B., & Silver, R.  C. (1989). The myths? of coping with loss. Journal of
Consulting and Clinical Psychology, 57, 349–357.
Wortman, C. B., & Silver, R. C. (1993). Reconsidering assumptions about coping with loss:
An overview of current research. In Filipp, S. H., Montada, L., & Lerner, M. (Eds.),
Life crises and experiences of loss in adulthood (pp. 341–365). Hillsdale, NJ: Erlbaum.
Zhang, B., Mitchell, S.  L., Bambauer, K.  Z., Jones, R., & Prigerson, H.  G. (2008).
Depressive symptom trajectories and associated risks among bereaved Alzheimer dis-
ease caregivers. American Journal of Geriatric Psychiatry, 16, 145–155.
Zisook, S., & Shuchter, S.  R. (1986). The first four years of widowhood. Psychiatric
Annals, 16, 288–294.
6 Complicated grief in children1
Atle Dyregrov and Kari Dyregrov

Introduction
A new grief disorder has been suggested for inclusion in the fifth edition of the
Diagnostic and Statistical Manual for Mental Disorders (DSM-5; see Prigerson
et al., 2009). However, children are not mentioned in the proposal. It is important
that our understanding and intervention efforts reflect the uniqueness of children’s
grief and that an adult diagnosis is not inappropriately used for children. In this
chapter we will first describe the consequences of and risk/protective factors
associated with childhood bereavement. Then we will discuss what constitutes
complicated grief in children, including why the proposed diagnosis of prolonged
grief disorder is inadequate for capturing the variety of complicated grief reac-
tions in children. Finally, we present what is known about intervention following
bereavement in childhood and discuss some important issues in this regard.

Consequences of childhood bereavement

It is estimated that around 4% of children in Western countries experience the
death of a parent before they reach the age of 18 (Pearlman, Schwalbe, & Cloitre,
2010). In addition, many young people experience the death of siblings, close
friends, and other loved ones. The death of a parent or child in a family will
usually result in a profound crisis in children and adolescents (Luecken, 2008;
Tremblay & Israel, 1998), increasing the occurrence of mental health problems
(Cerel, Fristad, Verducci, Weller, & Weller, 2006; Dowdney, 2000; Dyregrov &
Dyregrov, 2005; Pfeffer, Karus, Siegel, & Jiang, 2000; Silverman & Worden,
1993), decline in school performance, social withdrawal and behavioral problems
(Dowdney et al., 1999; Dyregrov, 2009; Luecken, 2008), and somatic complaints
(Silverman & Worden, 1992). A minority (approximately 20%) evidence more
severe problems (Dowdney, 2000; Worden & Silverman, 1996). A large Swedish
register study showed increased mortality following parental death, especially
when death was due to unnatural causes (Rostila & Saarela, 2011).
In addition to mental health problems, early parental loss has been associated
with physical health problems throughout life, in both epidemiological and cor-
relational studies (for a review see Luecken, 2008). The loss of a sibling results in
rates of problems equivalent to the loss of a parent (Worden, Davies, & McCown,
 CG in children  69
1999). Higher and more persistent rates of internalizing problems have been
found among bereaved girls than among boys (Schmiege, Khoo, Sandler, Ayers
& Wolchik, 2006; Worden & Silverman, 1996). The bulk of the research implies
an increased risk of depression in adulthood (Coffino, 2009; Jacobs & Bovasso,
2009).

Risk and protective factors

Although this is not based on a thorough review of all child bereavement studies,
we have extracted some important risk and protective factors found in numerous
studies. A variety of factors can increase or decrease the risk of a child developing
problems in the aftermath of a loss. Usually risk and protective factors mirror each
other: A good family climate will be protective whereas a negative family climate
is associated with risks.
To the extent that a death leads to massive changes in the child’s daily envi-
ronment – if the economic situation demands that the family move, resulting in
the loss of contact with friends, having to change school, and so on – the pos-
sibility of negative mental health changes increases (Coffino, 2009). Parents can
find themselves with increased responsibilities and tasks, and possible financial
difficulties, and thus have less time to spend with children. This may also make
them less emotionally available. Altogether, this may result in decreased parental
capacity. Substantial variance in children’s psychopathology could be explained
by subsequent life disruption and parental unavailability.
Good parental or primary caregiver capacity that succeeds in providing
warmth and discipline has been shown to be protective for bereaved children
(Haine, Wolchik, Sandler, Millsap, & Ayers, 2006; Lin, Sandler, Ayers, Wolchik,
& Luecken, 2004) and conversely poor-quality parenting and lack of support at
home are risk factors for a poor psychological outcome (Luecken, 2008; Tremblay
& Israel, 1998). Children’s (traumatic) grief levels have been found to be strongly
related to the emotional reactions of the caregiver at the time of death and the
degree of sadness in the home when retrospectively assessed 21 months after the
death (Brown et al., 2008). Saldinger, Porterfield, and Cain (2004) have shown
how mothers show greater child-centeredness than fathers and have also docu-
mented severe breakdown in parental attunement due to their preoccupation with
a dying spouse. Caregivers in bereaved families following a sudden parental loss
(suicide, accident, and sudden natural death) show increased rates of depression,
anxiety, posttraumatic stress disorder (PTSD), suicidal ideation, and functional
impairment compared with matched non-bereaved control families (Melhem,
Walker, Moritz, & Brent, 2008). This may place children at further jeopardy.
Losing a mother results in more problems than losing a father (Brent, Melhem,
Donahue, & Walker, 2009), including higher child mortality risk (Rostila &
Saarela, 2011). In families that lose a mother the burden of changes often falls
most heavily on daughters, who often assume more responsibility for household
tasks and care of siblings (Worden, 1996). Female gender is associated with
increased vulnerability (Schmiege et al., 2006). Riches and Dawson (2000), in an
70  Atle Dyregrov and Kari Dyregrov
excellent qualitative study, have explored how a widower’s early remarriage can
complicate the bereaved daughter’s grief resolution, leaving her with a sense of
betrayal and reduced opportunities for conversational remembering among family
and friends. This may prolong a daughter’s grief. Gender differences may change
as societal roles and expectations change.
If one follows a temporal organization of risk factors, pre-loss exposure to
death is a risk factor for later mental health problems (Worden, 1996), as is new
stressors in the post-loss environment (Hagan, Luecken, Sandler, & Tein, 2010).
Less is known about the risk factors associated with the death itself. There is some
indication that violent deaths result in more problems than non-violent causes
(Brent et al., 2009; Dyregrov, Nordanger, & Dyregrov, 2003; Worden, 1996), that
children who witness physical distress in their dying parents struggle more than
those who do not (Saldinger, Cain, & Porterfield, 2003), that stronger grief reac-
tions are related to the caregiver’s emotional reactions at the time of death, and
that children who themselves experience life-threat during the event in which a
parent dies experience stronger grief reactions than others (Brown et al., 2008).
From research on parental capacity and warmth (Lin et al., 2004), and on sadness
in the home (Brown et al., 2008), there is reason to believe that the immediate
post-loss emotional climate as well as the family’s communication and fact-shar-
ing propensity (Saler & Skolnick, 1992) are important for functioning over time.
Small children are totally dependent on adults for adequate information about
the loss and about the normality of their own reactions. Without being able to talk
freely with the surviving parent and others, risk of forming misinterpretations
increases (see Tremblay & Israel, 1998, for a review). In adults, catastrophic mis-
interpretations of grief reactions have been found to be associated with increased
levels of prolonged grief (Boelen, van den Bout, & van den Hout, 2010). Clinical
experience suggests that children form more misinterpretations than adults, as
they lack life experience and direct access to information about what happened,
and studies are needed to address this.

Normal and complicated grief in children

There is as yet no clear definition of what is normal and complicated grief in
children. The division is more difficult than for adults because children are under-
going development of important brain areas involved in emotion and cognition
as well as gaining rapid and new life experience (Gunnar & Quevedo, 2007).
The development of these brain areas forms the basis for regulating emotions
and understanding the loss and its long-term consequences. If, added to this,
one considers the influence of parental reactions and different family practice
regarding communication of facts, feelings, thoughts, and so on, the separation
of complicated grief from normal grief becomes difficult. In addition, grief may
manifest itself in different ways in children from in adults, and its manifestations
will vary across cultures.
Although no clear definition of healthy grieving exists, the focus usually is on
the individual’s ability to accept the reality of the death, cope with the emotional
 CG in children  71
ramifications, and adjust to the changes in one’s life. In addition, progressing
through normal developmental stages, resuming one’s life and being able to
remain functional in school or other domains of life are usually taken into account
(Dyregrov, 2009; Goodman et al., 2004; Worden, 1996). Although stage models
of progression through grief have largely been abandoned, healthy grieving is
often conceptualized as fulfilling certain tasks (Worden, 1996). However, many
studies originate in the United States and the focus on adaptability and return
to function found in this society may be less suited for understanding grief in
children in different cultural and historical contexts. What then constitutes com-
plicated grief in children?

A new diagnosis: made for adults

As mentioned at the beginning of this chapter, a new grief disorder has been sug-
gested for inclusion in the DSM-5 (Prigerson et al., 2009). Presently it is grouped
under adjustment disorders and is termed bereavement related disorder (http://
www.dsm5.org/ProposedRevision/Pages/proposedrevision.aspx?rid=367).
Besides having lost a significant other, it involves persistent yearning, preoc-
cupation with the lost person, different distress reactions, and social/identity
disruption. The diagnosis should not be made until at least 12 months after the
death and the disturbance should cause clinically significant impairment in social,
occupational, or other important areas of functioning.
There are several problems in adopting this disorder for children. It includes
only one type of grief: where grief continues unabated. Clinically we do not see
this type as frequently among children as among adults on account of children’s
lack of tolerance of emotions. Another problem with the adult proposal is the
heavy emphasis on functional impairment. This is more difficult to assess among
children, especially pre-school children. Within the trauma field this has led to the
development of an alternative set of scoring criteria for PTSD at pre-school age
(Scheeringa, 2008). A parallel development can be expected if a grief diagnosis
is adopted for children. The duration criterion of 12 months is also troublesome
for children. Waiting 12 months until diagnosis may be very unfortunate as it will
leave plenty of time for maladaptive parental strategies to solidify problems or for
misunderstandings or misperceptions to be more ingrained in children’s thoughts
and behavior. Our own clinical experience, in line with others’ (Dowdney et al.,
1999), suggests that 12 months is too long to wait before intervening. If children
have problems that seriously affect their functioning and relationships, or the loss
results in maladaptive family dynamics, mental health support should be insti-
gated earlier.

Bowlby’s concept of pathological mourning

A good starting point for describing complicated grief in children is John
Bowlby’s original work. Bowlby (1963) impressively depicted several types of
complicated grief that echoes what we meet in clinical practice. He used not the
72  Atle Dyregrov and Kari Dyregrov
term complicated grief, but the term pathological mourning. He used mourning
not so much to denote cultural practices and expressions of grief, as is common
today, but more like the term grief is used nowadays: defining emotional, physi-
ological, cognitive, and behavioral reactions to the death of someone significant
(Stroebe, Hansson, Stroebe, & Schut, 2002). Bowlby described the following
subtypes of pathological mourning: (a) persistent and unconscious yearning to
recover the lost object; (b) intense and persistent anger and reproach expressed
towards various objects, including the self; (c) absorption in caring for someone
else who has also been bereaved, amounting to compulsion; and (d) denial that
the object is permanently lost – absence of grief (here the death is not acknowl-
edged, but there is some awareness of it). Bowlby did not see these types as
mutually exclusive, and proposed that they could be found in combinations with
anxiety, depression, and hysterical and psychopathic behavior. He linked healthy
mourning to the restoration of function, whereas in pathological mourning the
bereaved person remained preoccupied with the lost person in thought and action.
He stressed that open expression of yearning for the lost object, accompanied by
sadness and crying, and anger and reproach at the object for its desertion, were
necessary conditions for healthy mourning. In our opinion, although his descrip-
tion of children’s complicated grief is very acute, his description lacks the type
in which children struggle with posttraumatic problems following their loss, and
he also underemphasizes the problems that ensue in children as a result of the
family climate or adults’ inadequate handling of the death or its consequences.
Bowlby’s types reflect the lack of recognition of trauma at that time. Although
basing the categorization on adults, Worden (1996) listed similar types of com-
plicated grief to those of Bowlby, adding a masked type: reactions that resemble
physical or other psychiatric disorders that the bereaved do not associate with
the death.

Traumatic experiences and loss

Traumatic aspects related to witnessing the death, finding the body, or having
fantasies about what happened can lead to complicated grief reactions. They are
prominent enough to be recognized as a distinct type, although the reactions
that ensue can also be subsumed under posttraumatic stress reactions or PTSD.
Grief and trauma often become intertwined. However, from reading the literature
published in recent years, it seems that the focus in childhood grief has become
centered on traumatic grief (see Goodman et al., 2004). Coming from the trauma
field, Cohen and colleagues (Cohen, Goodman, Brown, & Mannarino, 2004)
formed the construct of “childhood traumatic grief” (CTG) to conceptualize
debilitating grief reactions in children and adolescents. This is grief following
a loss that the child experiences as traumatic, when he or she evinces significant
PTSD symptoms that prevent the child’s grief processing. CTG also keeps the
child from approaching happy memories (Cohen, Mannarino, & Staron, 2006).
Trauma processing takes precedence over grief processing. Though the trauma
focus may have taken some of the attention away from other forms of complicated
grief, this focus has made us much more aware of the interaction of trauma and
 CG in children  73
grief following many deaths. The similarity in PTSD symptoms and cortisol levels
found among children who have experienced loss events and children who have
experienced traumatic events (e.g., abuse, witnessing violence) (Taylor, Weems,
Costa, & Carrión, 2009) also underscores the importance of understanding the
interaction of trauma and loss.
In the following subsections we outline a few important aspects that are spe-
cific to children and that may complicate their reactions following a loss.

Poor informational climate and loss

Family dynamics may aggravate complicated grief reactions in children or make
them more likely to develop. Intensified reactions may result when facts are not
communicated, information is kept from the child, or the emotional climate is
restricted or parents are unresponsive to their child’s needs (Lin et al., 2004).
This is an under-studied area, probably because it resides in the complex inter-
actional processes between the child and his or her environment. We believe it
may be fruitful to enhance the focus on these dynamics because children with
such living conditions become especially vulnerable. For this group of children,
trauma-specific grief therapy would not suffice and a family-oriented approach is
recommended to stimulate more open and direct communication about the death
and its consequences.

Lack of emotional tolerance and loss

Research on grieving adults has not found evidence of delayed grief (Bonanno &
Field, 2001). In clinical work with children we see that lack of emotional toler-
ance of the loss is evident and children try to regulate grief in tolerable doses and
use more avoidance and postponement than adults. It is known that frequent use
of an explicit strategy to regulate emotions can quickly render the initiation of
the strategy more implicit (automatic) over time (Gyurak, Gross, & Etkin, 2011).
Tremblay and Israel (1998) mention that children’s lack of fluency in identifying
and describing feeling states may also lead to more somatic expression of distress.
Saldinger and colleagues (2004) speculate that grieving children who score below
the norm for measures of anxiety and depression struggle to ward off feelings, in
order not to feel overwhelmed by them. Delayed or repressed reactions may be
so frequent in children that they can be seen as normal reactions, reflecting chil-
dren’s lack of emotional regulation capability. However, normal lack of emotional
responsiveness as a way to approach grief in tolerable doses is different from
almost total blockage of grief reactions. This form of overregulation will usually
be associated with functional decline if it continues over time. Clinically, some
children will deny any emotions tied to the lost person and may start avoiding
conversations about or memories of the deceased, often becoming dysfunctional
in some area, such as at school. Empirical work needs to be undertaken to clarify
when this way of modulating emotions become dysfunctional and represents a
type of complicated grief, and when it serves as a coping method associated with
a good long-term outcome.
74  Atle Dyregrov and Kari Dyregrov

Measuring grief
Several questionnaires are presently available to measure grief in children.

1 The Hogan Sibling Inventory of Grief (Hogan, 1990) has 46 items, half of
which measure positive growth. A recent version has been shortened to 20
items (Hogan et al., in review).
2 The Extended Grief Inventory developed by Layne, Savjak, Saltzman, and
Pynoos (2001) is currently being improved and renamed as the UCLA Grief
Reactions Scale (Layne, Kaplow, & Pynoos, 2011). This measures three
dimensions: traumatic grief; positive connection to the deceased; and com-
plicated grief reactions.
3 The Traumatic Grief Inventory for Children (Dyregrov et al., 2001) is built on
an early version of an adult scale, the Inventory of Traumatic Grief (Prigerson
et al., 1995).

An inspection of the two most used scales  – the Hogan scale and the UCLA
scale  – indicates little overlap. Perhaps a better strategy for scale development
would be to agree first on various subtypes of complicated grief and then develop
items that can map these dimensions.

Interventions for grief in children

The two meta-analytic studies of bereavement interventions for children have
looked not specifically at complicated grief, but at grief interventions generally.
They arrive at somewhat different conclusions. Currier, Holland, and Neimeyer
(2007) included 13 controlled studies and reported a small, non-significant, effect
size of 0.14. They concluded that the interventions do not generate the positive
outcomes found for other professional psychotherapeutic interventions, with the
exception of some studies that intervened in a time-sensitive manner for those
with a genuine need for intervention. They advocated early screening and focus-
ing interventions on high-risk groups such as those showing signs of “childhood
traumatic grief.” Rosner, Kruse, and Hagl (2010) included a total of 27 treatment
studies, and conducted two meta-analyses: one on controlled studies (n = 15) and
one on uncontrolled studies (n = 12). The effect size was 0.35 and 0.49, respec-
tively, and the authors concluded that the results favor was for bereaved children.
They reported that for uncontrolled studies larger effects were found with a longer
time interval between bereavement and intervention. Most studies have been
conducted using group treatments. There is some indication that individual and
group treatment may yield similar results (Salloum & Overstreet, 2008). Although
there are a variety of bereavement groups available for bereaved children, few
have been scrutinized for their effect on health. Although Currier and colleagues
(2007) assumed that participants with manifest symptoms and functional impair-
ment would benefit more from treatment, they found that controlled studies
showed only a trend in this direction, whereas uncontrolled studies confirmed
 CG in children  75
their assumptions. However, for the two studies that had identified children with
complicated grief, the weighted average effect size was 0.68. Inclusion of differ-
ent studies and a slightly different statistical procedure in the two meta-analyses
may explain the different results.
In general it seems that interventions that target those who struggle with symp-
toms – tertiary interventions – show most potential. A high level of symptoms is
not synonymous with a complicated grief reaction, as symptoms include various
conditions (depression, behavioral problems, anxiety, etc.), but it does indicate
that those evincing most problems in various areas benefit most from the interven-
tions studied.
Recently, interventions have been inspired by advances within cognitive–
behavioral therapy (CBT) for trauma and loss in adults, and trauma-focused grief
therapy in manualized forms has developed for children. These manuals have
proven efficient (Cohen et al., 2006; Shear, 2009). Data indicate that a trauma-
focused grief intervention can benefit children regardless of the manner in which
a parent died (McClatchey, Vonk, & Palardy, 2009). Generally, grief interventions
appear to be most successful when treatment is designed to attend to both trauma
and grief symptoms (Cohen, Mannarino, & Knudsen, 2004; Cohen et al., 2006;
Layne et al., 2008), and there are some data showing that, once trauma symptoms
have been resolved, grief symptoms may respond to brief interventions (Cohen et
al., 2006). However, although focus on the trauma aspects of the loss is important,
other forms of grief complications receive less attention and clinicians may be
less equipped to handle them, for lack of appropriate intervention strategies. A
child who is avoiding reminders at all costs will need to be helped differently from
a child who is crying all the time, yearning for the lost one.
Resource limitations and research on intervention to date justify a conservative
approach targeting service provision for those who show emotional or behav-
ioral problems. Still, a more proactive approach may be warranted as it is in
the early period following loss that adults make important decisions regarding
sharing information with their children, including them in rituals, and return to
kindergarten and school. Based on their data from suddenly bereaved adolescents,
Brent and colleagues (2009) suggest a window of opportunity in which to prevent
or attenuate further depressive episodes in youth shortly after a parent’s death.
Inclusion in funeral and memorial services serves the child’s grieving process
better than non-attendance (Silverman & Worden, 1992; Weller, Weller, Fristad,
Cain, & Bowes, 1988). However, from clinical experience we would emphasize
the importance of preparing children for these services and providing proper adult
support, as well as follow-up subsequent to the service (Dyregrov, 2008).

Strengthening parental capacity

From the research literature it seems of almost greater importance to secure con-
tinued parental capacity following a partner or child loss than to secure help for
children directly. As more than half of surviving parents (with children between
2 and 16 years old) evidence reactions indicating a probable psychiatric disorder
76  Atle Dyregrov and Kari Dyregrov
following a partner loss (Dowdney et al., 1999), or the loss of a child (Dyregrov
et al., 2003), securing early help and providing support to adult caretakers should
be a priority. Upholding or restoring parental capacity when struggling in the
aftermath of a partner or child loss may well produce results better than general
programs targeting children directly. Identifying depression, complicated grief,
and PTSD in caregivers by using screening measures may be an advisable strategy.
The importance of what has been termed “positive parenting,” consisting of
caregiver warmth and consistent discipline (Haine et al., 2006), has been identified
as an important protective resource for children through Sandler and colleagues’
program (Sandler et al., 2008). Saldinger and colleagues (2004) also found that
child-centered parenting was associated with fewer symptoms and more posi-
tive perceptions of the living parent by the child. Sandler and colleagues’ guided
preventive intervention, termed the Family Bereavement Program, which is theo-
retically based and is the most rigorously evaluated program to date, has targeted
the putative mediators of parental demoralization, negative life events, parental
warmth, and stable positive events in the family. Recently Sandler and his group
(Sandler et al., 2010) documented the long-term effects of their program in a ran-
domized controlled trial. Although the program was effective regarding intrusive
grief scores (intrusive thoughts about the loved one), it was less successful in
reducing present grief (e.g., still crying when thinking about the loved one). They
concluded that the program reduced problematic grief responses of parentally
bereaved youths over a 6-year period. However, the results of their intervention
must be regarded as modest and there seems to be room for improvement in the
model, maybe incorporating more strategies taken from trauma-focused pro-
grams. Their program is not targeting children identified with complicated grief,
but it may have the potential to prevent the development of such problems.
Tremblay and Israel (1998) mention that it may be conceivable that one can
foster too much preoccupation with grief feelings, something that can undermine
the child’s attempts to regain a positive focus on his or her experience. Parents
often bring children who do not talk, usually boys, to our clinical attention, or
they put pressure on them to talk. This happens regardless of any functional loss
in the young person. The parental expectation and pressure for overt grief from
children may place an extra burden on them (Silverman & Worden, 1993). Thus,
although an open communication climate seems to be important in families, there
is a delicate balance for caregivers and helpers not to put too much pressure on
children to talk about the death.
Although the meta-analytic studies do not indicate the need for intervention
for all, there should be a good possibility of preventing the complicated grief
resulting from inadequate parental/family handling of the death. Given the sparse
research on early family intervention we should be cautious about withholding
early help to the family, as early intervention may enhance parents’ efforts to
cope with what has happened and thereby also be a resource for their children
(Dyregrov, 2001; Dyregrov, 2002).
 CG in children  77

Elements in therapeutic help for children

From the studies adapting trauma-focused grief strategies it seems evident that
therapeutic work with bereaved children manifesting complicated grief involves
psychoeducation about normal grief and trauma reactions; building a narrative of
the events experienced; challenging misperceptions and misunderstanding about
the event or their own reactions (cognitive restructuring); targeting eventual
traumatic aspects of their loss, including handling traumatic reminders; attention
to aspects of grief, including longing, guilt, and others’ reactions; support from
others; broadening their coping skills (including ability to regulate emotions);
meaning making; and continuing of the relationship to the deceased.
Written disclosure interventions following trauma, in which individuals
express thoughts and emotions, have been associated with improved health in
adults (Pennebaker, 1997), and recently cognitive–behavioral writing therapy
using a structured form of writing has proven effective for various types of trauma
in children (van der Oord, Lucassen, van Emmerik, & Emmelkamp, 2010).
However, writing interventions have not shown the same promise following
loss experiences. Recently, more directed written exposure, especially focusing
on benefit finding, has proven more efficacious (Lichtenthal & Cruess, 2010).
The research on the use of such interventions with children and adolescents has
only just begun. A manualized written disclosure protocol for children who have
experienced disasters (Yule et al., 2005) is currently being tested out in several
countries and preliminary research indicates that it can be beneficial for bereaved
children (Kalantari, Yule, Dyregrov, Neshatdoost, & Ahmadi, 2012).

Conclusion
Being a child leads to unique challenges following a loss. It would be unfortunate
if the area of grief in children were not to establish its own knowledge base.
An insensitive extrapolation from adult complicated grief to explain complicated
grief in children, using a diagnosis established for adults, could possibly lead to an
unfortunate focus on one type of complicated grief in children. Children in devel-
opment have immature systems for emotional and cognitive regulation. They are
dependent on adults, who themselves may be grieving with limited capacity and
availability to support and provide the information that children need to cope.
Future studies must ensure that we better define different subtypes and dynamics
of complicated grief in children, develop appropriate measures to identify them,
and put in place appropriate intervention approaches to provide specific help for
various manifestations of complicated grief.

Note
1 This project was funded by the Egmont Foundation.
78  Atle Dyregrov and Kari Dyregrov

References
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2010). A prospective examination
of catastrophic misinterpretations and experiential avoidance in emotional distress fol-
lowing loss. Journal of Nervous and Mental Disease, 198, 252–257.
Bonanno, G. A., & Field, N. P. (2001). Examining the delayed grief hypothesis across 5
years of bereavement. American Behavioral Scientist, 44, 798–816.
Bowlby, J. (1963). Pathological mourning and childhood mourning. Journal of American
Psychoanalytic Association, 11, 500–541.
Brent, D., Melhem, N., Donohoe, M. B., & Walker, M. (2009). The incidence and course of
depression in bereaved youth 21 months after the loss of a parent to suicide, accident, or
sudden natural death. American Journal of Psychiatry, 166, 786–794.
Brown, E. J., Amaya-Jackson, L., Cohen, J., Handel, S., de Bocanegra, H. T., Zatta, E., et
al. (2008). Childhood traumatic grief: A multi-site empirical examination of the con-
struct and its correlates. Death Studies, 32, 899–923.
Cerel, J., Fristad, M. A., Verducci, J., Weller, R. A., & Weller, E. B. (2006). Childhood
bereavement: Psychopathology in the 2 years postparental death. Journal of the
American Academy of Child and Adolescent Psychiatry, 45, 681–690.
Coffino, B. (2009). The role of childhood parent figure loss in the etiology of adult
depression: Findings from a prospective longitudinal study. Attachment & Human
Development, 11, 445–470.
Cohen, J., Goodman, R. F., Brown, E. J., &Mannarino, A. (2004). Treatment of childhood
traumatic grief: Contributing to a newly emerging condition in the wake of community
trauma. Harvard Review of Psychiatry, 12, 213–216.
Cohen, J. A., Mannarino, A. P., & Knudsen, K. (2004). Treating childhood traumatic grief:
A pilot study. Journal of American Academy of Child and Adolescent Psychiatry, 43,
1225–1233.
Cohen, J. A., Mannarino, A. P., & Staron, V. R. (2006). A pilot study of modified cognitive–
behavioral therapy for childhood traumatic grief (CBT-CTG). Journal of American
Academy of Child and Adolescent Psychiatry, 45, 1465–1473.
Currier, J. M., Holland, J. M., & Neimeyer, R. A. (2007). The effectiveness of bereavement
interventions with children: A meta-analytic review of controlled outcome research.
Journal of Clinical Child and Adolescent Psychology, 36, 253–259.
Dowdney, L. (2000). Childhood bereavement following parental death. Journal of
Psychology and Psychiatry, 41, 819–830.
Dowdney, L., Wilson, R., Maughan, B., Allerton, M., Schofield, P., & Skuse, D. (1999).
Psychological disturbance and service provision in parentally bereaved children:
Prospective case–control study. British Medical Journal, 319, 354–357.
Dyregrov, A. (2001). Early intervention: A family perspective. Advances in Mind–Body
Medicine, 17, 9–17.
Dyregrov, A. (2008). Grief in children: A handbook for adults (2nd edn.). London: Jessica
Kingsley.
Dyregrov, A., Yule, W., Smith, P., Perrin, S., Gjestad, R., & Prigerson, H. (2001). Traumatic
Grief Inventory for Children (TGIC). Bergen, Norway: Children and War Foundation.
Dyregrov, K. (2002). Assistance from local authorities versus survivors’ needs for support
after suicide. Death Studies, 26, 647–669.
Dyregrov, K. (2009). The important role of the school following suicide: New research
about the help and support wishes of the young bereaved. Omega: Journal of Death
and Dying, 59, 147–161.
 CG in children  79
Dyregrov, K., & Dyregrov, A. (2005). Siblings after suicide: “The forgotten bereaved.”
Suicide and Life Threatening Behaviour, 35, 714–724.
Dyregrov, K., Nordanger, D., & Dyregrov, A. (2003). Predictors of psychosocial distress
after suicide, SIDS and accidents. Death Studies, 27, 143–165.
Goodman, R. F., Cohen, J., Epstein, E., Kliethermes, M., Layne, C., Macy, R. D., et al.
(2004). Childhood traumatic grief education materials. Childhood Traumatic Grief
Task Force Education Materials Subcommittee, National Childhood Traumatic Stress
Network. Retrieved September 5, 2010, from nctsnet.org/nctsn_assets/pdfs/reports/
childhood_traumatic_grief.pdf.
Gunnar, M., & Quevedo, K. (2007). The neurobiology of stress and development. Annual
Review of Psychology, 58, 145–173.
Gyurak, A., Gross, J. J., & Etkin, A. (2011). Explicit and implicit emotion regulation: A
dual-process framework. Cognition & Emotion, 25, 400–412.
Hagan, M. J., Luecken, L. J., Sandler, I. N., & Tein, J.-Y. (2010). Prospective effects of
post-bereavement negative events on cortisol activity in parentally bereaved youth.
Developmental Psychobiology, 52, 394–400.
Haine, R. A., Wolchik, S. A., Sandler, I. N., Millsap, R. E., & Ayers, T. S. (2006). Positive
parenting as a protective resource for parentally bereaved children. Death Studies, 30,
1–28.
Hogan, N. (1990). Hogan Sibling Inventory of Bereavement. In Touliatos, J., Perlmutter,
B., & Straus, M. (Eds.), Handbook of family measurement techniques (p. 524). Newbury
Park, CA: Sage.
Hogan, N. S., Schmidt, L. A., Camp, N., Barrera, M., Compas, B. E., Davies, B., et al. (in
review). Development and testing of the Hogan Inventory of Bereavement for Children
and Adolescents. Omega: Journal of Death and Dying.
Jacobs, J.  R., & Bovasso, G.  B. (2009). Re-examining the long-term effects of experi-
encing parental death in childhood on adult psychopathology. Journal of Nervous and
Mental Disease, 197, 24–27.
Kalantari, M., Yule, W., Dyregrov, A., Neshatdoost, H., & Ahmadi, S. J. (2012). Efficacy
of writing for recovery on traumatic grief symptoms of Afghan refugee bereaved ado-
lescents. Omega: Journal of Death and Dying, 65, 139–150.
Layne, C.  M., Kaplow, J., & Pynoos, R.  S. (2011). UCLA Grief Reactions Scale.
Unpublished psychological test, University of California, Los Angeles.
Layne, C. M., Saltzman, W. R., Poppleton, L., Burlingame, G. M., Pasalić, A., Duraković,
E., et al. (2008). Effectiveness of a school-based group psychotherapy program for war-
exposed adolescents: a randomized controlled trial. Journal of the American Academy
of Child and Adolescent Psychiatry, 47, 1048–1062.
Layne, C.  M., Savjak, N., Saltzman, W.  R., & Pynoos, R.  S. (2001). Extended Grief
Inventory. Unpublished psychological test, University of California, Los Angeles.
Lichtenthal, W. G., & Cruess, D. G. (2010). Effects of directed written disclosure on grief
and distress symptoms among bereaved individuals. Death Studies, 34, 475–499.
Lin, K. K., Sandler, I. N., Ayers, T. S., Wolchik, S. A., & Luecken, L. L. (2004). Resilience
in parentally bereaved children and adolescents seeking preventive services. Journal of
Clinical Child and Adolescent Psychology, 33, 673–683.
Luecken, L.  J. (2008). Long-term consequences of parental death in childhood:
Psychological and physiological manifestations. In Stroebe, M.  S., Hansson, R. O.,
Schut, H., & Stroebe, W. (Eds.), Handbook of bereavement research and practice
(pp. 397–416). Washington, DC: American Psychological Association.
80  Atle Dyregrov and Kari Dyregrov
McClatchey, I, S., Vonk, M. E., & Palardy, G. (2009). Efficacy of a camp-based interven-
tion for childhood traumatic grief. Research on Social Work Practice, 19, 19–30.
Melhem, N., Walker, M., Moritz, G., & Brent, D. A. (2008). Antecedents and sequelae
of sudden parental death in offspring and surviving caregivers. Archives of Pediatric
Adolescent Medicine, 162, 403–410.
van der Oord, S., Lucassen, S., van Emmerik, A. A. P., & Emmelkamp, P. M. G. (2010).
Treatment of post-traumatic stress disorder in children using cognitive behavioural
writing therapy. Clinical Psychology and Psychotherapy, 17, 240–249.
Pearlman, M. Y., Schwalbe, K. D., & Cloitre, M. (2010). Grief in childhood: Fundamentals
of treatment in clinical practice. Washington, DC: American Psychological Association.
Pennebaker, J. W. (1997). Opening up: The healing power of expressing emotions (rev.
edn.). New York: Guilford Press.
Pfeffer, C., Karus, D., Siegel, K., & Jiang, H. (2000). Child survivors of parental death from
cancer or suicide: Depressive and behavioral outcomes. Psycho-oncology, 9, 1–10.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K.,
et al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed
for DSM-V and ICD-11. PLoS Medicine, 6(8), e1000121. doi:10.1371/journal.
pmed.1000121
Prigerson, H.  G., Maciejewski, P.  K., Reynolds, C.  F., Bierhals, A.  J., Newsom, J.  T.,
Fasiczka, A., et al. (1995). Inventory of Complicated Grief: A scale to measure mal-
adaptive symptoms of loss. Psychiatry Research, 59, 65–79.
Riches, G., & Dawson, P. (2000). Daughters’ dilemmas: Grief resolution in girls whose
widowed fathers remarry early. Journal of Family Therapy, 22, 360–374.
Rosner, R., Kruse, J., & Hagl, M. (2010). A meta-analysis of interventions for bereaved
children and adolescents. Death Studies, 34, 99–136.
Rostila, M., & Saarela, J. M. (2011). Time does not heal all wounds: Mortality following
the death of a parent. Journal of Marriage and Family, 73, 236–249.
Saler, L., & Skolnick, N. (1992). Childhood parental death and depression in adult-
hood: Roles of surviving parent and family environment. American Journal of
Orthopsychiatry, 62, 504–516.
Saldinger, A., Cain, A., & Porterfield, K. (2003). Managing traumatic stress in children
anticipating parental death. Psychiatry, 66, 168–181.
Saldinger, A., Porterfield, K., & Cain, A.  C. (2004). Meeting the needs of parentally
bereaved children for child-centered parenting. Psychiatry, 67, 331–352.
Salloum, A., & Overstreet, S. (2008). Evaluation of individual and group grief and
trauma interventions for children post disaster. Journal of Clinical Child & Adolescent
Psychology, 37, 495–507.
Sandler, I. N., Ma, Y., Tein, J., Ayers, T. S., Wolchik, S., Kennedy, C., & Millsap, R. (2010).
Long-term effects of the family bereavement program on multiple indicators of grief
in parentally bereaved children and adolescents. Journal of Consulting and Clinical
Psychology, 78, 131–143.
Sandler, I. N., Wolchik, S. A., Ayers, T. S., Tein, J-Y., Coxe, S., & Chow, W. (2008). Linking
theory and intervention to promote resilience in parentally bereaved children. In
Stroebe, M. S., Hansson, R. O., Schut, H., & Stroebe, W. (Eds.), Handbook of bereave-
ment research and practice (pp. 531–550). Washington, DC: American Psychological
Association.
Scheeringa, M. S. (2008). Developmental considerations for diagnosing PTSD and acute
stress disorder in preschool and school-age children. American Journal of Psychiatry,
165, 1237–1239.
 CG in children  81
Schmiege, S. J., Khoo, S. T., Sandler, I. N., Ayers, T. S., & Wolchik, S. A. (2006). Symptoms
of internalizing and external problems: Modelling recovery curves after the death of a
parent. American Journal of Preventive Medicine, 31, 152–160.
Shear, M. K. (2009). Grief and depression: Treatment decisions for bereaved children and
adults. American Journal of Psychiatry, 166, 746–748.
Silverman, P. R., & Worden, J. W. (1992). Children’s reactions to the death of a parent in
the early months after the death. American Journal of Orthopsychiatry, 62, 93–104.
Silverman, P. R., & Worden, J. W. (1993). Determinants of adjustment to bereavement in
younger widows and widowers. In Stroebe, M., Stroebe, W., & Hansson, R. (Eds.),
Handbook of bereavement: Theory, research, and intervention (pp.  208–226). New
York: Cambridge University Press.
Stroebe, M. S., Hansson, R. O., Stroebe, W., & Schut, H. (2002). Handbook of bereavement
research: Consequences, coping, and care. Washington, DC: American Psychological
Association.
Taylor, L. K., Weems, C. F., Costa, N. M., & Carrión, V. G. (2009). Loss and the experience
of emotional distress in childhood. Journal of Loss and Trauma, 14, 1–16.
Tremblay, G. C., & Israel, A. C. (1998). Children’s adjustment to parental death. Clinical
Psychology: Science and Practice, 5, 424–438.
Weller, E. B., Weller, R. A., Fristad, M. A., Cain, S. E., & Bowes, J. M. (1988). Should
children attend their parent’s funeral? Journal of the American Academy of Child and
Adolescent Psychiatry, 27, 559–562.
Worden, J. W. (1996). Children and grief. New York: Guilford Press.
Worden, J. W., Davies, B., & McCown, D. (1999). Comparing parent loss with sibling loss.
Death Studies, 23, 1–15.
Worden, J. W., & Silverman, P. R. (1996). Parental death and the adjustment of school-age
children. Omega: Journal of Death and Dying, 33, 91–102.
Yule, W., Dyregrov, A., Neuner, F., Pennebaker, J., Raundalen, M., & van Emmerik, A.
(2005). Writing for recovery: A manual for structured writing after disaster and war.
Bergen: Children and War Foundation.
Part III

Diagnostic categorization
Scientific, clinical, and societal
implications
7 Prolonged grief disorder as a new
diagnostic category in DSM-5
Paul A. Boelen and Holly G. Prigerson

Introduction
A minority of bereaved individuals develops persistent, disabling, and distressing
symptoms of grief. As yet, there is no category for a bereavement related disorder
in the most frequently used classification system, the Diagnostic and Statistical
Manual for Mental Disorders (DSM). Since the mid-1990s, researchers and clini-
cians have increasingly pled for the inclusion of a syndrome of grief – by turns
referred to as pathological, complicated, and, more recently, prolonged grief dis-
order – in the DSM system. Horowitz and colleagues proposed criteria for “patho-
logical grief” in 1993 (Horowitz, Bonanno, & Holen, 1993) and refined criteria
for “complicated grief disorder” in 1997 (Horowitz et al., 1997). Comparable to
posttraumatic stress disorder (PTSD), this condition was conceptualized as con-
sisting of intrusive symptoms and signs of avoidance and failure to adapt. In that
same period, two influential studies were published by Prigerson and colleagues.
In one of these studies, Prigerson, Frank et al. (1995) differentiated symptoms of
complicated grief from bereavement-related depression and found complicated
grief to be associated with health impairments over and above depression. In
the other study, the Inventory of Complicated Grief (ICG) was introduced as a
19-item measure of complicated grief, together with data supporting the scale’s
psychometric properties (Prigerson, Maciejewski, et al., 1995).
The Prigerson, Frank et al. (1995) study was the first to provide evidence that
complicated grief is a distinct and disabling condition. From 1995 onwards, sev-
eral studies – some of which are cited below – have replicated this finding. The
second study provided researchers with a tool to study the validity and correlates
of complicated grief and is now the most commonly used and well-validated
measure of complicated grief. These initial studies culminated in publication of
“consensus criteria” for complicated grief in the late 1990s (Prigerson, Shear, et
al., 1999). According to these criteria, complicated grief encompassed symptoms
of separation distress (e.g., yearning) and traumatic distress (symptoms represent-
ing a sense of being traumatized by the loss) present to the point of functional
impairment for at least 2 months. Somewhat later, the time criterion was extended
to 6 months and symptoms of traumatic distress were no longer distinguished
from symptoms of separation distress because these symptoms were found to load
on a single dimension (Prigerson & Jacobs, 2001).
Since the publication of the DSM-IV-TR (APA, 2000) and these early stud-
ies, dozens of studies have been published supporting the diagnostic validity of
86  Paul A. Boelen and Holly G. Prigerson
complicated grief (CG), or prolonged grief disorder (PGD), as it is now termed –
henceforth abbreviated as PGD/CG. In this chapter, we will elucidate that PGD/
CG meets the definition of a mental/psychiatric disorder and, as such, should be
included in the DSM system. In addition, proposed standardized criteria for PGD/
CG will be described.

Can PGD/CG be defined as a formal disorder?

Robins and Guze (1970) described five taxonomic principles for establishing
the validity of a disorder. They proposed that diagnostic validity could be incre-
mentally improved by increasingly precise clinical description, delineation of
syndromes from other disorders, and empirical studies on underlying psychobio-
logical mechanisms. This publication guided refinement of the definition of mental
disorder in later editions of the DSM and research on psychometric properties of
existing and new syndromes. Stein et al. (2010) recently proposed a revised set of
criteria for the definition of a mental/psychiatric disorder for DSM-5, building on
this earlier work. In this section, we will argue that PGD/CG meets these criteria
(for similar reviews see Prigerson & Jacobs, 2001; Prigerson et al., 2009; Shear
et al., 2011).

A behavioral or psychological syndrome or pattern that occurs in an

individual
People suffering from PGD/CG are essentially stuck in a state of chronic grief –
grief being defined as a state of wanting what you cannot have, a sense of craving
for the deceased that does not abate with time (Prigerson & Maciejewski, 2008).
Their inability to accept the separation and reluctance to make adaptations to life
cause a continued desire to be close to their loved ones, and protest and pain
whenever they confront the impossibility of fulfilling this desire. These responses
are maintained and exacerbated by confusion about self-identity, a sense of
foreshortened future, and a sense that life is empty and meaningless without
the lost person. Although these reactions may occur in uncomplicated grief, in
PGD/CG these reactions are persistently distressing and disabling beyond the
first few months post loss. Thus, PGD/CG constitutes a clearly identifiable and
recognizable cluster of symptoms that can be reliably assessed using the ICG
(Prigerson, Maciejewski, et al., 1995) or its updated version the PG-13 (Prigerson
et al., 2009). As described below, the syndrome can be distinguished from normal/
uncomplicated grief and depression and anxiety.

Consequences of the syndrome are clinically significant distress or

disability
There is evidence from concurrent and prospective longitudinal studies that
elevated levels and clinical caseness of PGD/CG are associated with sleep dis-
turbances, suicidal ideation, increased substance misuse, reduced quality of life,
and other physical and mental health impairments (Prigerson et al., 2009; Shear
 Prolonged grief disorder in DSM-5  87
et al., 2011). Critically, many studies have shown that these associations emerge
even when controlling for the shared variance between PGD/CG and concomitant
depression and anxiety (e.g., Bonanno et al., 2007; Prigerson et al., 2009). Thus,
PGD/CG symptoms have proven to have incremental validity. That is, many
bereaved people suffering from emotional problems would be missed if clinicians
relied only on conditions currently found in the DSM. Findings of linkages of
PGD/CG symptoms with distress and disability are robust across subgroups of
bereaved people divided by age, gender, mode of death, kinship to the deceased,
and cultural background (Prigerson & Jacobs, 2001).

The syndrome is not merely an expectable response to common

stressors
Several studies have estimated prevalence rates of PGD/CG in different sam-
ples (e.g., Fujisawa et al., 2010; Kersting, Brähler, Glaesmer, & Wagner, 2011;
Morina, Von Lersner, & Prigerson, 2011). Although different criteria were used
to identify PGD/CG cases, in all these studies, prevalence rates were below 25%
indicating that meeting criteria for PGD/CG “caseness” is not the norm. In a series
of studies Bonanno and colleagues examined trajectories of grief symptoms over
time (e.g., Bonanno et al., 2002; Chapter 5 of this volume). These studies also
indicated that only a minority of mourners (≤ 25% dependent on the sample under
investigation) displayed a pattern of persistent intense grief. In fact, across stud-
ies, more than half of the people displayed a pattern of consistently low levels of
distress, suggesting that resilience and not chronic grief is the expectable response
to bereavement.
Some authors have argued that a categorical distinction between PGD/CG and
normal/uncomplicated grief  – evidence that PGD/CG symptoms are phenom-
enologically distinct from those of normal grief – should be established before
PGD/CG can become an official disorder (Hogan, Worden, & Schmidt, 2004).
However, these concepts are better described as two extremes of a single dimen-
sion of grief severity. A recent taxometric investigation confirmed this (Holland,
Neimeyer, Boelen, & Prigerson, 2009).
This is not to say that PGD/CG and normal/uncomplicated grief overlap
completely. Studies by Boelen and van den Bout (2008), Boelen and colleagues
(2003), and Prigerson, Maciejewski, and colleagues (1995) showed that scores on
the ICG were more strongly associated with quality of life impairments than were
scores on the Texas Revised Inventory of Grief (TRIG; Faschingbauer, Zisook, &
DeVaul, 1987), a measure tapping relatively benign grief reactions (e.g., crying).
These findings suggest that persistently high levels of PG/CG symptoms, but not
necessarily other grief symptoms, represent a maladaptive bereavement reaction.

The syndrome reflects an underlying psychobiological dysfunction

There is evidence that PGD/CG meets this criterion. Reynolds et al. (1999) found
that depressive symptoms but not grief symptoms declined in bereaved people
taking nortriptyline, suggesting that different biological dysfunctions underlie
88  Paul A. Boelen and Holly G. Prigerson
both conditions. Bonanno et al. (2007) found that, in mourners interviewed about
their loss, PGD/CG severity was associated with reduced heart rate and PTSD
severity with increased heart rate, a finding also pointing to a specific physiologi-
cal response. O’Connor et al. (2008; also Chapter 15 in this volume) conducted
an intriguing functional magnetic resonance imaging (fMRI) study in which they
found that women with PGD/CG, but not those without, showed reward-related
activity in the nucleus accumbens that was positively correlated with self-reported
yearning when confronted with photos of their deceased mother. This suggests
that persistent activation of reward areas in the brain is a biological correlate of
PGD/CG. Apart from these biological correlates, several psychological dysfunc-
tions have been identified as having a specific linkage with PGD/CG symptoms.
These include impairments in the ability to retrieve specific memories following
loss (Chapter 13) and negative thinking and avoidance behaviors (Chapter 16).

The syndrome is not primarily a result of social deviance or conflicts

with society
This point is an important consideration more than a defining criterion of mental or
psychiatric disorder (Stein et al., 2010). It cautions that the PGD/CG label should
never be used when grief reactions are consistent with cultural and religious
norms. The importance of cultural diversity in the expressions and manifestations
of grief is recognized in recent proposals of grief-related disorders for DSM-5
(discussed below) that include the criterion that “the bereavement reaction must
be out of proportion or inconsistent with cultural or religious norms” to qualify as
disordered (APA, 2011). Notably though, consistent psychometric performance
of PGD/CG symptom criteria has been observed across different non-Western
cultures, suggesting that it is not purely a “Western” construct (e.g., Fujisawa et
al., 2010; Morina et al., 2011).

The syndrome has diagnostic validity

The syndrome should have diagnostic validity in terms of prognostic signifi-
cance, psychobiological disruption, and response to treatment. The prognostic
significance and underlying psychobiological disruptions of PGD/CG have been
discussed above. With respect to treatment, there is growing evidence that treat-
ments specifically designed to target PGD/CG symptoms are efficacious. Shear
et al. (2005) found complicated grief treatment to be more effective in the treat-
ment of PGD/CG than interpersonal psychotherapy originally designed for the
treatment of depression. Boelen, de Keijser, van den Hout, and van den Bout
(2007; see also Chapter 16) found cognitive–behavioral grief therapy to be more
effective than supportive counseling. Thus, in these studies, interventions focus-
ing on underlying mechanisms of PGD/CG were more effective than non-specific
interventions also used for the treatment of other disorders. This indicates that, if
not adequately diagnosed, a person with PGD/CG may receive a treatment that
has insufficient effect for PGD/CG, whereas, if adequately identified, he or she
can be referred to an effective treatment.
 Prolonged grief disorder in DSM-5  89

The syndrome has clinical utility

Inclusion of PGD/CG in DSM-5 has benefits in terms of clinical utility: It aids in
the identification and treatment of a significant minority of bereaved people who
were thus far not identified at all, or diagnosed with the wrong label. Illustrative
are Shear et al.’s (2011) findings that 85% of a large sample seeking treatment for
PGD/CG had made multiple attempts to receive adequate help before receiving
effective PGD/CG-targeted psychotherapy. Clinical utility of PGD/CG is none-
theless an important area for further study (see First et al., 2004). Accordingly,
research is now being prepared that addresses the acceptability of standardized
PGD/CG criteria and their impact on clinical decision making and treatment.

No definition perfectly specifies boundaries for the concept “mental/

psychiatric disorder”
As with the criterion that “The syndrome is not primarily a result of social devi-
ance or conflicts with society” (above) this point is an important consideration
more than a defining feature of mental/psychiatric disorder (Stein et al., 2010)
and – as such – applies no more or less to PGD/CG than to other syndromes in
DSM. It could be taken as an encouragement to continue research on the nature,
duration, and severity of grief symptoms defining the boundaries between normal/
uncomplicated grief and PGD/CG.

Diagnostic validators and clinical utility should help to differentiate

the syndrome from “nearest neighbors”
Factor analytic studies support the differential diagnosis of PGD/CG. In many
different samples PGD/CG symptoms have been found to be distinct from, rather
than loading on a single dimension with, symptoms of depression, anxiety, and
other nearest neighbors (Bonanno et al., 2007; Prigerson, Frank, et al., 1995).
Moreover, as noted, PGD/CG is associated with many adverse outcomes when
controlling for these symptoms, providing further evidence that it is distinct from
its nearest neighbors.

Potential benefits of adding a new syndrome to DSM-5 should

outweigh potential harms
Two frequently reported concerns about inclusion of PGD/CG in DSM-5 include
the risk of medicalization of normal/uncomplicated grief and the stigmatization of
those receiving the diagnosis. The risk of medicalization of normality is inherent
to the DSM as a categorical system. That is, all signs and symptoms of all mental/
psychiatric disorders in DSM can also occur in “normal” individuals and they
point at a disorder only when they are present in particular combinations, at par-
ticular levels of duration and severity. Because the boundaries between normality
and psychopathology are hard to draw, there is always a risk of inflated rates of
false positive diagnosis when normal signs and symptoms are falsely labeled as
90  Paul A. Boelen and Holly G. Prigerson
markers of some mental/psychiatric disorder. All this applies to grief: PGD/CG
symptoms resemble reactions seen in normal/uncomplicated grief and there is
indeed a risk of false positive diagnosis of PGD/CG. To some extent, pathologiza-
tion of normal grief can be avoided by making a careful distinction between PGD/
CG symptoms and PGD/CG caseness. If someone has PGD/CG symptoms, this
does not necessarily mean that he or she has a mental/psychiatric disorder; this
is so only when these symptoms are present at particular levels of duration and
severity.
The risk of stigmatization is also not specific to PGD/CG. Although being
diagnosed with any DSM label presents such a risk, this does not imply that these
labels are not valid. In the case of grief, not including PGD/CG in DSM would
leave a significant minority of afflicted individuals undiagnosed. The choice of
words is important here: The label should be used to indicate not that some people
grieve “abnormally” according to some rule, but, instead, that their suffering is
so intense and causes so much distress and disability that it requires professional
care. Concerns about stigmatization have not been supported by research: in a
recent study, 98.5% of a community-based sample of bereaved people noted that
they would be willing to receive help for PGD/CG if they were diagnosed with
this condition and a similar percentage would be relieved to know they had a
recognizable condition (Johnson et al., 2009).

Recent proposals of criteria for DSM-5

In light of the research findings reviewed above, three criteria sets have recently
been proposed for inclusion in DSM-5: criteria for Prolonged Grief Disorder put
forth by Prigerson et al. (2009), criteria for Complicated Grief from Shear et al.
(2011), and criteria for Adjustment Disorder (AD) Related to Bereavement and
Bereavement Related Disorder recently proposed by the DSM-5 Anxiety Disorder
Working Group (APA, 2012).

Prolonged Grief Disorder

Prigerson developed a diagnostic algorithm for PGD. The 1999 consensus criteria
(Prigerson et al., 1999) and criteria proposed by Horowitz et al. (1997) formed
the starting point for this endeavor. Validation of the criteria proceeded through
a series of analyses, using data from 291 mostly spousally bereaved people, who
were all interviewed using the rater version of the ICG (Prigerson, Maciejewski,
et al., 1995).
In the first phase of the analyses, methods from item response theory (IRT)
were used to examine properties of 22 candidate PGD symptoms. In these analy-
ses, six items were removed because they had poor discriminative ability, mean-
ing that they were unable to distinguish reliably between different levels on the
latent dimension of PGD severity. Another four items were removed because they
showed differential item functioning, meaning that they were endorsed differ-
ently by people from different subgroups (e.g., men and women) who had similar
scores on the underlying PGD dimension.
 Prolonged grief disorder in DSM-5  91
In the second phase, a criterion standard for PGD “caseness” was determined.
Because, obviously, no such standard was available at this stage, the authors
decided to define “caseness” as the score on the PGD dimension that had the
greatest agreement with an expert clinician’s judgment of whether or not a person
represented a true case of PGD.
In Phase 3 the most parsimonious combination of symptoms that best dis-
tinguished between “cases” and “non-cases” of PGD was determined. Yearning
was specified as a mandatory symptom, because this is a hallmark symptom of
grief and, indeed, was among the items with the strongest ability to distinguish
between low and high scores on the underlying dimension of PGD severity. Using
combinatorics, a mathematical approach to study the performance of different
combinations of items, almost 5,000 different algorithms were tested. The opti-
mal, most efficient algorithm included yearning and at least five of nine additional
symptoms, described under criterion C in Table 7.1.
In Phase 4 the predictive validity for three temporal subtypes of PGD was
determined: acute PGD (meeting criteria for caseness at 0–6 months, but not at
6–12 months), delayed PGD (caseness at 6–12 months but not at 0–6 months),
and persistent PGD (caseness at 0–6 months and 6–12 months). Findings revealed
that delayed and persistent PGD, but not acute PGD, were related to adverse health
outcomes at 12–24 months (including depression and PTSD, suicidal ideation,
and poor quality of life). This phase thus provided evidence that PGD symptoms
should not be labeled as a disorder when present in the first 6 month post loss. In
the next phase, the formulation of the criteria was finalized with the addition of
criterion D, requiring that symptoms were present to a disabling degree.
The predictive validity of the proposed criteria was examined in the sixth and
last phase of the analyses. These analyses showed that those meeting criteria for
PGD at 6–12 months post loss but not meeting criteria for concomitant depres-
sion, PTSD, or generalized anxiety disorder had a significantly greater chance
than non-cases of suffering one of these disorders, suicidal ideation, functional
disabilities, and reduced quality of life at 12–24 months post loss.

Complicated Grief
Shear et al. (2011; see also Simon et al., 2011) developed another set of standard-
ized criteria for CG. To inform this process, the authors gathered data from three
groups who all completed the ICG: bereaved healthy controls (n = 95), patients
diagnosed with a mood or anxiety disorder (n  =  369), and a group of patients
presenting for treatment of CG (n = 318) that included 288 people identified as
“cases” of CG. Cases were those who had a score of at least 30 on the ICG and
who were identified as such by expert clinicians. The authors then conducted IRT,
factor, and sensitivity/specificity analyses on these data to inform their proposal
for CG criteria.
In the first step of their analyses, factor analysis on data of the full sample
(n = 782) showed that ICG items represented a single underlying dimension. In
the second step, IRT methods were used to explore the performance of items.
These analyses showed that “feeling that life is empty,” “loneliness,” and “feeling
Table 7.1  Criteria for prolonged grief disorder, complicated grief, and bereavement related disorder

Prolonged Grief Disorder (Prigerson) Complicated Grief (Shear) Bereavement Related Disorder (DSM proposal)
A Event: Bereavement (loss of a A The person has been bereaved, i.e. experienced the death A The person experienced the death of a close
significant other) of a loved one, for at least 6 months relative or friend at least 12 months earlier
B Separation distress: The bereaved B At least one of the following symptoms of persistent B Since the death at least one of the following
person experiences yearning (e.g., intense acute grief has been present for a period longer symptoms is experienced on more days than not
craving, pining, or longing for the than is expected by others in the person’s social or and to a clinically significant degree:
deceased; physical or emotional cultural environment: 1. Persistent yearning/longing for the deceased
suffering as a result of the desired, but 1. Persistent intense yearning or longing for the person 2. Intense sorrow and emotional pain because of
unfulfilled, reunion with the deceased) who died the death
daily or to a disabling degree 2. Frequent intense feelings of loneliness or as if life is 3. Preoccupation with the deceased person
empty or meaningless without the person who died 4. Preoccupation with the circumstances of the
3. Recurrent thoughts that it is unfair, meaningless, death
or unbearable to have to live when a loved one has
died, or a recurrent urge to die in order to find or to
join the deceased
4. Frequent preoccupying thoughts about the person
who died, e.g., thoughts or images of the person
intrude on usual activities or interfere with
functioning
C Cognitive, emotional, and C At least two of the following symptoms are present for C Since the death at least six of the following
behavioral symptoms: The bereaved at least a month: symptoms are experienced on more days than not
person must have five (or more) of the 1. Frequent troubling rumination about circumstances and to a clinically significant degree:
following symptoms experienced daily or consequences of the death, e.g., concerns about Reactive distress to the death
or to a disabling degree: how or why the person died, or about not being 1. Marked difficulty accepting the death
1. Confusion about one’s role in life able to manage without their loved one, thoughts of 2. Feeling shocked, stunned, or emotionally numb
or diminished sense of self (i.e., having let the deceased person down, etc. over the loss
feeling that a part of oneself has 2. Recurrent feeling of disbelief or inability to accept 3. Difficulty in positive reminiscing about the
died) the death, as if the person cannot believe or accept deceased
2. Difficulty accepting the loss that their loved one is really gone 4. Bitterness or anger related to the loss
3. Avoidance of reminders of the 3. Persistent feeling of being shocked, stunned, dazed 5. Maladaptive appraisals about oneself in relation
reality of the loss or emotionally numb since the death to the deceased or the death (e.g., self-blame)
 4. Inability to trust others since the 4. Recurrent feelings of anger or bitterness related to 6. Excessive avoidance of reminders of the loss
loss the death (e.g., avoiding places or people associated with
5. Bitterness or anger related to the 5. Persistent difficulty trusting or caring about other the deceased)
loss people or feeling intensely envious of others who Social/identity disruption
6. Difficulty moving on with have not experienced a similar loss 7. A desire not to live in order to be with the
life (e.g., making new friends, 6. Frequently experiencing pain or other symptoms that deceased
pursuing interests) the deceased person had, or hearing the voice of or 8. Difficulty trusting other people since the death
7. Numbness (absence of emotion) seeing the deceased 9. Feeling alone or detached from other people
since the loss 7. Experiencing intense emotional or physiological since the death
8. Feeling that life is unfulfilling, reactivity to memories of the person who died or to 10. Feeling that life is meaningless or empty
empty, or meaningless since the reminders of the loss without the deceased, or the belief that one
loss 8. Change in behavior due to excessive avoidance cannot function without the deceased
9. Feeling stunned, dazed or shocked or the opposite, excessive proximity seeking, e.g., 11. Confusion about one’s role in life or a
by the loss refraining from going places, doing things, or having diminished sense of one’s identity (e.g., feeling
contact with things that are reminders of the loss, that a part of oneself died with the deceased)
or feeling drawn to reminders of the person, such 12. Difficulty or reluctance to pursue interests
as wanting to see, touch, hear, or smell things to since the loss or to plan for the future (e.g.,
feel close to the person who died. (Note: sometimes friendships, activities)
people experience both of these seemingly
contradictory symptoms.)
D Timing: Diagnosis should not be D The duration of symptoms and impairment is at least 1 D The disturbance causes clinically significant
made until at least 6 months has month distress or impairment in social, occupational, or
elapsed since the death other important areas of functioning
E Impairment: The disturbance causes E The symptoms cause clinically significant distress or E Mourning shows substantial cultural variation; the
clinically significant impairment impairment in social, occupational, or other important bereavement reaction must be out of proportion or
in social, occupational, or other areas of functioning, where impairment is not better inconsistent with cultural or religious norms
important areas of functioning (e.g., explained as a culturally appropriate response
domestic responsibilities)
F Relation to other mental disorders:
The disturbance is not better
accounted for by major depressive
disorder, generalized anxiety disorder,
or posttraumatic stress disorder
94  Paul A. Boelen and Holly G. Prigerson
stunned/dazed” discriminated best between high and low scores on the underlying
CG dimension, whereas the items “experiencing pain as deceased,” “avoidance,”
and “feeling drawn to things associated with deceased” emerged as poor indica-
tors of CG.
Then, in a third step, the authors performed factor analysis on ICG scores
obtained in the subsample of CG cases. The 19 items clustered into six underly-
ing factors: (1) yearning and preoccupation, (2) anger and bitterness, (3) shock
and disbelief, (4) estrangement from others, (5) hallucinations of the deceased,
and (6) behavior change. This factor structure was used to inform the proposed
criteria (see Table 7.1). For instance, the five ICG items that clustered into the first
factor were grouped into four “symptoms of persistent intense acute grief” under
criterion B. The two ICG items of factor 2 were combined in criterion C4.
In a fourth step, the authors examined the sensitivity (proportion of actual CG
cases identified as such) and specificity (proportion of non-cases of CG identified
as such) of separate ICG items and combinations of items. Of the individual items,
“yearning and longing” was the most sensitive one, confirming its centrality to
CG. Additional analyses showed that the presence of at least one symptom from
the “yearning and preoccupation” cluster yielded the largest sensitivity, whereas
the presence of at least one symptom from the “hallucinations” cluster yielded
the lowest sensitivity. Then the authors examined in how many of the six clusters
one symptom had to be present to optimally distinguish cases and non-cases.
This analysis showed that having at least one symptom from the “yearning and
preoccupation” cluster and from two other symptom clusters yielded the optimal
identification of cases and non-cases.
In a next step, Shear et al. (2011) added several items to the CG criteria that
were not included in their quantitative analyses. First, based on evidence for an
association between CG and suicidal thinking, “suicidality” was added as part of
the separation distress criterion (criterion B). Secondly, “rumination” was added
as one of the symptoms under criterion C “because there is data for importance
of this symptom” (pp.  108–109). Finally, “emotional or physical reactivity to
reminders of the loss” was added under criterion C, because it proved important
in studies by Bonanno et al. (2007).
In keeping with Prigerson’s proposal, a time criterion described that the diag-
nosis should not be made in the first 6 months post loss. In keeping with the ICG
instruction asking for symptom presence in the preceding month, a timeframe
of at least 1 month of symptoms was required. Finally, a distress and disability
criterion was added (criterion E).

Adjustment Disorder Related to Bereavement and Bereavement

Related Disorder
The DSM-5 has embraced the idea that a disorder of grief should be included. The
current state of affairs (April 2012) is that two disorders of grief have been pro-
posed by the Anxiety Disorder Working Group. The first is called AD Related to
Bereavement in the main body of the DSM-5. The second is called Bereavement
Related Disorder, proposed for inclusion in its Appendix.
 Prolonged grief disorder in DSM-5  95
AD related to bereavement is defined as present when the person experiences
on more days than not “intense yearning/longing for the deceased, intense sorrow
and emotional pain, or preoccupation with the deceased or the circumstances of
the death, at least 12 months following the death of a close relative or friend”
(APA, 2012). Other symptoms that may qualify as indicative of the disorder are
“difficulty accepting the death, intense anger over the loss, a diminished sense of
self, a feeling that life is empty, or difficulty planning for the future or engaging in
activities or relationships” (APA, 2012).
Criteria for Bereavement Related Disorder have been proposed for the
Appendix – a part of the DSM where “unofficial disorders” are described that still
need further study. Criteria are listed in Table 7.1. Criteria for this disorder are met
if, at least 12 months after the death of the loved one, the person experiences one
symptom placed under criterion B, and 6 of 12 symptoms representing “reactive
distress” and “social/identity disruption” placed under criterion C, which cause
significant distress and impairment.

Discussion
We summarized evidence that PGD/CG meets criteria for being included as a
new condition in DSM-5. Working groups preparing DSM-5 have embraced this
idea. Although this is an important step toward recognition of a distinct disorder
of grief, critical comments can be made about the proposals for AD Related to
Bereavement and Bereavement Related Disorder. Most importantly, no studies
are yet available that have examined the validity, reliability, or dimensionality
of criteria for the proposed two criteria sets. For instance, looking at criteria for
Bereavement Related Disorder, there is no empirical validation for the 12-month
time criterion, for the distinction between symptoms of “reactive distress” and
“social/identity disruption,” and for the requirement that 6 of 12 symptoms from
criterion C need to be present to optimally identify cases.
Important also is that criteria for both proposed disorders differ considerably
from the ones that have so far been used in clinical practice and research. Thus,
their inclusion in DSM-5 will cause a significant discontinuity in clinical practice
and research (see First et al., 2004). That is, in clinical practice, those who meet
criteria for AD Related to Bereavement or Bereavement Related Disorder, based
on the proposed criteria, differ from those who were thus far diagnosed with PGD/
CG, based on Prigerson’s or Shear’s criteria, or the frequently used cut-off of a
score of > 25 on the ICG. Moreover, none of the research findings regarding the
prevalence, risk factors, and treatment of PGD/CG are directly applicable to AD
Related to Bereavement and Bereavement Related Disorder.
These problems could be averted if an empirically examined criterion set
were included in DSM-5, such as the ones proposed by Prigerson et al. (2009)
or Shear et al. (2011). This is not to say that these proposals require no further
study. For instance, criteria from Prigerson et al. are based on a relatively small
sample, mainly consisting of elderly bereaved spouses, and thus require replica-
tion. Criteria from Shear et al. (2011) seem complex (see below) and are not
clearly linked to empirical research. For instance, it is not clear what data justify
96  Paul A. Boelen and Holly G. Prigerson
inclusion of unstudied criteria (e.g., rumination). Notable too is that several items
have proven to have poor psychometric properties in prior studies; for instance,
the identification and hallucinatory symptoms represented in criterion C6 proved
to be poor indicators of PGD/CG in an earlier study by Prigerson et al. (1999).
Further, the specification that the grief is present “longer than is expected by
others in the person’s social or cultural environment” (criterion B) would seem
difficult to determine in a standardized and sufficiently reliable way.
An alternative would be to include in DSM-5 only symptom criteria that have
been assessed in prior research. For instance, if DSM-5 criteria relied on symp-
toms assessed with the ICG, this would allow for re-analyses of the many data
sets that have been collected with this scale in different countries, with different
bereaved groups. Given that symptoms such as suicidality, positive reminiscing,
or self-blame, which are all part of the criteria for Bereavement Related Disorder,
are not included in the ICG or other scales used in prior studies, there is no chance
that re-analyses of data could be helpful in testing the now proposed criteria.
Irrespective of their precise form, there is a need for standardized criteria for
PGD/CG. Such criteria would enable clinicians to identify and treat the condition,
and to receive reimbursement for this treatment. Moreover, standardized criteria
would enable researchers to study causes, consequences, and treatment of this
condition, and to compare findings between studies. Comparison of research find-
ings has so far been difficult, given that researchers have not relied on the exact
same criteria to define PGD/CG across studies. The time criterion, for instance,
has changed. This does not diminish the value of studies supporting the diagnostic
validity and clinical utility of the PGD/CG construct summarized in this chap-
ter, particularly because most of these studies used the same measure (the ICG)
and investigated PGD/CG as a continuous rather than a categorical construct.
However, the variation in descriptions of PGD/CG in prior research should be
considered in the process toward standardization of criteria and, at the same time,
emphasizes the importance of achieving such standardization.
There are many topics that need further study. For instance, researchers
should continue to search for diagnostic algorithms that best distinguish between
bereaved individuals who are and those who are not at risk for persistent health
impairments, with a time criterion that optimally balances false positives and
negatives. This algorithm should be examined across heterogeneous groups of
mourners. It is also important to strive for criteria that are parsimonious, even
though these should cover the many different forms the clinical picture of PGD/
CG may take. For example, the B3 criterion from Shear et al.’s (2011) criteria is
not a good example in this respect because it encompasses four distinct symptoms
that should be disaggregated. The algorithm of symptoms should be parsimonious
as well: needing to have only two of eight criteria (Shear) is simpler to diagnose
(and thus has stronger inter-rater reliability) than 6 of 12 criteria (Bereavement
Related Disorder). Moreover, the latter proposal produces many more variations
than the former one. Yet still, the former criteria may be too easy to meet and
thereby inflate prevalence rates. Regardless, these are all empirical issues that can
be investigated.
 Prolonged grief disorder in DSM-5  97
It is also important to distinguish symptoms of PGD/CG from the causes
and consequences of these symptoms. For instance, in accord with the broader
literature (also see Chapter 12 in this volume), rumination, included in Shear’s
criteria, is perhaps better seen as a cause than a symptom of PGD/CG. Suicidality
is perhaps more a consequence than a symptom of PGD/CG.
Notwithstanding these considerations, it seems timely to include a formal
category for PGD/CG in DSM-5. This would facilitate empirical research and
would imply recognition of the suffering of a significant minority of mourners
who experience difficulties in their process of recovery. Inclusion of a disorder
of grief in DSM-5 would imply not a pathologization of something normal but,
instead, a normalization of something that mostly is not, but sometimes is, indeed,
pathological.

References
APA. (2000). Diagnostic and statistical manual of mental disorders (4th edn. Text
Revision). Washington, DC: American Psychiatric Association.
APA. (2012). Proposed revision for adjustment disorder. Retrieved April 23, 2012, from
http://www.dsm5.org/ProposedRevisions/Pages/proposedrevision.aspx?rid=367.
Boelen, P. A., & van den Bout, J. (2008). Complicated grief and uncomplicated grief are
distinguishable constructs. Psychiatry Research, 157, 311–314.
Boelen, P. A., van den Bout, J., de Keijser, J., & Hoijtink, H. (2003). Reliability and validity
of the Dutch version of the Inventory of Traumatic Grief. Death Studies, 27, 227–247.
Boelen, P. A., de Keijser, J., van den Hout, M. A., & van den Bout, J. (2007). Treatment of
complicated grief: A comparison between cognitive behavioral therapy and supportive
counseling. Journal of Consulting and Clinical Psychology, 75, 277–284.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring, M., Sonnega, J.,
et al. (2002). Resilience to loss and chronic grief. Journal of Personality and Social
Psychology, 83, 1150–1164.
Bonanno, G. A., Neria, Y., Mancini, A. D., Coifman, D., Litz, B., & Insel, B. (2007). Is
there more to complicated grief than depression and PTSD? A test of incremental valid-
ity. Journal of Abnormal Psychology, 116, 342–351.
Faschingbauer, T.  R., Zisook, S., DeVaul, R. (1987). The Texas Revised Inventory of
Grief. In Zisook, S. (Ed.), Biopsychosocial aspects of bereavement (pp.  127–138).
Washington, DC: APA Press.
First, M. B., Pincus, H. A., Levine, J. B., Williams, J. B., Ustun, B., & Peele, R. (2004).
Clinical utility as a criterion for revising psychiatric diagnoses. American Journal of
Psychiatry, 161, 946–954.
Fujisawa, D., Miyashita, M., Nakajima, S., Ito, M., Kato, M., & Kim, Y. (2010). Prevalence
and determinants of complicated grief in the general population. Journal of Affective
Disorders, 127, 352–358.
Hogan, N., Worden, J. W., & Schmidt, L. (2004). An empirical study of proposed compli-
cated grief disorder criteria. Omega, 48, 263–277.
Holland, J. M., Neimeyer, R. A., Boelen, P. A., & Prigerson, H. G. (2009). The underlying
structure of grief: A taxometric investigation of prolonged and normal reactions to loss.
Journal of Psychopathology and Behavioral Assessment, 31, 190–201.
Horowitz, M. J., Bonanno, G. A., & Holen, A. (1993). Pathological grief: Diagnosis and
explanation. Psychosomatic Medicine, 55, 260–273.
98  Paul A. Boelen and Holly G. Prigerson
Horowitz, M.  J., Siegel, B., Holen, A., Bonanno, G. A., Milbrath, C., & Stinson, C.  H.
(1997). Diagnostic criteria for complicated grief disorder. American Journal of
Psychiatry, 154, 904–910.
Johnson, J. G., First, M. B., Block, S., Vanderwerker, L. C., Zevin, K., Zhang, B. H., &
Prigerson, H. G. (2009). Stigmatization and receptivity to mental health services among
recently bereaved adults. Death Studies, 33, 691–711.
Kersting, A., Brähler, E., Glaesmer, H., & Wagner, B. (2011). Prevalence of complicated
grief in a representative population-based sample. Journal of Affective Disorders, 131,
339–343.
Morina, N., Von Lersner, U., & Prigerson, H.  G. (2011). War and bereavement:
Consequences for mental and physical distress. PLoS One, 6, e22140.
O’Connor, M.  F., Wellisch, D.  K., Stanton, A.  L., Eisenberger, N.  I., Irwin, M.  R., &
Lieberman, M. D. (2008). Craving love? Enduring grief activates brain’s reward center.
Neuroimage, 42, 969–972.
Prigerson, H. G., Frank, E., Kasl, S. V., Reynolds, C. F., Anderson, B., Zubenko, G. S.,
Houck, P. R., George, C. J., & Kupfer, D. J. (1995). Complicated grief and bereave-
ment-related depression as distinct disorders: Preliminary empirical validation in
elderly bereaved spouses. American Journal of Psychiatry, 152, 22–30.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged Grief Disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine, 6(8): e1000121.
Prigerson, H. G., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale,
consensus criteria, and a preliminary empirical test. In Stroebe, M. S., Hansson, R. O.,
Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research: Consequences,
coping, and care (pp. 613–647). Washington, DC: APA Press.
Prigerson, H. G., & Maciejewski, P. K. (2008). Grief and acceptance as opposite sides of
the same coin: setting a research agenda to study peaceful acceptance of loss. British
Journal of Psychiatry, 193, 435–437.
Prigerson, H.  G., Maciejewski, P.  K., Reynolds, C.  F., Bierhals, A.  J., Newsom, J.  T.,
Fasiczka, A., et al. (1995). Inventory of Complicated Grief: A scale to measure mal-
adaptive symptoms of loss. Psychiatry Research, 59, 65–79.
Prigerson, H. G., Shear, M. K., Jacobs, S. C., Reynolds, C. F., Maciejewski, P. K., Davidson,
J., et al. (1999). Consensus criteria for traumatic grief. British Journal of Psychiatry,
174, 67–73.
Reynolds, C. F. III, Miller, M. D., Pasternak, R. E., Frank, E., Perel, J. M., Cornes, C., et
al. (1999). Treatment of bereavement-related major depressive episodes in later life.
American Journal of Psychiatry, 156, 202–208.
Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric ill-
ness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987.
Shear, K., Frank, E., Houck, P. R., Reynolds, C. F. III. (2005). Treatment of complicated
grief: A randomized controlled trial. JAMA, 293, 2601–2608.
Shear, M.  K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Duan, N., et al. (2011).
Complicated grief and related bereavement-issues for DSM-5. Depression and Anxiety,
28, 103–117.
Simon, N. M., Wall, M. M., Keshaviah, A., Dryman, M. T., LeBlanc, N. J., & Shear, M. K.
(2011). Informing the symptom profile of complicated grief. Depression and Anxiety,
28, 118–126.
Stein, D. J., Phillips, K. A., Bolton, D., Fulford, K. W. M., Zadler, J. Z., & Kendler, K. S.
(2010). What is a mental/psychiatric disorder? From DSM-IV to DSM-V. Psychological
Medicine, 40, 1759–1785.
8 Is complicated/prolonged grief a
disorder?
Why the proposal to add a category
of complicated grief disorder to the
DSM-5 is conceptually and empirically
unsound
Jerome C. Wakefield

Should prolonged intense grief be classified as a mental disorder in DSM-5? This

is what two research groups studying “prolonged grief disorder” (PGD) or “com-
plicated grief disorder” (CG) have proposed (Prigerson et al., 2009; Shear et al.,
2011). Rather than considering prolonged intense grief generally to be a painful
but normal-range phenomenon that is on the upper end of the continuum of grief
severity, and rather than seeing such severity as reflecting normally varying fac-
tors such as closeness of the lost relationship and the griever’s circumstances and
temperament, existential nature, and history, these researchers argue that grief
that satisfies certain symptom and duration criteria should be considered in and of
itself a psychiatric disorder.
Any biologically designed response can malfunction, so it is plausible that
there exist some grief disorders. The question is whether the proposed CG diag-
nostic criteria identify a class of such disorders with reasonable validity.
There are many versions of the CG proposal, with varying mixes of symptom
criteria and with duration requirements ranging from 2 months to 14 months.
Recent articles have tended to defend a 6-month duration, and I will focus on
these proposals. However, the DSM-5 workgroup has recently placed on the APA
website a proposed CG-like variant of adjustment disorder with a 12-month dura-
tion requirement. In addition to the standard requirements for adjustment disorder,
the bereavement-related variant is defined as follows:

[Adjustment Disorder] Related to Bereavement: For at least 12 months fol-

lowing the death of a close relative or friend, the individual experiences on
more days than not intense yearning/longing for the deceased, intense sorrow
and emotional pain, or preoccupation with the deceased or the circumstances
of the death. The person may also display difficulty accepting the death,
intense anger over the loss, a diminished sense of self, a feeling that life is
empty, or difficulty planning for the future or engaging in activities or rela-
tionships. Mourning shows substantial cultural variation; the bereavement
reaction must be out of proportion or inconsistent with cultural or religious
norms. (American Psychiatric Association, 2010a)
100  Jerome C. Wakefield
These criteria appear to require only one symptom, such as yearning for the
lost person (the symptoms in the first sentence are disjunctively linked by “or,”
and the symptoms in the second sentence “may” also be present but apparently
need not be). The rationale offered on the DSM-5 website for thus pathologiz-
ing grief lasting more than 12 months as an adjustment disorder notes “a need
for a diagnosis to characterize an individual who is having clinically significant
distress as a result of the death of a loved one” (American Psychiatric Association,
2010b), but offers no evidence or systematic argument that these conditions are
in fact disorders. Instead, it refers to the research literature on grief: “This matter
has been the subject of considerable research on abnormal mourning which has
been named, in some circles, ‘prolonged grief’ and as ‘complicated grief’ by other
researchers” (American Psychiatric Association, 2010b). This reference circularly
takes it as given that the grief researchers are correct in claiming that what they
are studying is abnormal mourning rather than intense normal mourning.
The DSM-5 also proposes for further study a separate bereavement related
disorder, defined using symptoms similar to other recent proposals. Diagnosis of
bereavement related disorder would be warranted if, at least 12 months after the
loss, the individual continued to experience yearning, intense sorrow, or contin-
ued preoccupation with the lost individual or the circumstances of the loss, plus at
least six out of the following symptoms relating to the death that cause distress or
role impairment beyond what is culturally expected: difficulty accepting the death;
feeling shocked or stunned; difficulty positively reminiscing about the deceased;
bitterness or anger related to the loss; self-blame; avoidance of reminders of the
loss; desire not to live in order to be with the deceased; difficulty trusting other
people; feeling alone or detached; feeling life is meaningless or empty; confusion
about one’s role in life; difficulty pursuing one’s interests.
The bereavement related disorder proposal includes the same 12-month dura-
tion requirement and the same sorts of symptoms as the bereavement-related
adjustment disorder proposal, but with a much more demanding symptom thresh-
old. Anyone who qualified as disordered under “bereavement related disorder”
or similar CG proposals would certainly qualify as disordered under “adjustment
disorder related to bereavement.” So, the DSM-5 plans to pathologize all those
cases that would be pathologized as “complicated grief” and more. The question
of whether the specified conditions are really disorders is thus now of more than
theoretical interest.
I review here the arguments for the disordered status of these likely newly
pathologized conditions. Regarding terminology, the DSM-5’s label of “bereave-
ment related disorder” is insufficiently specific because there are many disorders
other than disordered grief itself that can be related to bereavement, ranging from
major depression to heart failure. Consequently, I use the term “complicated
grief” (CG).

Bowlby’s warning
John Bowlby’s work set the stage for most current work on loss, and CG research-
ers cite his attachment theory as a framework for thinking about grief (Shear et
 Is complicated/prolonged grief a disorder?  101
al., 2007; Shear & Shair, 2005). I preface my discussion of recent grief research
with a reminder of Bowlby’s warning about seeing normal grief as briefer or less
severe than it is.
Bowlby was concerned that clinicians and researchers might misconstrue the
potential severity and duration of normal grief and thus pathologize this gradual
and difficult process. He thus cautioned:

Loss of a loved person is one of the most intensely painful experiences any
human being can suffer .  .  . [T]here is a tendency to under-estimate how
intensely distressing and disabling loss usually is and for how long the
distress, and often the disablement, commonly lasts. Conversely, there is a
tendency to suppose that a normal healthy person can and should get over
bereavement not only fairly rapidly but also completely. Throughout this
volume I shall be countering those biases. Again and again emphasis will be
laid on the long duration of grief. (Bowlby, 1980, p. 8)

What did Bowlby have in mind when he emphasized “the long duration of
grief”? According to Bowlby, widows who fully recover:

are more likely to take two or three years to do so than a mere one .  .  . I
emphasize these findings, distressing though they are, because I believe that
clinicians sometimes have unrealistic expectations of the speed and complete-
ness with which someone can be expected to get over a major bereavement.
(p. 101)

As I will show, current grief researchers seem to have forgotten Bowlby’s warning.

Three arguments based on the DSM’s definition of mental

disorder
I begin by briefly considering three arguments for the pathological status of pro-
longed grief that are linked in one way or another to the DSM’s definition of
mental disorder.

Impairment as the criterion for disorder

Prigerson and colleagues sometimes claim that the essence of disorder is impair-
ment (e.g., “Thus, impairment is the criterion for determining pathology”;
Prigerson & Maciejewski, 2006, p. 15). This criterion is so invalid that it would
imply that sleep and pregnancy are disorders. Moreover, normal grief can be
highly distressing and role impairing, so impairment alone cannot distinguish
normal from disordered grief. In support of the impairment claim, Prigerson and
Maciejewski cite the DSM’s definition of mental disorder, which requires that,
to be a disorder, a condition must be “associated with present distress (e.g., pain-
ful symptom) or disability (i.e., impairment in one or more important areas of
functioning)” (American Psychiatric Association, 2000, p. xxi). However, this is
102  Jerome C. Wakefield
presented by the DSM as a necessary condition for disorder, not as sufficient by
itself. Prigerson and Maciejewski fail to cite the definition’s additional require-
ments (see below).

Disorder as statistical deviance

Those arguing that CG is a disorder frequently fall back on statistical criteria that
rely on the degree of severity of grief symptoms as a “gold standard” supporting
disorder attributions, whether setting the threshold at the upper 50% (Horowitz
et al., 1997) or 20% (Prigerson et al., 1997, 2009). Yet it is plainly a fallacy to
identify statistical deviance with disorder (Stroebe et al., 2000; Wakefield, 1992a).
Deviant features need not be disorders (even normal grief is “deviant” relative
to typical functioning, and those with greater intensity of normal responses will
always by definition be statistically deviant from the majority), and common fea-
tures need not be normal.
The statistical approach may be inspired by the DSM’s statement in its defini-
tion of disorder that, for a condition to be a disorder, it “must not be merely an
expectable and culturally sanctioned response to a particular event, for example,
the death of a loved one” (American Psychiatric Association, 2000, p. xxxi). This
clause eliminates certain socially sanctioned, expectable conditions from disorder
status, but it does not imply that every unexpectedly intense response is a disorder.
There is an especially powerful reason for dismissing the statistical argument
in the CG debate. The alternative hypothesis to the CG disorder claim is that the
identified conditions are by and large simply the upper end of the severity con-
tinuum of normal grief. Consequently, the statistical argument begs the question,
because the competing hypotheses – the normal and the disordered construal of
the identified conditions – equally assert that the specified conditions are statisti-
cally deviant.

The argument that CG satisfies DSM’s definition of mental disorder

Prigerson’s group repeatedly claims that CG satisfies the DSM’s definition of
mental disorder, and quotes chapter and verse. The problem is that these quo-
tations consistently leave out a crucial sentence in the definition that states the
“dysfunction” requirement that a harmful (impairing, distressing) condition, to
be a disorder, must result from a dysfunction: “Whatever its original cause, it
must currently be considered a manifestation of a behavioral, psychological, or
biological dysfunction in the individual” (p. xxi). The centrality of the dysfunc-
tion requirement and an elaboration of “dysfunction” as failure of biologically
designed processes to perform their functions is pursued in my “harmful dysfunc-
tion” analysis of disorder (Wakefield, 1992b, 1999, 2006).
So, to satisfy the DSM’s definition of mental disorder, the CG proposal’s fun-
damental challenge is to show that the conditions identified as CG are due to
dysfunction. CG researchers have attempted to meet this challenge. I consider the
 Is complicated/prolonged grief a disorder?  103
three most prominent arguments they offer for the dysfunction etiology of CG:
the distinctive symptom argument, the derailment of grieving argument, and the
predictive validity argument.

CG as categorical pathology: the distinctive symptom

argument
It is sometimes claimed that CG’s symptoms are qualitatively different phe-
nomena from those of normal grief, lending plausibility to the claim that its
symptoms are pathognomonic for a disorder with a distinct etiology. However,
this appealing notion has not turned out to fit the facts. Misleading statements
suggestive of such distinct symptoms continue to appear in the literature (e.g.,
“CG is characterized by a unique pattern of symptoms following bereavement”;
Lichtenthal, Cruess, & Prigerson, 2004, p. 637), but CG researchers have mostly
come to acknowledge that there is no clear qualitative difference between the
symptoms of CG and normal grief and have consequently rejected this as a
test of validity: “The issue is not whether the symptoms sort themselves into
seemingly pathological versus seemingly normal symptom clusters” (Prigerson
& Maciejewski, 2006, p. 15).
A glance at the proposed DSM diagnostic criteria above, ranging from yearn-
ing and disbelief to lack of interest in usual activities, reveals that all of the
proposed symptoms of CG occur in intense acute grief. In 1998, a panel of
experts acknowledged that there are no “smoking gun” symptoms revealing
a pathological grief process: “The panel acknowledged that a wide range of
‘symptoms’ that occur after a loss could be considered within normal limits”
(Prigerson et al., 1999, p. 68). For example, when Prigerson et al. (1997) used a
symptom intensity criterion for CG that identified the top 20% at 6 months post
loss as disordered, they found that 57% of the grief-stricken met the criterion at
2 months post loss.
Shear and Mulhare (2008) claim that CG differs qualitatively from normal grief
in being accompanied by rumination and avoidance (Table 2, p. 664). However,
these symptoms are not in fact unique to CG. Shear and Mulhare’s definition of
rumination is extraordinarily broad (even worrying about your prolonged grieving
qualifies), and their list of normal-grief symptoms includes the rumination-like
item “thoughts and memories of the deceased are prominent and preoccupying.”
Regarding avoidance, grieving individuals normally use avoidance to regulate the
degree of painful sadness (Stroebe & Schut, 1999), and research indicates that
avoidance occurs during acute stages of grief (Bonanno, Papa, Lalande, Zhang, &
Noll, 2005, Figure 3, p. 92).
The view that CG cannot be symptomatically distinguished from acute grief
was confirmed by a recent taxometric study by a group of leading CG researchers.
They concluded: “As a whole, these taxometric analyses offered little support for
a categorical conceptualization of normal grief and PGD” (Holland, Neimeyer,
Boelen, & Prigerson, 2009, p. 198).
104  Jerome C. Wakefield

CG as derailed grief: the interminability argument

Grief varies enormously from individual to individual and according to circum-
stances. Recent research challenges traditional stage models of grief and suggests
that grief is an idiosyncratic process of integrating a loss (Bonanno & Kaltman,
1999), suggesting that any durational and intensity CG cut-off points are arbitrary.
Those with lengthier grieving processes may be dealing with closer relationships
or deeper existential issues, for example, and simply experiencing longer integra-
tion trajectories rather than suffering from dysfunctions.
CG proponents have come up with an ingenious and powerful reply to this
objection. They argue that, as a matter of empirical fact, once grief goes on for
the specified 6- to 12-month duration, then it goes on more or less indefinitely.
According to this argument, normal grief surely should have a trajectory of recov-
ery over time to be normal, so interminable intense grief strongly suggests that
something has gone wrong with the grief process: It has been “derailed” from its
biologically designed course. Such “derailment” implies internal dysfunction that
is blocking the natural course of the recovery process, and thus justifies attribution
of disorder:

Complicated grief, unlike normal or uncomplicated grief, is not a self-limited

process. (Prigerson et al., 1995, p. 23)

In some bereaved individuals . . . the intense pain and distress festers, can go
on interminably (as “complicated grief”) . . . Once established, complicated
grief tends to be chronic and unremitting. (Zisook & Shear, 2009, pp. 67, 69)

CG entails harmful dysfunction in that a normal healing process has been

derailed. (Shear et al., 2011, p. 105)

According to this argument, the primary reason for the validity of the 6-month
or 12-month duration threshold for complicated grief is that symptoms that exist
intensely at 6 or 12 months tend to persist indefinitely after that rather than to
diminish with time, on account of a derailed or frozen or otherwise malfunction-
ing grieving process. The alternative hypothesis is simply that at 6 or 12 months,
although there may be some cases of frozen grief processes, most symptoms are
part of a longer but still normal grieving process that is moving towards resolution
at a slow pace. The question is: What do the data actually show? Do they confirm
or falsify the “derailment/interminability” hypothesis?
The “interminability” claim is bewildering in light of the history of grief
research. As Bowlby makes clear, many individuals continue to heal after 6
months or a year. Case after case in classic texts portrays individuals with severe
grief reactions at 1 or 2 years grappling with deep issues triggered by the loss and
eventually progressing to recovery.
For example, Parkes and Weiss (1983) describe the difficult course of grief
of Mrs. Webley, a woman in her early thirties with a 9-year-old daughter, whose
 Is complicated/prolonged grief a disorder?  105
husband, on whom she was greatly dependent, died of diabetes. Her acute grief
immediately after the death was severe, including disbelief, intense yearning, and
difficulty with usual chores. Mrs. Webley’s situation was less acute but not much
improved a year after the loss: “When we saw Mrs. Webley at the end of the
first year of her bereavement [i.e., 13 months post loss], she appeared depressed
and apathetic . . . Despite the passage of the year, her husband was constantly in
Mrs. Webley’s thoughts.” Although she had gotten over the disbelief (“I have got
myself to knowing now that he’s gone”; p. 140), she was despairing (“You’re in a
lost world”) and unmotivated (“I don’t feel like working any more. I go in because
I force myself to go in”; p. 140), and still missed and thought about her husband
frequently (“But I do miss him terribly. I’m still involved with him too much . . .
I keep saying, Why did you have to go?”; pp. 140–141). Yet she was gradually
experimenting with greater autonomy: “I know the decisions are wrong, but I’m
trying to learn how to do it myself” (p. 141).
Mrs. Webley at 1 year without doubt would have met currently proposed crite-
ria for complicated grief. Did she have an interminable mental disorder in which
the grieving progress was derailed? Or was she simply dealing with greater and
deeper challenges than most? Parkes and Weiss continue:

We saw Mrs. Webley again at three years after her husband’s death, two years
after the interview in which she made the comments quoted above. Much
had happened in her life. She had begun seeing another man, had become
pregnant by him, and had given birth to a second daughter. She hoped the
man she was seeing would marry her . . . Mrs. Webley was no longer lonely.
She still at times thought of her husband, but she no longer did so constantly.
(pp. 141–142)

Mrs. Webley clearly needed a relationship to feel happy. The loss of her
husband left her not only bereaved, but lonely and adrift, experiencing chronic
distress. Is it a disorder to be lonely and to desperately need to find someone
with whom to share your life, and to dwell meanwhile on the past? She had the
capacity to change and succeeded in doing so, but it took a few years and she was
still in intense grief at the 13-month mark. Perhaps the critical point was not her
suffering but the degree of progress she had made, and the beginnings of growth
that were discernible. Although Mrs. Webley might have benefited from more
support and therapeutic help, it seems doubtful that she or other individuals with
such trajectories are best served by pathologizing their experiences and placing
them within the category of the mentally disordered. Indeed, it is possible that
such an approach could “derail” such individuals from the hard work they need to
do to change their circumstances and themselves to create a new life.
To mention a more famous example, Viederman (1995) notes that the 29-year-
old Richard Feynman, later to become a Nobel laureate in physics, wrote to his
beloved former wife, then 2 years dead, that he was unable to experience anything
for other women. He subsequently remarried happily and lived an extraordinarily
productive life with a well-known contagious joie de vivre (p. 2).
106  Jerome C. Wakefield
These are not isolated cases. The interminability argument is falsified by
the very research data on which CG proponents rely. No doubt there are some
interminable cases of intense grief. Shear’s group, although supporting a 6-month
durational threshold for CG (Shear et al., 2011), reports that the mean time post-
loss of the clinical sample recruited to refine its criteria was 5.5 years (Simon et
al., 2011). The evidence indicates that the proposed criteria for CG do not in fact
identify such an interminably grieving group.
For example, Horowitz et al. (1997) measured 30 symptoms of grief at 6
months and 14 months post loss in a sample of 70 bereaved individuals. Rather
than symptoms at 6 months being persistent, results indicated that “The frequency
of these symptoms declined significantly in the interval between 6 months and
14 months after the death of a significant other” (Horowitz et al., 1997, p. 909).
For example, here are the changes from 6 to 14 months in the percentages of the
sample judged to have severe levels of six symptoms from Horowitz et al.’s crite-
ria: unbidden memories, 72% vs. 42%; strong yearning, 58% vs. 35%; emotional
spells, 47% vs. 20%; feeling alone and empty, 59% vs. 38%; avoids reminders
of the deceased, 44% vs. 17%; and loss of interest in important activities, 62%
vs. 19%. In another study, Prigerson et al. (2008; see also Maciejewski, Zhang,
Block, & Prigerson, 2007) analyzed longitudinal data on grief and depressive
symptoms, documenting the rise and fall of various symptoms over time. The
results reveal gradually decreasing levels of symptoms over the first 18 months of
grieving (see Fig. 8.1 in Prigerson et al. 2008, p. 169). For example, 5–6 months
post loss the average occurrence of yearning is almost daily; after a year, about
every other day; after almost 2 years, weekly. Prigerson et al. (2008) attempt to
explain their rationale for calling the CG cases “chronic” and “persistent” despite
decreasing symptom levels:

[T]hose diagnosed with PGD revealed persistently high levels of grief that
did not drop to the level of those without PGD throughout the study observa-
tion period . . . Thus, the PGD diagnosis after 6 months postloss identified a
group of bereaved individuals who would remain persistently grief stricken.
(p. 172)

However, the fact that the most symptomatically severe at 6 months remained
higher in symptoms over time than those who were less symptomatic at 6 months
just shows that those higher in symptoms at 6 months tend to stay higher in
symptoms later on as well. They are the more intense responders. This form of
“persistence” does not imply lack of improvement in symptoms over time. It just
implies lack of change of relative position as the group overall improves.
Prigerson et al. also claim that, because indicators of grief peak at about 6
months, after that the symptoms are likely to represent pathology. This is a bewil-
dering claim. After the point of greatest average intensity of normal symptoms,
there inevitably will be a period as the normal symptoms gradually subside – but
the symptoms remain predominantly normal. This absurd argument would imply
that every individual taller than the 5′9″ average height must be suffering from a
growth hormone pathology.
 Is complicated/prolonged grief a disorder?  107
In sum, the available data decidedly falsify the “persistence” claim for those
having CG-level symptoms at 6–18 months. Beyond that interval, persistence
remains relatively unstudied. The “derailment” and “interminability” claims as
applied to 6- to 12-month-threshold CG are scientific myths.

CG as risk factor: the predictive validity argument

Predictive validity is currently Prigerson and colleagues’ central argument for CG
being a pathology:

[T]he symptoms of CG are associated with and predictive of substantial

morbidity (e.g., depression, suicidal ideation, high blood pressure), adverse
health behaviors (e.g., increased smoking, alcohol consumption, poor sleep),
and quality of life impairments. Thus, the symptoms are indicative of pathol-
ogy . . . What our results demonstrate is that the set of CG symptoms that
we have identified . . . are predictive of many negative outcomes and that is
the basis for distinguishing them from normal grief symptoms. (Prigerson &
Maciejewski, 2006, p. 15)

Prigerson et al. in their definitive 2009 study attempt to show not only that CG
differs from intense normal grief (the issue on which I focus), but also that CG is
different from other common disordered responses to loss and stress, particularly
major depressive disorder (MDD), posttraumatic stress disorder (PTSD), and gen-
eralized anxiety disorder (GAD). Prigerson et al. in their final predictive validity
analyses eliminate individuals with one or more of these other disorders, and test
the predictive validity of “pure” CG (only 3.3% of the bereaved sample). I focus
on these analyses as most relevant, so henceforth CG will refer to “pure” CG.
In their bereaved sample, Prigerson et al. (2009) compare the risk for four
negative outcomes at 12–24 months of those who qualify for CG at 6–12 months
(i.e., they have the most severe symptoms) with those who do not. For CG versus
other grievers at 6–12 months, the outcomes at 12–24 months are other mental
disorders (MDD, GAD, PTSD), 28.6% vs. 3.4%; suicidal ideation, 57.1% vs.
10.1%; functional disability, 71.4% vs. 35.9%; and poor quality of life, 83.3% vs.
14.7%, respectively (Table 4, p. 9). Note that the initial 6- to 12-month baseline
levels of the predicted variables are not controlled or reported in these com-
parisons. Before evaluating these empirical results, I comment on the conceptual
status of the predictive validity argument.

Conceptual problems with the predictive validity argument

Prigerson et al. argue that, if CG predicts statistically heightened negative out-
comes, then it must be pathological. This is a grossly invalid inference; later nega-
tive outcomes are not always evidence of earlier disorder. The fact that individuals
who grieved more intensely had increased negative outcomes no more implies
that their grief was disordered than does the fact that there are heightened risks
with pregnancy imply that pregnancy itself is a disorder. Risk permeates life and
108  Jerome C. Wakefield
does not itself imply dysfunction. Risk management to avoid negative outcomes
is one thing, diagnosis and treatment of disorder another, and confusing the two
spuriously pathologizes life’s many difficulties.
In mounting a “risk” argument for CG’s disordered status, Prigerson et al.
(2009) might be construed as relying on the “risk” clause of the DSM’s definition
of mental disorder, which states that a disorder must be associated either with
present distress or impairment “or with a significantly increased risk of suffering
death, pain, disability, or an important loss of freedom” (American Psychiatric
Association, 2000, p. xxxi). However, the definition does not say that risk is suf-
ficient for disorder. In addition, the risk must be caused by a dysfunction. Because
risk is omnipresent in life, risk is evidence of disorder only if the presence of a
dysfunction can be independently supported.
Using predictive validity for disorder validation has the inherent weakness that
both disorders and non-disorders can have negative outcomes of similar kinds.
For example, bad marriages with resultant chronic high stress are known to pre-
dict various later physical ailments, but that has not moved the DSM to accept
bad marriages as mental disorders. Shortness and homeliness within their normal-
range domains are known to have negative impacts on a variety of outcomes, yet
neither is a disorder.
Another limitation of predictive validity is that the statistical associations on
which claims of predictive validity for a group are based are sometimes due to
relatively small numbers of individuals within the group experiencing negative
outcomes. It is then questionable whether pathologizing the entire group is justi-
fied based on associations due to a minority. For example, coughing is statisti-
cally associated with an increased likelihood of tuberculosis, but no one suggests
pathologizing everyone who is coughing, because most coughing is a normal
biologically designed adaptive response to abnormal environmental factors such
as dust in the air. The statistical association between coughing and tuberculosis is
based on a few individuals whose coughs are due to tuberculosis. Those individu-
als – not necessarily the others – suffer from pathology. To support the case for
CG at 6–12 months being a disorder, the negative outcomes measured at 12–24
months must plausibly imply that the condition at 6–12 months involved a dys-
function of grief processes, not just intense normal grief.
The use of risk of negative outcomes to support disorder attribution is suf-
ficiently problematic that a committee considering DSM-5 revision of the defini-
tion of mental disorder proposed eliminating the “risk” clause:

Regarding the phrase pertaining to “increased risk”, risk factors are important
to bear in mind and perhaps even to treat . . . At the same time, we would
note that disorder and risk factors should not be conflated . . . We therefore
tentatively suggest simplifying this criterion by omitting the phrase on risk.
(Stein et al., 2010, p. 1762)

Prigerson’s “risk” argument suffers from precisely the fallacy identified by

the committee. “Treatment” of risk factors may be warranted, but raises different
 Is complicated/prolonged grief a disorder?  109
issues from treatment of disorder. The DSM lists non-disordered conditions that
are frequently targets of clinical intervention under what have come to be known
as the “V Codes” because of their diagnostic codes in the manual, and risk without
dysfunction belongs there.

Mortality
Turning to negative outcomes of CG, Prigerson et al. (2009) mention in passing a
CG-as-disorder argument not addressed in their data, namely the mortality associ-
ated with grief. However, they assume without evidence that CG will encompass
the majority of such deaths. To the contrary, it is known that the most significant
mortality increase occurs in the first days and weeks post loss, decreasing almost
to population levels by about 6 months post loss (Buckley, McKinley, Tofler, &
Bartrop, 2010). That is before CG would be diagnosed. Some continued excess
mortality may be related to pre-existing physical conditions or lifestyle tenden-
cies exacerbated by stress, not requiring a CG disorder explanation (Parkes &
Prigerson, 2010).

Physical disorders
Some studies have shown that CG predicts small increases in some later physical
problems, such as high blood pressure or cardiovascular disease and even cancer.
These results are based on few cases, and causality has not been established.
Moreover, all mental and physical stresses are risk factors for such disorders.
Running for a bus and having a marital argument both raise the risk of a heart
attack, yet neither is a disorder. There is no reason to attribute such outcomes to
disordered grief as opposed to intense normal grief.

Other mental disorders

Of four outcomes examined in the 2009 paper, I set aside development of another
mental disorder. The identified mood and anxiety disorders are of high prevalence
(their validity can be questioned as well; Horwitz & Wakefield, 2007), and have
many high-stress predictors that are not disorders. The fact that CG, like many
other stressful circumstances, contains a risk of developing these stress-related
disorders does not make it a disorder, any more than losing a job and marital
dissolution, which also increase risk for these disorders, are themselves disorders.

Functional impairment and quality of life

By definition, those qualifying for CG had higher symptom severity than those
with non-CG at 6–12 months. As noted, one can expect the more intense respond-
ers with highest 6- to 12-month grief symptoms generally to stay highest at 12–24
months, even if symptoms subside on a lengthy trajectory. Thus, in consider-
ing whether 12- to 24-month outcomes support an inference to 6- to 12-month
110  Jerome C. Wakefield
dysfunction, the “dysfunction” hypothesis must be compared with the alternative
hypothesis that higher grief severity at 6–12 months can explain the results at
12–24 months.
Prigerson et al.’s (2009) measures for functional impairment and quality of
life were intentionally formulated with low thresholds to maximize the chances
of finding an effect. For example, “Individuals with at least ‘some difficulty’ with
at least one of the 14 tasks (e.g., bathing) were considered functionally impaired
in order to make the measure sensitive to impairment in a highly functioning
sample” (p. 3). So, if an individual sometimes did not engage in just one activity
(e.g., walking a mile) because of grief symptoms, that was enough to qualify as
functionally impaired. This is problematic because normal grief symptoms are
often moderately impairing as a result of withdrawal, reduced energy and motiva-
tion, and focus on grieving.
Moreover, one would expect those higher in grief symptoms, disordered or not,
to experience a lower quality of life. Quality of life in itself is a dubious indica-
tor of disorder, especially among those emotionally affected by death of a loved
one. Bonanno et al. (2005) show that the related measure of self-perceived health
status increases dramatically from 4 to 18 months as grief distress decreases. One
would expect quality of life to track grief distress as well.
So, the quality of life and functional impairment outcomes might mean nothing
more than that the CG group, having had higher grief intensity at 6–12 months,
still had high enough intensity at 12–24 months to display negative outcomes
generally associated with emotional suffering, including lower subjective qual-
ity of life and moderate functional impairment. The modest thresholds for being
“positive” on these outcomes are consistent with normal grief and do little to
address the disorder question.

Suicidal ideation
The outcome variable most often cited as evidence of CG pathology is suicidal
ideation. As noted, of CG versus non-CG at 6–12 months, 57.1% versus 10.1%,
respectively, were positive for suicidal ideation at 12–24 months. To measure
this variable, “Positive responses to one or more of the four Yale Evaluation of
Suicidality screening questions were categorized as having suicidal ideation”
(Prigerson et al., 2009, p. 3). The study’s measure of suicidal ideation was aimed
at maximizing sensitivity; any positive (non-zero) responses to questions concern-
ing feelings about living versus dying, wish to live, wish to die, and thoughts of
killing oneself were considered “positive” for suicidal ideation (Holly Prigerson,
personal communication, November 5, 2010).
The preamble to the questions notes that “Sometimes people with [grief] feel
that this experience has affected their feelings about living.” Such reaching for
positive responses would be justified in a screening instrument where sensitivity
is paramount, but this study attempts to validate the presence of disorder, and
reaching for positive responses regarding, for example, “feelings about living”
is straying far from clearly pathological terrain. A weaker will to live in the face
 Is complicated/prolonged grief a disorder?  111
of real loss and sustained emotional suffering is not the same as suicidal ideation
indicative of pathology. One would expect individuals with higher levels of feel-
ings of meaninglessness, lack of role clarity, impoverished friendship networks
and interests, and yearning for their lost loved one (all CG symptoms) to be more
likely to entertain the thought that life may not be worth living, quite aside from
any pathology.
However, the most serious validity problem in inferring earlier disorder from
later suicidal ideation arises from the wording of the one question asking directly
about suicidal ideation: “In light of [the loss], have you ever had thoughts of
killing yourself?”, with possible positive replies “yes” or “possibly.” An “ever”
question logically implies that the positives will increase with time, because one
never exits from a positive answer that one has “ever” had such thoughts. As time
goes on, those who continue to have intense suffering and who thus may tran-
siently think about whether suicide would be preferable are added to the number
of positives. (The number who already had such thoughts at 6–12 months is not
presented.) Indeed, a “positive” on this question does not indicate current suicidal
ideation at all; someone whose only suicidal ideation was 6 months earlier would
still appropriately answer “yes.”
The likelihood of false positives for suicidality is increased even further by
“possibly” being a positive response, encompassing those who have only tran-
siently or vaguely entertained the notion of ending it all. Many people under stress
entertain such thoughts. For example, in one high school screening for suicidal
thoughts over the past 3 months, 10% responded positively, yet the stability of such
answers over a mere 8-day test–retest period was low. The designers of the instru-
ment observed that “Poor test–retest reliability could be related to the ephemeral
nature of suicidal ideation and depressive feelings among teens” (Shaffer et al.,
2004, p. 77). The same is likely to be true of those suffering from grief, and the
“possibly” option exacerbates the problem. So, the increased “suicidal ideation”
among the CG-diagnosed group could be a near-tautologous result of the way
this question is phrased plus the CG group’s greater distress and the inevitable
desperate thoughts that occur during lengthy periods of suffering. Taking all the
problems together with this and the other outcomes, there are no grounds for
plausibly inferring an earlier disorder from these outcome measures.

Conclusion
I identified several arguments presented to support the claim that intense grief
lasting 6–12 months or more is pathological, rather than a severe normal variant,
and thus that CG should be added as a new category of disorder to the DSM-5.
These arguments fail either because of conceptual flaws or because they are con-
trary to the research evidence. The current CG proposals are thus scientifically
unwarranted.
Granting that in principle there are grief disorders, my analysis suggests the
need for far more stringent diagnostic criteria than those proposed, if massive
false positive diagnoses are to be avoided. The proposals’ relabeling of millions
112  Jerome C. Wakefield
of intensely grieving individuals as disordered is serious enough, but a much
greater false positives problem would occur subsequently when the diagnosis is
on the books and out of the research setting, and general practitioners are treating
grief in a public sensitized by pharmaceutical advertising to think about grief as
a disorder.
The research effort supporting the CG proposal, clearly motivated by com-
passion for the grief-stricken, has been a win–win gamble for grief studies. The
research has highlighted clinical phenomena previously ignored or marginalized,
and added immeasurably to our understanding of grief’s symptoms and trajectory.
The contributions of the Prigerson and Shear CG research groups is a watershed
in grief studies, irrespective of the merits or the outcome of the proposal to add
CG as a new category of disorder to DSM-5.

References
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental
disorders, DSM-IV-TR (4th edn., text revision). Washington, DC: APA.
American Psychiatric Association. (2010a). Adjustment disorders. Retrieved April 28, 2011,
from http://www.dsm5.org/ProposedRevisions/Pages/proposedrevision.aspx?rid=367.
American Psychiatric Association (2010b). Adjustment disorders: Rationale. Retrieved
April 28, 2011, from http://www.dsm5.org/ProposedRevisions/Pages/proposedrevi-
sion.aspx?rid=367#.
Bonanno, G. A., & Kaltman, S. (1999). Toward an integrative perspective on bereavement.
Psychological Bulletin, 126(6), 760–776.
Bonanno, G. A., Papa, A., Lalande, K., Zhang, N., & Noll, J. G. (2005). Grief process-
ing and deliberate grief avoidance: A prospective comparison of bereaved spouses and
parents in the United States and the People’s Republic of China. Journal of Consulting
and Clinical Psychology, 73, 86–98.
Bowlby, J. (1980). Loss. Sadness and depression (Attachment and loss, Vol. 3). New York:
Basic Books.
Buckley, T., McKinley, S., Tofler, G., & Bartrop, R. (2010). Cardiovascular risk in early
bereavement: A literature review and proposed mechanisms. International Journal of
Nursing Studies, 47(2), 229–238.
Holland, J. M., Neimeyer, R. A., Boelen, P. A., & Prigerson, H. G. (2009). The underlying
structure of grief: A taxometric investigation of prolonged and normal reactions to loss.
Journal of Psychopathology and Behavioral Assessment, 31, 190–231.
Horowitz, M.  J., Siegel, B., Holen, A., Bonanno, G. A., Milbrath, C., & Stinson, C.  H.
(1997). Diagnostic criteria for complicated grief disorder. American Journal of
Psychiatry, 154, 904–910.
Horwitz, A.  V., & Wakefield, J.  C. (2007). The loss of sadness: How psychiatry trans-
formed normal sorrow into depressive disorder. New York: Oxford University Press.
Lichtenthal, W., Cruess, D., & Prigerson, H. G. (2004). A case for establishing complicated
grief as a distinct mental disorder in DSM-V. Clinical Psychology Review, 24, 637–662.
Maciejewski, P., Zhang, B., Block, S., & Prigerson, H. (2007). An empirical examina-
tion of the stage theory of grief. Journal of the American Medical Association, 297(7),
716–722.
Parkes, C. M., & Prigerson, H. G. (2010). Bereavement. New York: Routledge.
 Is complicated/prolonged grief a disorder?  113
Parkes, C. M., & Weiss, R. S. (1983). Recovery from bereavement. New York: Basic Books.
Prigerson, H. G., Bierhals, A. J., Kasl, S. V., Reynolds, C. F., Shear, M. K., Day, N., et al.
(1997). Traumatic grief as a risk factor for mental and physical morbidity. American
Journal of Psychiatry, 154, 616–623.
Prigerson, H. G., Frank, E., Kasl, S. V., Reynolds, C. F., Anderson, B., Zubenko, G. S., et
al. (1995). Complicated grief and bereavement-related depression as distinct disorders:
Preliminary empirical validation in elderly bereaved spouses. American Journal of
Psychiatry, 152, 22–30.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine, 6(8), e1000121.
Prigerson, H.  G., & Maciejewski, P.  K. (2006). A call for sound empirical testing and
evaluation of criteria for complicated grief proposed for DSM-V. Omega, 52, 9–19.
Prigerson, H. G., Shear, M. K., Jacobs, S. C., Reynolds, C. F., Maciejewski, P. K., Davidson,
J. R. T., et al. (1999). Consensus criteria for traumatic grief: A preliminary empirical
test. British Journal of Psychiatry, 174(1), 67–73.
Prigerson, H.  G., Vanderwerker, L.  C., & Maciejewski, P.  K. (2008). A case for inclu-
sion of prolonged grief disorder in DSM-V. In Stroebe, M. S., Hansson, R. O., Schut,
H., & Stroebe, W. (Eds.), Handbook of bereavement research and practice: Advances
in theory and intervention (pp.  165–186). Washington, DC: American Psychological
Association.
Shaffer, D., Scott, M., Wilcox, H., Maslow, C., Hicks, R., Lucas, C. P., et al. (2004). The
Columbia Suicide Screen: Validity and reliability of a screen for youth suicide and
depression. Journal of the American Academy of Child and Adolescent Psychiatry, 43,
71–79.
Shear, M. K., Monk, T., Houck, P., Melhem, N., Frank, E., Reynolds, C., & Sillowash, R.
(2007). An attachment-based model of complicated grief including the role of avoid-
ance. European Archive of Psychiatry and Clinical Neuroscience, 257, 453–461.
Shear, M.  K., & Mulhare, E. (2008). Complicated grief. Psychiatric Annals, 38(10),
662–670.
Shear, M. K., & Shair, H. (2005). Attachment, loss, and complicated grief. Developmental
Psychobiology, 47, 253–267.
Shear, M.  K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Duan, N., et al. (2011).
Complicated grief and related bereavement issues for DSM-5. Depression and Anxiety,
28, 103–117.
Simon, N. M., Wall, M. M., Keshaviah, A., Dryman, M. T., LeBlanc, N. J., & Shear, M. K.
(2011). Informing the symptom profile of complicated grief. Depression and Anxiety,
28(2), 118–126.
Stein, D. J., Phillips, K. A., Bolton, D., Fulford, K. W. M., Sadler, J. Z., & Kendler, K. S.
(2010). What is a mental/psychiatric disorder? From DSM-IV to DSM-V. Psychological
Medicine, 40(11), 1759–1765.
Stroebe, M., & Schut, H. (1999). The dual process model of coping with bereavement:
Rationale and description. Death Studies, 23(3), 197–224.
Stroebe, M., van Son, M., Stroebe, W., Kleber, R., Schut, H., & van den Bout, J. (2000).
On the classification and diagnosis of pathological grief. Clinical Psychology Review,
20, 57–75.
Viederman, M. (1995). Grief: Normal and pathological variants. American Journal of
Psychiatry, 152(1), 1–4.
114  Jerome C. Wakefield
Wakefield, J. C. (1992a). The concept of mental disorder: On the boundary between bio-
logical facts and social values. American Psychologist, 47, 373–388.
Wakefield, J. C. (1992b). Disorder as harmful dysfunction: A conceptual critique of DSM-
III-R’s definition of mental disorder. Psychological Review, 99, 232–247.
Wakefield, J. C. (1999). Evolutionary versus prototype analyses of the concept of disorder.
Journal of Abnormal Psychology, 108, 374–399.
Wakefield, J. C. (2006). The concept of mental disorder: Diagnostic implications of the
harmful dysfunction analysis. World Psychiatry, 6, 149–156.
Zisook, S., & Shear, K. (2009). Grief and bereavement: What psychiatrists need to know.
World Psychiatry, 8, 67–74.
9 Lessons from PTSD for
complicated grief as a new DSM
mental disorder
Jan van den Bout and Rolf J. Kleber

Introduction
In 1980 Post-Traumatic Stress Disorder (PTSD) was introduced as a diagnos-
tic category in the Diagnostic and Statistical Manual of Mental Disorders (3rd
Edition) (DSM-III; American Psychiatric Association, 1980). It is the only disor-
der in the current DSM nomenclature in which a cause of the disorder is included,
namely a (psycho)trauma (or more than one). The proposed DSM-5 disorders
Prolonged Grief Disorder (Prigerson et al., 2009) and Complicated Grief (CG;
Shear et al., 2011) resemble PTSD in that a necessary cause is also specified,
namely the loss of a significant other, in particular the death of a loved one. Both
could fit in a general rubric of event-related disturbances, as was already argued in
the 1990s (Brom, Kleber, & van den Bout, 1993). For that reason it is illuminating
to review what the scientific, clinical, and societal consequences have been of
the introduction of PTSD in the DSM system. What have been the pros and cons
of the construct PTSD since its inclusion in DSM? And, after more than three
decades of research and clinical work, what lessons are there to be learnt for the
concepts of prolonged grief disorder and complicated grief, in case one of them
(or an amalgam of these) is ‘canonized’ in DSM-5?

PTSD and DSM

The introduction of PTSD in 1980 in DSM-III has undoubtedly led to an impres-
sive bulk of empirical research covering a wide range of traumatic experiences.
As the editor of the Journal of Traumatic Stress Paula Schnurr (2010) mentioned,
the number of publications on trauma grew over ninefold between 1980–1984 and
1995–1999 (from 930 to 8,606). Not only in absolute figures, but also relatively
speaking, publications on trauma became more frequent: within the anxiety disor-
ders literature between 1981 and 1985, 16% of the publications were on trauma,
growing to 38% between 2000 and 2005 (Boschen, 2008).
Before going into detail on the vicissitudes of the construct PTSD in science
(specifically in psychiatry and [clinical] psychology) and in society (including
persons who experienced a traumatic event), we first provide a short historical
account of the introduction of this construct.
116  Jan van den Bout and Rolf J. Kleber

A historical note on the introduction of PTSD in DSM

A system like the DSM does not fall from heaven nor is it a ‘necessary’ conse-
quence of scientific findings. Rather, it is the result of human endeavour and is
ultimately created through consensus meetings with experts. In these consensus
meetings scientific findings of course play a role, but frequently other concerns are
decisive, sometimes because there simply are no scientific data to rely on. PTSD
is in this respect a nice illustration: the experts were to a large extent influenced
by the Zeitgeist. What were these societal influences that led to the introduction
of PTSD in DSM?
In the 1960s and 1970s, the USA fought a long war in Vietnam, a war that
was far from popular in the home country. During many years numerous anti-war
demonstrations were organized. When U.S. veterans returned to their home com-
munities, they did not receive a warm welcome (as the veterans of the Second
World War had received). For instance, they were often called ‘baby killers’. Partly
as a result of this unfriendly welcome and the idea of having fought for nothing
(the USA ultimately lost this war), a substantial number of veterans suffered from
chronic distress, severe feelings of malaise, marital problems, alcoholism, and
social instability after discharge from the army. A group of anti-war psychiatrists
(most notably Chaim Shatan and Robert Lifton, who earlier had written about the
victims of the atomic bombs on Hiroshima and Nagasaki) and Vietnam veterans
(most notably Charles Figley) proposed a relatively new view on these problems:
the veterans exhibited a ‘post-Vietnam Syndrome’ (a term first used by Shatan
in 1972). That is, their health problems were the result of the traumatic experi-
ences in Vietnam. The central new assumption was that ‘– in contrast to standard
psychiatric opinion  – problematic temperaments, personalities, and behaviours
were not shaping a subject’s response to trauma but were products of it’ (McHugh
& Treisman, 2007, p.  215). In short, the atrocities experienced during combat
in Vietnam were considered as the main cause of the veterans’ difficulties, but
one should realize that the poor homecoming climate and the lack of recognition
reinforced the problems.
At the same time, the American Psychiatric Association (APA) was working on
a fundamental revision of the DSM system, a revision in the direction of a revival
of nosological entities (van Praag, 2000). Thus emphasis was laid on a description
of concrete symptoms belonging to a circumscribed disorder. The ‘post-Vietnam
Syndrome’ ideally fulfilled the criteria for such a disorder: it appeared to be a clear
disorder with highly identifiable symptoms. In an effective lobby of the Vietnam
Veterans Working Group of psychiatrists the editorial committees of DSM (most
notably Nancy Andreasen) were persuaded to adopt the concept as a new disorder.
Once this was accomplished, three groups worked together to let the diagnosis
grow after it had been adopted (see McHugh & Treisman, 2007). Psychiatrists
realized that not only the atrocities in Vietnam but also civilian traumata could
lead to a similar disorder: they were also ‘victims of [traumatic] stress’. Vietnam
veterans realized that a diagnosis of war-caused disorder was more honourable
than a personality disorder, substance abuse, or an adjustment disorder. Somewhat
 Lessons from PTSD for CG  117
later this was also realized by victims of other severe events, such as rape, physical
abuse, and accidents. Finally, the Veterans Administration was eager to embrace
the diagnosis as an excellent justification for its work with Vietnam veterans.
Essential for the conceptual formulation of PTSD was the scientific work of
the psychiatrist Mardi Horowitz. He conducted a series of elegant experiments in
the early 1970s, in which he analysed the central responses to traumatic events.
In 1976 he published his seminal book Stress Response Syndromes (Horowitz,
2003), in which he formulated disordered forms of coping with traumatic stress.
This work formed the basis for the content of the PTSD concept.
To conclude: primarily societal developments (a draining war that was finally
lost, resulting in veterans with, among others, severe feelings of malaise) led to
the introduction of PTSD in DSM-III.

Critical voices on the mental disorder PTSD

As already indicated, the introduction of PTSD gave rise to much research.
Consequently, we now know much more about traumatic stressors and their nega-
tive consequences. Theoretical models on the origin and maintenance of posttrau-
matic symptomatology have been formulated and tested. Posttraumatic pathology
has become one of the important foci of psychology and psychiatry, leading to
more funding of research on trauma. Effective therapy methods for PTSD have
been developed and tested, and much has been learnt about the possibilities as
well as limitations (including noxious effects) of prevention and treatment (e.g.
debriefing). These scientific and societal accomplishments have been accompa-
nied by the foundation of societies for professionals in the trauma field, such as
the International Society for Traumatic Stress Studies (ISTSS), with branches in
nearly all continents. In addition, there are specialized trauma journals, such as the
Journal of Traumatic Stress and the European Journal of Psychotraumatology.
By and large, in view of the just-mentioned accomplishments, one can state that
the introduction of PTSD in DSM has been a nearly unprecedented success.
However, during the 30 years of its existence there have been dissenting voices,
which seem to become only louder with the coming of DSM-5. These critical
voices relate to the characterization of traumatic events and resulting symptoms
and to the suggestion that PTSD may be primarily a product of social construction
(Young, 2004). In a rather sceptical review, Rosen and Lilienfeld (2008) stated
that virtually all core assumptions of the construct of PTSD lack compelling or
consistent empirical support. Other authors (for example, McNally, 2004, 2009)
have also analysed the conceptual limitations. In the next paragraphs we will dis-
cuss the most salient issues raised by PTSD critics. These issues are also highly
relevant to a future inclusion of complicated grief in DSM-5.

The stressor criterion

The central tenet of the concept of PTSD is that a special class of events (criterion
A: the stressor criterion) is causally related to a certain set of reactions (criteria
118  Jan van den Bout and Rolf J. Kleber
B–D: the symptom criteria). It is evident that criterion A events are not a sufficient
cause for PTSD (because even the most severe traumatic events do not always
lead to PTSD), but these events are seen as a necessary condition for PTSD. To
put it simply: the concept of PTSD implies that there can be no PTSD without
having experienced traumatic events. However, what does research show about
this relation?
People can show the full range of PTSD symptoms without having experi-
enced criterion A events. These symptoms appear after non-criterion A events
such as marital disruption. As an illustration: Mol et al. (2005) conducted an open
population study among nearly 1,500 respondents. They found that non-traumatic
life events can generate at least as many PTSD symptoms as ‘traumatic’ events,
casting doubt on the specificity of traumatic events as a cause of PTSD. It thus
appears that full-blown PTSD can develop in the absence of the occurrence of a
criterion A event. This result could be caused by the fact that several PTSD symp-
toms are rather generally defined. Nevertheless, the specificity of the concept of
trauma (defined as extreme experiences of powerlessness and disruption) is lost.

Dose–response relationship between traumatic stressors and

symptoms
A corollary of the foregoing is the assertion that the more ‘heavy’ the event is
(‘traumatic’ or otherwise), the ‘heavier’ the symptoms are. Research shows that
this hypothesis has not always been corroborated (see Rosen & Lilienfeld, 2008).
Factors other than the event per se may be more important. Traumatic events
are often not the largest contributor to outcome. Thus, criterion A events are not
necessary and surely not sufficient to instigate PTSD; but, in addition, factors
extraneous to the traumatic event may contribute more variance to clinical out-
come than the event itself.

Changes in the stressor criterion

In DSM-III the stressor criterion A was relatively simple: ‘Existence of a
recognizable stressor that would evoke significant symptoms of distress in almost
everyone’ (American Psychiatric Association, 1980, p. 238). In later editions the
objective part of the stressor criterion was expanded, while at the same time a
subjective criterion was added. In DSM-IV it reads:

The person has been exposed to a traumatic event in which both of the
following were present: (1) the person experienced, witnessed, or was con-
fronted with an event or events that involve actual or threatened death or
serious injury, or a threat to the physical integrity of self or others (2) the
person’s response involved intense fear, helplessness, or horror. (American
Psychiatric Association, 1994, p. 427)

‘Confronted with’ could mean ‘hearing about’, and thus the implications of this
expansion were considerable. Watching television and being confronted with a
 Lessons from PTSD for CG  119
disaster or hearing or reading narratives of survivors could potentially lead to
PTSD. This led to quite a lot of debate. Spitzer, First, and Wakefield (2007) plead
for more stringent criteria in the interest of ‘Saving PTSD from itself’. Most likely,
this indirect confrontation will be dropped from the criterion definition in DSM-5.
A consequence of this change in criterion A for the relation between PTSD and
CG can be illustrated with the results of a community survey in Detroit by Breslau
et al. (1998). Criterion A stipulates: ‘confronted with . . . the actual . . . death’. Of
the total 2,181 respondents, 60% indicated that they had experienced the sudden,
unexpected death of a friend or relative. No fewer than 31% of the persons with
PTSD had experienced such a sudden loss. So, loss became more or less part of
the PTSD definition. One could comment that with this change in criterion A the
clinical picture of PTSD has changed dramatically: to a large extent PTSD now
also consists of grief phenomena, ignoring salient differences between posttrau-
matic and grief reactions (Boelen, van den Hout, & van den Bout, 2006; Raphael,
Martinek, & Wooding, 2004; see also Chapter 10 of this volume).

The issue of comorbidity

One of the appealing aspects of PTSD is that it strikes mental health practitioners
and laymen (including patients) as a self-evident syndrome: traumatic stressors
can be so vehement and intruding that ‘flashbacks’ and forms of avoidance and
denial are only ‘natural’. However, research has clearly shown that a person con-
fronted with a traumatic experience can suffer from more mental disorders than
PTSD. There is substantial comorbidity observed, so much that a ‘pure’ PTSD
disorder appears to be a rarity. So, is PTSD simply a sort of amalgam of other
disorders and is the diagnosis of PTSD, although appealing, in fact redundant?
For instance, one third to one half of those suffering from PTSD also show a clini-
cal depression (Keane, Marshall, & Taft, 2006). That someone may suffer from
depression at the same time is not surprising when one looks at the symptoms
that are listed for both PTSD and depression: loss of interests, social indiffer-
ence, trouble sleeping, and difficulty concentrating. Other concomitant disorders
regularly occurring with PTSD are substance abuse disorders, complicated grief,
and generalized anxiety disorder.

Overemphasis or simulation of complaints?

If a patient reports having experienced a traumatic experience, most health
care practitioners usually will not check whether this event actually happened.
However, it was found among Vietnam veterans in the United States suffering
from combat-induced PTSD that some of them had experienced no combat in
Vietnam and that some of them had not even been in Vietnam (McNally, 2004).
The symptoms were heavily emphasized, exaggerated, or even simulated.
Undoubtedly, the legal possibility of financial compensation for combat-related
PTSD played a large role. As mentioned before, PTSD is the only diagnosis in
DSM in which an external cause is mentioned. Therefore it is a diagnosis that can
be used to get (financial) compensation. It might have secondary gains. The extent
120  Jan van den Bout and Rolf J. Kleber
of this problem is not known; it occurs probably more after experiences for which
objective data about the (f)acts are not easily available.

Society and trauma/PTSD

The term (psycho)trauma has become very popular in society. The concept of
(psycho)trauma and, in its vestige the disorder PTSD, has been enormously
popularized, even to the extent that nowadays PTSD is a term that is known by
many readers of newspapers. Also, most people now ‘know’ and realize that very
distressing events can lead to severe mental effects. In some countries the official
mental disorder of PTSD has led to the possibility of compensation for damage
for clients with PTSD. Among professionals working with ‘traumatized’ clients
the mental disorder PTSD has led to more status (and to more funding). However,
the enormous popularity of the terms trauma and PTSD has some consequences
that might be or may become problematic, for laymen without traumata as well as
for people who experienced traumatic events. These effects may also occur after
the introduction of a diagnosis of complicated grief.
First of all, there is the danger of ‘trauma inflation’. The term trauma is often
used in everyday speech to point to almost every negative event, such as the tran-
sition from kindergarten to elementary school. This inflation is hardly present in
scientific and professional literature (the Journal of Traumatic Stress publishes
mainly about war, disasters, and abuse). Nevertheless, the term has crept into
common usage as a label for minor problems and daily hassles as well. The focus
on severe events and their long-lasting and serious disturbances is lost and the
term runs the risk of becoming meaningless.
Perhaps more importantly, by and large the phrase posttraumatic stress disor-
der seems to imply that a traumatic event is the main determinant of the (stress)
reactions and malfunctioning of a person (or client/patient). Above we saw that
such a link is far from self-evident. The danger is that such a diagnosis obscures
other possible pathogenic features (Rosen & Lilienfeld, 2008) that might be
important, such as prior and current vulnerabilities, which could therapeutically
be targeted. Admittedly, for some clients such a presentation is pleasant: they can
attribute their emotional problems to an external event (and one might speculate
that this is a major reason for the popularity of the term trauma). However, it is
questionable whether such a notion is beneficial for them: their own responsibility
for and influence on their prior and current emotional problems no longer seem
at stake.
Somewhat paradoxically, recent developments in the field of traumatic stress
underline the point that is made here. More and more emphasis is laid on the
finding that some people do quite well under ‘traumatic’ circumstances: they
are resilient (Bonanno, 2004; Brom & Kleber, 2009; see also Chapter 5 of this
volume). In fact these resilient people are the best proof of the assertion that it is
not a traumatic event itself, but the way a person conceptualizes and reacts to a
‘traumatic’ event, that determines his or her emotions and behaviours.
 Lessons from PTSD for CG  121

DSM: a pragmatic system with drawbacks

Do we really need elaborate and precise definitions of mental disorders? In psy-
chiatric epidemiology some decades ago, a strong movement arose to come to
unequivocal descriptions of mental disorders. Researchers needed unambiguous
criteria to assess mental disorders, especially in the case of collaborative research
in different countries. Also for mental health practitioners across countries it
became desirable to use the same diagnostic phrases for the same mental phenom-
ena. Thus, both research diagnostic criteria and classification systems for mental
disorders were formulated (or revised).
As noted earlier, characteristic of the third edition of DSM in 1980 was a
revival of the nosological way of looking at psychiatric problems. Subsequently,
DSM-IV went further along those lines. The nosological disease model holds
that ‘psychiatric disorders are characterised by a particular symptomatology,
course, outcome, treatment response . . .’ (van Praag, 2000, p. 151). Van Praag
rightly stresses that for mental health problems/disorders ‘[t]his disease model is
a premise, not an empirically based concept. The premise holds that disturbances
of the psychic “apparatus” manifest themselves as discrete entities’ (ibid.). In the
DSM-IV the very first feature of the definition of mental disorder is: ‘A clini-
cally significant behavioral or psychological syndrome or pattern that occurs in an
individual’. We note that pattern seems a much weaker expression than syndrome,
whereas the word syndrome appears weaker than disorder, the term that is com-
monly used in the case of somatic disorders. Recently Stein et al. (2010) reviewed
the DSM-IV definition of mental disorder and proposed some changes with an eye
on the coming DSM-5. They explicitly state that no definition perfectly specifies
precise boundaries for the concept of mental disorder. Nevertheless, the current
diagnostic criteria for the mental disorders are quite strict.
The DSM system, originally primarily a tool for researchers, who of course
need some form of categorization and classification, has subsequently been used
as a tool for mental health professionals assessing their clients. It is a tool that is
prototypical for the medical model: there is a disorder or there is no disorder. It
is also a tool with severe implications: only when a disorder has been established
should reimbursed help be given. In other words: a mental health professional can
bill the insurance company only in cases where there is a diagnostic code, which
thus requires a diagnosis in terms of some form of ‘official’ disorder.
The strict focus on having to decide whether a particular client/patient ‘has’ a
disorder (or syndrome or pattern) or not is understandable in some settings (a yes
or no decision makes sense in various insurance, medical, or forensic contexts),
but it also has disadvantages. Tucker wrote in an editorial in the American Journal
of Psychiatry some words which almost every clinician will recognize:

by using DSM, sometimes clinicians are treating the diagnosis and not the
patient . . . We are not looking at or studying the patient’s phenomenology
any more, but are looking for the symptoms needed to make the diagnosis
. . . Accurate observation and the story of the patient must be included in our
122  Jan van den Bout and Rolf J. Kleber
diagnostic processes. All are necessary for the effective care of our patients,
which in the long run, is what it is all about. (Tucker, 1998, p. 161)

Cooper (2004; see also Chapter 2, this volume) has argued cogently that the
answer to the question ‘what is a (mental) disorder?’ is determined not only by
scientific findings but also by social factors (such as the perceived need that
people with severe emotional problems should get help) and financial factors
(such as – a consideration of the mental health practitioner – the wish to make it
easier for patients to obtain reimbursement, or – a consideration of the insurance
companies – the wish to restrict the number of patients who qualify), irrespec-
tive of the question whether their mental problems fall within the definition of
a mental disorder. She gives several examples to underline her argument, one
from a (former) DSM Working Group on PTSD which noted that ‘requiring a
minimum duration before a diagnosis of PTSD could be made might reduce help-
seeking behaviour as well as reimbursement for treatment’ (Davidson, Foa, et
al., 1996; cited in Cooper, 2004). Cooper stresses also that social factors (such
as the perceived need that people with severe emotional problems should get
help) are determinants for including certain severe emotional problems within
DSM, which reflects a widespread conviction among many health practitioners:
the DSM should furnish a licence for doing reimbursed interventions with clients
who need psychotherapeutic help, and undoubtedly there are intensely grieving
people for whom this is the case.
To recapitulate: (1) although a precise definition of mental disorder is lacking,
the diagnostic criteria for the separate mental disorders are quite strict; (2) there
are clear indications that the very focus on classification issues (consisting of
two steps: is there a disorder and, if so, what is the disorder?) may hinder the
diagnostic and the therapeutic process; and (3) social, clinical, and financial
factors influence what is considered a disorder and/or influence what is included
in the DSM or not. Hence, a consideration for including CG within the DSM
could be that almost all grief experts are of the opinion that some form(s) of
complicated grief exist(s), but they have as yet dissenting opinions on the issue
of whether the currently proposed DSM-5 grief disorders cover their ideas on
CG adequately.
More generally speaking, one should perhaps go a step further. Perhaps less
attention and effort should be paid to deciding whether something is a disorder or
not. Instead, the efforts and skills and time of clinicians should rather be devoted
to alleviating human mental suffering, irrespective of the answer to the question
whether this suffering is an indication of a mental disorder or not (see also Bolton,
2008, for a philosophical treatise on this subject). Additionally, it should be
remembered that, if there is an official disorder, there is frequently no one-to-one
relation between a certain disorder and intervention. For example, for the large
majority of Axis I disorders (including the frequent mood and anxiety disorders)
the intervention of choice is mainly cognitive–behavioural therapy. This being the
case one could wonder what the merits are of all the efforts towards classification,
when the chosen intervention hardly has a relation with the classification, possibly
 Lessons from PTSD for CG  123
because ‘transdiagnostic’ processes are more important than previously thought.
An additional consideration is that, in the case of some ‘official’ disorders, hardly
anything can be done, whereas in cases where no ‘official’ disorder has been
diagnosed there are frequently effective intervention activities available. Also,
most mental health practitioners assess whether a patient ‘has’ a DSM disorder
primarily for administrative reasons (partly resulting from the insurance system),
not for therapeutic reasons.

Complicated grief and DSM-5: explorations

We now return to our main question: what lessons are to be learnt for the disorder
complicated grief from the adventures of PTSD in the 30 years since its incep-
tion? In our section on critical voices on the mental disorder PTSD we discussed
issues that have a clear relevance to a possible introduction of CG in DSM-5.
Should complicated grief (or whatever other term might be chosen) be included
in the DSM system or not? If the answer is yes: what can be predicted about the
future of this new disorder? Let us try to extrapolate from the findings about
PTSD to the eventual inclusion of CG in DSM.

Acknowledgement of suffering
The inclusion of the mental disorder of PTSD led to much research. The field of
traumatic stress is booming in science. The same can be said of society: trauma(tic
stress) is a much discussed topic in the media and among the public. That traumatic
events sometimes can lead to severe mental problems is now common wisdom in
society, and some victims acquire a set of symptoms that can be labelled as PTSD.
Although we are not aware of research on this matter, persons with PTSD seem
not to be stigmatized for having this psychiatric label. On the contrary, there is
nowadays recognition of the (pathological) effects of traumatic events. Similarly,
for patients with severe mental problems in the aftermath of severe events, the
‘official’ disorder PTSD has made reimbursement for psychotherapeutic help
possible.
Including CG in the DSM might have similar consequences. Currently, the
dominant attitude towards grief appears to be: grief is part of life, so please restrict
the whining. Inclusion of CG would lead to the realization that complications in
grief are a serious matter, which can take the form of a real disorder and need
special care. Such a label provides patients with acknowledgement. They will be
taken (more) seriously by professionals.

Growth of effective interventions

Since 1980 numerous controlled studies have been published on interventions
for PTSD. In the literature, the suggestion has repeatedly been made that experts
on grief interventions should look at what has been found and accomplished in
the field of interventions for PTSD. This suggestion is somewhat ironic, because
124  Jan van den Bout and Rolf J. Kleber
before 1980 there were already grief intervention studies that were in fact proto-
typic for the later studies on PTSD (e.g. Ramsay, 1977). However, in the decades
since 1980 far fewer controlled intervention studies have been published on CG
than on PTSD. Based on the experiences with PTSD, it seems safe to conclude that
an official mental disorder in DSM would lead to more and better-suited interven-
tion approaches for people suffering from CG. In addition, it will probably also
lead to better guidelines about whether one should refrain from intervention. For
instance, some acute interventions after trauma (such as emotional debriefing) that
originally appeared to be sensible have been found inadvisable (Bisson, Brayne,
Ochberg, & Everly, 2007). The parallel between grief and trauma is intriguing
in this respect: treatments of disorders have been found to be effective, but brief
forms of preventative counselling have not always been found to be effective
(Schut & Stroebe, 2010).
Above we mentioned that in the case of a mental disorder in which the cause
is implied (PTSD) there is the danger that such a diagnosis obscures prior and
current vulnerabilities, which could be targeted therapeutically. By implication,
this would possibly also apply for CG. However, just as this possible danger de
facto is no problem in concrete therapeutic activities (because psychotherapy is
always about changeable factors within a person), it is to be expected that this will
not be a problem for the treatment of CG either.

The introduction of a disorder as self-fulfilling prophecy

As noted earlier, since the inclusion of PTSD in DSM-III, public attention to
trauma and the consequences of traumatic experiences has grown enormously.
From a societal perspective, PTSD’s inclusion has led to consequences that gen-
erally seem more positive than negative. However, scientifically speaking, the
picture is less positive. Robert Spitzer, one of the main protagonists of the (new)
DSM approach in 1980, wrote with two noted scholars: ‘Since its introduction
into DSM-III in 1980, no other DSM diagnosis . . . has generated so much contro-
versy in the field as to the boundaries of the disorder, diagnostic criteria, central
assumptions, clinical utility, and prevalence in various populations’ (Spitzer, First,
& Wakefield, 2007, p. 233). The question arises: if we had known then what we
now know about the construct of PTSD, would PTSD have been accepted as an
official mental disorder? The answer to this question may possibly be yes, but we
are not certain. However, the paradox is: we know what we know now, for the
very reason that PTSD was introduced as a disorder (resulting in much research).
And the paradox goes even further: because of its societal success and impact, the
somewhat ramshackle construct PTSD will undoubtedly survive in the DSM-5.
In addition, the case of PTSD has shown that its inclusion in DSM has led to
research that eventually led to changes in PTSD criteria. Along similar lines, one
could predict that inclusion of CG in DSM with (some amalgam of) the proposed
diagnostic criteria could lead to fruitful research, on the basis of which better
diagnostic criteria could be formulated.
Although grief is part of human life, the phenomena of normal and abnormal
grief are hardly known among laypeople, medical and health professionals, or
 Lessons from PTSD for CG  125
clergy (Wortman & Silver, 2001). The introduction in DSM of some form of CG
will undoubtedly lead to more knowledge among professionals and laypeople
about normal and pathological grief (see also Stroebe et al., 2000). However, it
will also lead to more demand for ‘services’ for people with CG, be it by means
of psychotherapy or by medication, and thus to more reimbursements. On the
other hand, as Kendler (2010) rightly notes, assessing a DSM disorder provides
the possibility, but by no means the requirement, that treatment should be started.
Sometimes it is more appropriate to wait (and perhaps eventually abstain) than to
begin with interventions.

Blurring normal and abnormal adaptation to trauma: the danger of

medicalization
One unfortunate result of the introduction of PTSD was that all consequences of
extreme events were primarily considered from the perspective of PTSD, in short,
from a psychopathology perspective. However, only a minority of the persons
who experience a traumatic event develop PTSD. For instance, prevalence rates
of PTSD vary in studies of natural and technological disasters between 2% and
45% (Neriah, Nandi, & Galea, 2008). The reported percentages vary on account
of differences in event characteristics, instruments, and length of time since the
event. Unfortunately, as a result of the focus on pathology, normal reactions to
traumatic events have hardly been examined. Therefore, some authors even spoke
of a ‘tyranny’ of PTSD (e.g. Kleber, 1995). This risk may also occur after the
introduction of the concept of complicated grief; it may be beneficial for people
suffering from disturbances, but not for the majority of people confronted with
loss.
There is the risk that normal reactions will be considered as mental problems or
disorders that need professional help. Some have raised concerns about the possi-
bility that normal grief would be medicalized as a consequence of a DSM disorder
CG (Stroebe et al., 2000). However, this argument holds not only or in particular
for normal grief, but for almost any mental problem or disorder. ‘Normal’ social
anxiety runs the risk of becoming labelled social phobia, ‘normal rumination’ runs
the risk of becoming labelled generalized anxiety disorder (GAD), problems with
eating may change into ‘eating disorders’.
A related concern is the possibility that the family will withdraw if a treatment
for CG is being carried out. De Keijser (1997) reported that the social network
appears to withdraw when a family member makes use of grief counselling; this
finding is in need of further exploration. Bereaved seem not to envisage such a
reaction: Prigerson and Vanderwerker (2005) report that none of their CG respond-
ents expected that, if they met criteria for CG, their family members would be less
understanding of their distress.

The stressor criterion

Earlier we mentioned that research suggests that full-blown PTSD might develop
in the absence of the occurrence of any ‘traumatic’ event. As a corollary, it can be
126  Jan van den Bout and Rolf J. Kleber
expected that the symptom pattern of CG (such as yearning about what has been
lost) will be found in absence of the loss of a loved one. This is to be expected
because grief is about the dissolution of bonds, but in the grief literature it is hardly
stressed that people have bonds not only with other people (and animals), but also
with other (non-)material objects, such as one’s work situation or one’s culture.
For example, people who suddenly lose their jobs, after many years of having
made personal investments for their employers, may show grief symptoms, which
occasionally are even more intense and long-standing than the grief symptoms
as a result of, for example, the loss of one’s partner. The same applies for losing
‘objects’ such as one’s mental health. For example, there are reports (Appelo et al.,
1993) suggesting that some of the characteristic symptoms of schizophrenia are
primarily grief symptoms, arising from the realization that one’s life perspective
has been broken and that one has a pervasive, life-long illness. This suggests that
grief (with possible disordered forms) is a much more all-embracing phenomenon
than thought before.

Epilogue
Without any doubt, the growth of the PTSD field has been multifaceted: massive,
impressive, and successful as well as complicated, questionable, and sometimes
unconvincing. In spite of the criticisms, PTSD will most definitely remain in
DSM-5. To be fair, the critical arguments apply also to other well-known disor-
ders such as depression and schizophrenia. In these fields there are also debates
about the weak and strong boundaries of these disorders. It is wise to make use of
all the PTSD-related pros and cons in the development (and lobbying) with regard
to the future diagnosis of CG. Many issues with regard to PTSD hold true for a
future disorder of CG. Perhaps the main lesson is: try – for the sake of persons
with enduring intense grief problems – to include CG in DSM, but be careful and
do not overstretch the significance of the concept.

References
American Psychiatric Association. (1980). Diagnostic and statistical manual of mental
disorders (3rd edn.). Washington DC: APA.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4th edn.). Washington DC: APA.
Appelo, M. T., Slooff, C. J., Woonings, F., Carson, J., & Louwerens, J. (1993). Grief: Its sig-
nificance for rehabilitation in schizophrenia. Clinical Psychology and Psychotherapy,
1, 53–59.
Bisson, J.  I., Brayne, M., Ochberg, F.  M., & Everly, G.  S. Jr. (2007). Early psychoso-
cial intervention following traumatic events. American Journal of Psychiatry, 164,
1016–1019.
Boelen, P. A., van den Hout, M., & van den Bout, J. (2006). A cognitive–behavioral con-
ceptualization of complicated grief. Clinical Psychology: Science and Practice, 13,
109–128.
Bolton, D. (2008). What is mental disorder? Oxford: Oxford University Press.
 Lessons from PTSD for CG  127
Bonanno, G. A. (2004). Loss, trauma, and human resilience: Have we underestimated the
human capacity to thrive after extremely aversive events. American Psychologist, 59,
20–28.
Boschen, M.  J. (2008). The growth of PTSD in anxiety disorder research. Psychiatry
Research, 138, 262–264.
Breslau, N., Kessler, R. C., Chilcoat, H. D., Schultz, L. R., Davis, G. C., & Andreski, P.
(1998). Trauma and posttraumatic stress disorder in the community: The 1996 Detroit
area survey of trauma. Archives of General Psychiatry, 55, 626–632.
Brom, D., & Kleber, R. J. (2009). Resilience as the capacity for processing traumatic expe-
riences. In Brom, D., Path-Horenczyk, R., & Ford, J. D. (Eds.), Treating traumatized
children: Risk, resilience and recovery (pp. 133–149). New York: Routledge.
Brom, D., Kleber, R. J., & van den Bout, J. (1993). Loss and trauma: Unity and diversity.
In Malkinson, R., Rubin, S. S., & Witztum, E. (Eds.), Loss and bereavement in Jewish
society in Israel (pp. 39–50). Jerusalem: Cana Publishing House.
Cooper, R. (2004). What is wrong with the D.S.M.? History of Psychiatry, 15, 5–25.
Davidson, J., Fox, E., & Blank, A. (1996). Post traumatic stress disorder. In Widiger, T.,
Frances, A., Pincus, H., Ross, R., First, M., Davis, W., & Kline, M. (Eds.), DSM-IV
sourcebook, vol. 2 (pp. 577–605). Washington, DC: American Psychiatric Association.
Horowitz, M. J. (2003). Stress response syndromes (4th edn.). San Francisco: Jossey-Bass.
Keane, T. M., Marshall, A. D., & Taft, C. T. (2006). Posttraumatic stress disorder: Etiology,
epidemiology and treatment outcome. Annual Review of Clinical Psychology, 2,
161–197.
de Keijser, J. (1997). Sociale steun en professionele begeleiding bij rouw [Social support
and professional counselling for the bereaved]. Amsterdam: Thesis Publishers.
Kendler, K. S. (2010). Mood disorders. Retrieved March 22, 2011, from DSM-5 website:
http://www.dsm5.org/about/Documents/grief%20exclusion_Kendler.pdf.
Kleber, R.  J. (1995). Epilogue: Towards a broader perspective of traumatic stress. In
Kleber, R. J., Figley, C. R., & Gersons, B. P. R. (Eds.), Beyond trauma: Cultural and
societal dimensions (pp. 299–306). New York: Plenum.
McHugh, P. R., & Treisman, G. (2007). PTSD: A problematic diagnostic category. Journal
of Anxiety Disorders, 21, 211–222.
McNally, R. J. (2004). Conceptual problems with the DSM-IV criteria for posttraumatic
stress disorder. In Rosen, G. M. (Ed.), Posttraumatic stress disorder: Issues and contro-
versies (pp. 1–14). Chichester, UK: John Wiley & Sons.
McNally, R. J. (2009). The cutting edge: Can we fix PTSD in DSM-V? Depression and
Anxiety, 26, 597–600.
Mol, S.  S., Arntz, A., Metsemakers, J.  F., Dinant, G.  J., Vilters-van Montfort, P. A., &
Knottnerus, J. A. (2005). Symptoms of post-traumatic stress disorder after non-trau-
matic events: Evidence from an open population study. British Journal of Psychiatry,
186, 494–499.
Neria, Y., Nandi, A., & Galea, S. (2008). Posttraumatic stress disorder following disasters:
A systematic review. Psychological Medicine, 38, 467–480.
van Praag, H.  M. (2000). Nosologomania: A disorder of psychiatry. World Journal of
Psychiatry, 1, 151–158.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged Grief Disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine, 6(8): e1000121.
Prigerson, H. G., & Vanderwerker, L. C. (2005). Final remarks. Omega, 52, 91–94.
Ramsay, R. W. (1977). Behavioural approaches to bereavement. Behaviour Research and
Therapy, 15, 131–135.
128  Jan van den Bout and Rolf J. Kleber
Raphael, B., Martinek, N., & Wooding, S. (2004). Assessing traumatic bereavement. In
Wilson, J. P., & Keane, T. M. (Eds.), Assessing psychological trauma and PTSD (2nd
edn., pp. 492–510). London: Guilford.
Rosen, G.  M., & Lilienfeld, S. O. (2008). Posttraumatic stress disorder: An empirical
evaluation of core assumptions. Clinical Psychology Review, 28, 837–868.
Shear, M.  K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Duan, N., et al. (2011).
Complicated grief and related bereavement-issues for DSM-5. Depression and Anxiety,
28, 103–117.
Schnurr, P. P. (2010). PTSD 30 years on. Journal of Traumatic Stress, 23, 1–2.
Schut, H.  A.  W., & Stroebe, M.  S. (2010). Effects of social support, counselling, and
therapy before and after the loss: Can we really help bereaved people? Psychologica
Belgica, 50, 89–102.
Spitzer, R. J., First, M. B., & Wakefield, J. C. (2007). Saving PTSD from itself in DSM-V.
Journal of Anxiety Disorders, 21, 233–241.
Stein, D. J., Phillips, K. A., Bolton, D., Fulford, K. W. M., Sadler, J. Z., & Kendler, K. S.
(2010). What is mental/psychiatric disorder? From DSM-IV to DSM-V. Psychological
Medicine, 40, 1759–1765.
Stroebe, M., van Son, M. J. M., Stroebe, W., Kleber, R. J., Schut, H. A. W., & van den Bout,
J. (2000). On the classification and diagnosis of pathological grief. Clinical Psychology
Review, 20, 57–75.
Tucker, G.  J. (1998). Putting DSM-IV in perspective. American Journal of Psychiatry,
155, 159–161.
Wortman, C., & Silver, R. (2001). The myths of coping with loss revisited. In Stroebe,
M., Hansson, R., Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement
research: Consequences, coping, and care (pp. 405–429). Washington, DC: American
Psychological Association.
Young, A. (2004). When traumatic memory was a problem: On the historical antecedents
of PTSD. In Rosen, G. M. (Ed.), Posttraumatic stress disorder: Issues and controver-
sies (pp. 127–146). Chichester, UK: John Wiley & Sons.
10 Complicated grief in the context
of other psychiatric disorders
PTSD
Beverley Raphael, Jennifer Jacobs, and
Jeff Looi

Introduction
The consideration of the possibility of “pathologies” of grief has a substantial
history in literature, reflected in prolonged and debilitating mourning overwhelm-
ing the bereaved’s life: they live on, as it were, in their continuing relationship
with the deceased to the exclusion of other life. Freud (1917), in “Mourning and
Melancholia,” discussed the potential for grief to lead to depression. “Traumatic”
grief has also been considered as a possible entity (Raphael, 1983). Recent
research has highlighted the distinction between complicated grief, depression,
and posttraumatic stress disorder (PTSD) as separate syndromes (Boelen, van de
Schoot, van den Hout, de Keijser, & van den Bout, 2010; Golden & Dalgleish,
2010).
The comorbidity of complicated grief and PTSD, the phenomena of reactions
to the stressors of loss and trauma, the possible etiology identified in scientific
studies to date, and implications for management will be the focus of this chapter.
The potential for trauma syndromes such as PTSD to sit alongside, interact
with, or contribute to complicated grief is a challenge for multiple reasons. First,
early in the development of the concept “Traumatic Stress,” theory identified
bereavement or loss of a loved one as a stressor. It has subsequently been difficult
to disengage loss and define it as a distinct stressor and “bereavement” as a dis-
tinct entity. It has been and still is, to a major degree, included by many workers in
the field as a “traumatic” stressor. Second, “Complicated Grief” was also at times
referred to as “traumatic grief,” as in the original development of this concept and
its evolution into “Prolonged Grief Disorder.” The progressive development of
the understanding of “traumatic grief” as the coexistence of grief and trauma phe-
nomena has come with studies which have better clarified the different reactive
phenomena that may follow these different stressor experiences. The attention to
violent deaths, particularly those associated with human malevolent intent such
as homicide or terrorism, has further contributed to recognition of the trauma of
the loss as well as the grief (Neria & Litz, 2003; Rynearson & McCreery, 1993).
Traumatic bereavement could lead to a mixture of trauma and grief, or indeed
complicated grief disorder and comorbid trauma symptoms to the level of PTSD.
Thus PTSD could result from the way a loved one died, or other co-occurring
traumatic stressors.
130  Beverley Raphael et al.

Phenomena of traumatic stress reactions and bereavement

reactions in children and adults
Important early studies of children by Pynoos and colleagues (Pynoos, Frederick,
et al., 1987; Pynoos, Nada, Frederick, Ginda, & Stuber, 1987) examined reactions
in a group of school children who experienced a sniper attack, in which several
children were killed. The researchers developed a Traumatic Stress Reaction
Index (16 items) and a Grief Reaction Index (nine items). Severity of exposure
to life threat was correlated with high levels of traumatic stress reactive phenom-
ena and symptoms. The children more closely associated with a child who died
showed higher levels of grief reactive phenomena on the Grief Reaction Index.
Furthermore, the traumatic stress reaction phenomena were more likely to be
associated with the development of PTSD, whereas loss/bereavement phenomena
were more likely to be associated with depression. This work has extended to
current studies with children so affected, and intervention strategies to address
this pathology have been developed. For instance, Pynoos, Steinberg, and Brymer
(2007) have developed and tested a model of postdisaster intervention that identi-
fies “traumatic bereavement” and recommends strategies, school-based or other,
to address this. Researchers and clinicians have increasingly recognized the coex-
istence of traumatic stress and grief phenomena for children experiencing certain
losses, and developed models to assess and manage this (Cohen & Mannarino,
2008).
Studies examining bereavement phenomena over time have assessed patterns
of reaction for adult populations of bereaved community members following the
death of a loved one (Byrne & Raphael, 1993; Middleton, Moylan, Raphael, &
Martinek, 1998). These are non-clinical samples and so represent the spectrum of
potential reactions. The reactive phenomena demonstrate patterns of progressive
decline in levels in the early months after a major bereavement. About 9% were
found to continue in a “chronic” pattern for 6 months or more, with high levels of
acute grief phenomena reported in the populations studied, representing patterns
like those of prolonged grief disorder or complicated grief (Raphael & Minkov,
1999).
The decline in levels of reactive phenomena or “symptoms” after exposure
to the stressor of loss of a loved one is similar to the patterns of traumatic stress
reactive phenomena or symptoms over time. If symptom patterns reach the level
of dysfunction or disorder, they have the potential to meet criteria for complicated
grief and PTSD after a specific period.
Both these sets of phenomena relate directly and specifically to the stressors
experienced. Studies have included clear, systematic, and validated measures of
the specific phenomena; for instance, for bereavement, the Core Bereavement
Items scale (CBI) (Burnett, Middleton, Raphael, & Martinek, 1997) and trau-
matic stress measures such as the reliable and valid items of PTSD checklists. The
specific stressors, as opposed to reactive phenomena, are, however, infrequently
measured systematically and consistently, thus complicating comparison across
different studies in these fields.
 CG and other psychiatric disorders  131

Clinical observations of phenomena

Characteristic reactions of grief and posttraumatic stress (PTS) reactions are
important in demonstrating specificity of reactive phenomena as well as similari-
ties. These can be considered in terms of:

• cognitive phenomena;
• affective phenomena;
• avoidance phenomena;
• arousal phenomena;
• other related phenomena.

These have been reported previously (Raphael, Martinek, & Wooding, 2004).
They have been considered to be useful themes in clinical assessment, and could
also be considered as potential domains in research studies. These phenomena are
presented in Tables 10.1 and 10.2.

Bereavement and pathology: complicated grief, PTSD,

depression
The potential for the loss of a loved one to lead to complicated or prolonged
grief is widely discussed in this volume, as are criteria for such a syndrome or
disorder. Conceptually this syndrome is strongly related to attachment theory, and
suggests an ongoing, complicated attachment of the bereaved to the person who
is now deceased. The stressor is the disruption of the attachment, its loss. There
is considerable research examining the relationship between loss and depression,
and other complications after bereavements. It is only recently that studies have
specifically included the potential for the distinct entity PTSD as a comorbidity.
It is of note that many of the more recent studies considering bereavement,
complicated grief, and PTSD have been of populations affected by disaster or
terrorism. In some of these studies, recording the potential for loss syndromes has
been considered chiefly because there are concurrently multiple deaths and losses.
The stressors of life threat, as indicated in criterion A of DSM-IV, have also been
investigated, as has their contribution to the subsequent development of PTSD.
High mortality rates have been found to be associated with higher levels of psy-
chopathology, for instance, explaining 20% of the variance in disaster effect size
in a well-controlled meta-analysis (Norris & Wind, 2009). This probably relates
to the fact that bereavement is an additional stressor, as well as to the fact that
many of those surviving are likely to have been exposed to severe life-threatening
circumstances themselves, with associated heightened risk of PTSD. In addition
they may have been exposed to the multiple, possibly mutilating, deaths of others.
Studies after September 11, 2001 (9/11), such as those of Galea et al. (2002),
reported that, at 5–8 weeks, having a friend or relative who had been killed, that is,
a bereavement stressor, was associated with depression, not PTSD. Complicated
grief was not measured. However, a subsequent study of primary care patients
Table 10.1  Phenomena of posttraumatic reactions and bereavement

Posttraumatic phenomena Bereavement phenomena

Cognitive phenomena
Intrusions of scene of trauma (e.g., death) Image of lost person constantly comes to mind
not associated with yearning or longing (unbidden or bidden)
Associated with distress, anxiety at image Associated with yearning or longing
Preoccupation with the traumatic event Distress that person is not there
and circumstances of it Preoccupation with the lost person and intense
Memories usually of the traumatic scene images of him or her
Re-experiencing of threatening aspects of Memories of person associated with affect relevant to
the event memory (often positive)
Re-experiencing of person’s presence, as though he
or she were still there (e.g., hallucinations of sound,
touch, sight)

Affective phenomena

Anxiety
Anxiety is the principal affect Anxiety, when present, is separation anxiety
And is general and generated by threat Is specific and generated by separation from lost
Fearful of threat/danger person
Precipitated by reminders, intrusions Is generated by imagined future without lost person
Precipitated by his or her failure to return

Yearning/longing
These are not prominent features Yearning for lost person is a core grief phenomenon
Not person oriented; if occurs, is for things Is person oriented, intense, painful, profound,
to have been as they were before – for the triggered by reminders of him or her; yearning for
return of “innocence of death” and for the him or her to return, to be there
sense of personal invulnerability

Sadness
Sadness not commonly described Sadness frequent and profound
Nostalgia for event not described Feelings of nostalgia common and persistent

Avoidance phenomena
Avoids reminders of events, including May search for and seek out places of familiarity,
places treasured objects (e.g., linking objects, photos and
Attempts to lessen affect; numbing, images)
lessened feelings generally May try to avoid reminders of the absence of the lost
May have great difficulty talking of event person; may try to mitigate pangs of grief but only
during avoidance times, although at others temporarily, including distracting, but also seeks to
may be powerfully driven to talk of the express grief as normal
experience (but not person) May be very driven to talk of lost relationship and
Withdrawal from others (protective of self) lost person
May seek others for support or to talk of deceased

Arousal phenomena
Oriented to threat and danger Oriented to lost person
General scanning and alertness to danger, General scanning of environment for lost one or cues
fearfulness of him or her
Exaggerated startle response (i.e., response Arousal drives searching behavior
to minimal threat) Overresponse to cues of lost person
Overresponse to cues of trauma

Source: Adapted from Raphael et al. (2004).

 CG and other psychiatric disorders  133
Table 10.2 Other phenomena: signs of reactive process

Posttraumatic phenomena Bereavement phenomena

Occur on witnessing something horrific, torture, “Contraction of the grief muscles . . . Appears
etc., fear and threat to be common to all the races of mankind”
“probably that horror would generally be (p. 185)
accompanied by strong contraction of the Obliquity of the eyebrows; contraction of
brow, but as far as fear is one of the elements, central fascia of frontal muscle
the eyes and mouth would be opened, and the Inner ends of eyebrows (p. 188) puckered into
eyebrows raised – as far as antagonistic action bunch
of the corrugations permitted this movement” Transverse furrows across the middle parts of
(pp. 322–323) the forehead
“Contraction of platysma does add greatly to the Depression of corners of mouth
expression of fear” (p. 317) Mouth closed
Eyes somewhat staring Corners drawn downward and outward
Pupils may be dilated (pp. 201–202)
Curved mouth concavely downward

Note: Page numbers are from Darwin (1872/1998).

who knew someone who died in the attack found that they were twice as likely to
report problems in functioning, work, and social and family life, and at least one
mental disorder (Neria et al., 2008). Although depression was the most prevalent
condition, the stressor experience reported was most strongly connected with
PTSD. In another study, Neria et al. (2007) used a web-based survey to study the
long-term grief reactions of 704 adults bereaved after 9/11 at 2.5–3.5 years after
the attack. They specifically assessed complicated grief as a distinct syndrome,
and found that it was often comorbid with depression and PTSD. Whereas most
of the participants reported some complicated grief symptoms, 43% met study
criteria for complicated grief as a diagnosis.
A comprehensive study of bereaved Norwegians who had lost loved ones in
the 2004 South East Asian tsunami was carried out, assessing those who had been
directly exposed to the disaster and those not so exposed (Kristensen, Weisaeth,
& Heir, 2009). The authors used diagnostic criteria interviews to diagnose PTSD
and major depressive disorder and a self-report scale to measure prolonged grief
disorder (PGD). Rates of psychiatric disorders were twice as high in those directly
exposed. They reported that loss of a child and low education correlated with
PGD whereas exposure correlated with PTSD. Each disorder was independently
correlated with functional impairment.

Stressor exposures, PTSD, and bereavement

Multiple studies have highlighted the complexity of etiological processes that
might contribute to comorbidity of PTSD and complicated grief. Questions
arise whether this is related to traumatic circumstances of the death, life threat
exposures for the bereaved, or other concurrent stressors. This is not clearly
identifiable in the methodologies and findings of many studies – particularly the
degree to which “circumstances of the death” is the identified stressor, vis-à-vis
the multiple other potential life threat circumstances that may occur in association
134  Beverley Raphael et al.
with loss. There is also the possibility of PTSD being the initial condition, with
associated effects on attachment relationships being a factor contributing to the
development of complicated grief. Other vulnerabilities, such as past trauma and
loss, children’s developmental trauma, or multiple concurrent adversities could
also increase the risk of complicated grief.
The potential ways people have died may be “traumatic,” as described in the
stressor criterion A of PTSD, and be associated with greater risk of PTSD for
those so bereaved. This could be the case if, for example (as per criterion A of
PTSD), the person witnessed, or was confronted with, an event or events that
involved actual or threatened death or serious injury or threat to the physical
integrity of the self or others, and the person’s response involved intense fear
responses, helplessness, and horror.
Anderson, Arnold, Angus, and Bryce (2008) have studied complicated grief
and PTS in family members of patients in an intensive care clinic. Family mem-
bers were enrolled when their loved one was still alive in intensive care, and
assessed at three time points: enrolment, 1 month, and 6 months. Interviews and
assessment covered anxiety, depression, PTS, and complicated grief. This study
found that complicated grief was not associated with anxiety or depression at any
time point. Forty-six percent of the bereaved participants had complicated grief.
Thirty-five percent of participants had PTS symptoms according to the cut-off
point on the Impact of Event Scale (IES), which indicates probable PTSD, and
this correlated with pre-existing and concurrent anxiety scores. Of significance,
however, was the finding that all the bereaved participants with high levels of PTS
(IES score > 30) also had complicated grief. The authors expressed concern at
the high prevalence of “post-traumatic stress” in this cohort of family members,
because of the “profound impact of such disorders on physical and mental health
and social functioning” (p. 1874).
In a study of adults who had lost first-degree relatives to war-related violence
(Kosovo), Morina, Rudari, Bleichhardt, and Prigerson (2009) examined the rates
of PGD or complicated grief disorder and its association with PTSD and major
depressive disorder (MDD). This preliminary investigation is of interest in that
the particular stressors were assessed as well as symptoms/syndrome patterns,
for instance if there was an immediate threat to one’s own life, and other ongoing
“strains.” The loss of a loved one was associated with PGD, which affected 38%
of this sample of 60 bereaved people; PTSD was associated with immediate threat
to one’s own life and affected 55% of this population. Major depressive disorder
was predicted by ongoing strains and occurred in 38%. Prolonged grief disorder
was significantly associated with MDD, anxiety symptoms, sleeping difficul-
ties, and feelings of embitterment. Whereas two thirds (65%) of the participants
with PGD also met criteria for PTSD, 49% of those with PTSD did not meet
criteria for PGD. There was no significant relationship between PTSD and PGD,
nor between PGD or PTSD and MDD. These findings would suggest that, even
though these bereavements were likely to have been very traumatic in terms of
the circumstances of the way people died, the PTSD in this study was specifically
related to the extent of personal life threat. The PTSD and PGD were two distinct
 CG and other psychiatric disorders  135
entities, highlighting the etiological and phenomenological distinctions, despite
the prevalence and overlap of comorbidities. Of particular concern, as noted by
these authors, was that “even seven years after the war, 73.3% of the participants
fulfilled criteria for PGD, PTSD or MDD” (p. 7). They also note that a major area
of psychopathology, that related to complicated or prolonged grief disorder, would
not have been addressed if bereavement-related pathology had not been included.
Benedek and Ursano (2006) describe related issues with military communities.
The nature of bereavement through death by homicide has been studied in
detail and reported in a review drawing together many of these themes (Rynearson,
2006). Rynearson’s early work in this field highlighted the traumatic nature of
such deaths, even for those not present when the death occurred. They could be
tormented by images and flashbacks of such potential circumstances of death,
as well as the complexity and intensity of their grief (Rynearson & McCreery
1993; see also Chapter 20 in this volume). In a recent national study of mental
health correlates of losing a loved one to homicide, Zinzow, Rheingold, Hawkins,
Saunders, and Kilpatrick (2009) found heightened risks for PTSD (odds ratio
1.88), major depressive episode (odds ratio 1.64), and drug use dependence (odds
ratio 1.77). Unfortunately, complicated grief was not considered in this study.
Suicide deaths have been studied in terms of those who survive such family
losses. Brent, Melhelm, Donohoe, and Walker (2009) studied the incidence and
course of depression in bereaved youths 21 months after the death of a parent by
suicide, accident, or natural death, and also examined sequelae of such deaths for
caregivers (surviving parent) (Melhelm, Walker, Moritz, & Brent, 2008). They
reported an increased risk for both depression and PTSD above and beyond other
vulnerability factors. In terms of the “sudden” deaths studied, being bereaved by
suicide was not associated with heightened risk compared with other acute deaths.
This study is of interest in that a specific measure of “Circumstances of Exposure
to Death” had been developed and tested in a prior related study (Brent et al.,
1992). Complicated grief was assessed using the Inventory of Complicated Grief
(Prigerson et al., 1995). Both risk and protective factors that might influence the
impact of bereavement were also assessed, such as family cohesion, social sup-
port, self-esteem, and coping styles, plus pre-existing psychopathology as reported
for the deceased, the surviving parent, and offspring. This comprehensive study
thus addressed a significant variable in the circumstances of the death, and it also
examined specific types of sudden death that, in and of themselves, may have
contributed to increased risk of adverse bereavement outcomes. Complicated
grief was found in 32% of the suicide caregivers, 25% of accident caregivers, and
31% of sudden natural death caregivers, as a new-onset psychiatric condition. In
terms of the offspring complicated grief was reported in 21% of suicide bereaved,
18% of accident bereaved, and 19% of those bereaved by natural sudden death.
This study found that caregiver and offspring symptoms of depression, anxiety,
PTSD, suicidal ideation, and complicated grief were correlated. The study exam-
ined correlations of new-onset of PTSD in the bereaved offspring, finding that
prior psychiatric history (odds ratio 9.4), bipolar disorder in the deceased parent
(odds ratio 1.9), and family history of PTSD (odds ratio 6.8) were predictors. It
136  Beverley Raphael et al.
is of particular significance that type of death per se was not a significant predic-
tor of PTSD, but that “being at the scene when the death occurred” (odds ratio
8.3) (Melhelm et al., 2008, p.  407) was. As will be discussed below, this may
be a better indicator of the potential impact of “circumstances of the death” as a
possible traumatic stressor exposure. Brent et al. (2009) interviewed the initial
cohort described above approximately 9 months after the loss. This second study
reported on subsequent data with interviews about a year later, that is, 21 months
after the death. Compared with the non-bereaved control group, the bereaved
young people had higher rates of major depression or alcohol or substance use
problems. Post hoc correlates indicated that the level of complicated grief was
higher in young people whose parents had died by accident. The centrality of
depression as a pathological outcome, with onset associated with bereavement,
was further demonstrated in the cumulative incidence and course of depression
in bereaved compared with non-bereaved control youths. Cumulative incidence
and course of PTSD in those bereaved was highest in the first 9 months, about
8.5%, but not in the subsequent period. Remission was common. The authors
conclude that the “direct effect of bereavement on incident depression and PTSD
was limited to the first 9 months after the loss” (p. 791).
Shocking, horrific ways of dying may be the consequence of different life
threats, including homicide, mass disaster, terrorism, or war, where criterion A of
PTSD may be met. All may lead to suffering, profound agony, helplessness, grief,
and the specific suffering associated with the deaths of children. September 11,
2001, exemplifies this, the effects of which were increased by extensive media
depiction and amplification, that is, witnessing, often repeatedly, horrific deaths.
It is likely that such circumstances would encompass both trauma and loss. This
and other terrorist or similar shocking and deliberate perpetrations of mass death
and destruction could lead to the potential for traumatic exposures as well as loss
and grief. Shear, Jackson, Essock, Donahue, and Felton (2006) screened people
presenting to Project Liberty community counseling outreach services after 9/11.
These researchers found that, at 1.5 years after the event, 44% of the counseling
recipients who knew someone who had been killed in the attacks screened posi-
tive for complicated grief. Positive screening, which was highest amongst those
who had lost a family member, was associated with functional impairment and
subthreshold or diagnostic levels of major depression and PTSD. These authors
highlighted the complex interfaces of trauma and grief pathologies, in that the
symptoms of complicated grief with separation anxiety may interact with threat
anxiety aspects of PTSD. Higher PTS symptom scores and higher grief scores
were associated with higher levels of functional impairment, and neither grief
alone nor PTSD alone correlated with such impairment.

Assessment and management: complicated grief with trauma

syndromes as comorbidity
The issues discussed above constitute a complex and sensitive challenge to man-
agement. This is the more so because of overlapping symptom patterns, as with
many such comorbidities, the extra burden that may result for the person who is
 CG and other psychiatric disorders  137
affected, and the skills required for treatment. Shear et al. (2001) conducted a pilot
study and subsequently a randomized controlled trial of a manualized but sensi-
tive treatment for complicated grief (Shear, Frank, Houck, & Reynolds, 2005).
They have contributed a very important set of strategies for this field.
The potential adaptations of this model to treat comorbid trauma syndromes
are well exemplified in the development of a protocol for the treatment of compli-
cated grief and substance use disorders. This study (Zuckoff et al., 2006) included
the set of the complicated grief strategies, but also had the goal of helping patients
“to achieve sufficient initial improvement in substance use behaviours” (p. 207).
They would then learn skills for managing intense emotions safely so that these
would not, if evoked in the grief therapy, lead to relapse of substance use behav-
iors related to this emotional response. The model developed was called CGSUT
(Complicated Grief Substance Use Treatment). It combined components for
achieving abstinence from substances; enhanced tolerance for emotional reac-
tiveness (including motivational interviewing and emotion coping skills); and
the established treatment for complicated grief. Two of the strategies, those for
enhancing recognition and management of distressing emotions and skills for
communicating feelings and empathy for others, are likely to be of use across
other comorbidities such as PTSD. This small open trial, the authors suggest,
“is promising” and “a grief focussed treatment, combined with MI [motivational
interviewing] and skills building for emotion coping and communication” consti-
tutes a feasible model for further study (Shear et al., 2007, p. 209). A challenge
lies in the fact that comorbidity is frequently entrenched, particularly years after
the loss, so that studies of early intervention (e.g., Fiegelman, Jordan, & Gorman,
2009) should be considered of particular importance. Longer therapies (e.g., 24
sessions) also challenge commitment.
Hensley (2006) took up the earlier conceptualization of complicated grief as
traumatic grief, and considered comorbid depression, again demonstrating the
difficulties of terminology in this field. From this review she suggested that treat-
ment may require “specific psychotherapeutic techniques addressing the trauma
and separation distress central to this disorder” (p. 124).
The increasing focus on cognitive–behavioral therapies (CBT) that are trauma
focused, as a treatment for trauma syndromes, has provided a basis for the use of
this type of intervention in the management of such comorbidity. Nevertheless
there is significant overlap with complicated grief treatment elements. This is par-
ticularly likely if the trauma stressor is directly related to the circumstances of the
death, the way the person died, whereby emotional responsiveness, fear reactions,
anger, and distress may be intense. The management of these emotional aspects
is likely to be critical. In the first author’s experience with clinical management
of these issues, the discussion of the death itself, what actually happened, can be
facilitated with empathy, so that the exposure component is progressively man-
aged and is a less dominating element of the pathology, that is, “the story of the
death” is told.
Shear et al. (2007) further examined the attachment aspect of complicated grief
and highlighted the significance of the avoidance theme in management in this
context. The reviewing of the bonds of the bereaved person to the lost person,
138  Beverley Raphael et al.
and the capacity to review memories and deal with the dependent and ambivalent
aspects of the relationship with the deceased, may be part of therapeutic exposure
or confrontation processes. Having a more real set of memories of the relationship
may help the bereaved. Attachment issues and the ways in which they contribute
to complicated grief need clinical and research clarification. Furthermore, the
significance of the attachment theme needs to be more specifically researched in
therapy models and clinical trials.
The role of pharmacotherapy in the management of complicated grief and
PTSD comorbidity remains to be established. Simon et al. (2008) describe the
association of “naturalistic” pharmacotherapy use in the treatment trial of Shear
et al. (2005). They found that those having such antidepressant pharmacotherapy
were more likely to complete the full course of therapy sessions for complicated
grief treatment than those undergoing the control intervention psychotherapy
treatment. Pharmacotherapy could be indicated for comorbid PTSD. The pharma-
cological rationale behind any such interventions should also be clarified.
Stroebe, Schut, and Stroebe (2005) looked at grief counseling more broadly,
showing in their review that there was little evidence that the emotional disclosure
aspect of counseling facilitated adjustment to loss in normal bereavement, and the
same issues are also relevant for trauma syndromes. This further emphasizes the
importance of research addressing vulnerabilities and possible protective factors
with the aim of lessening the risk of complicated grief disorder, related comor-
bidities such as trauma syndromes, substance use disorders, and depression, and
other adverse health or behavioral outcomes. Prevention and early intervention
strategies have been suggested (Zhang, El-Jawahri, & Prigerson, 2006) and some
earlier studies (randomized controlled trials) demonstrated that interventions
with high-risk bereaved persons could potentially lead to improved outcomes
(Raphael, 1977).
The needs of children and adolescents with complicated grief and trauma syn-
dromes have been increasingly recognized, particularly in disaster-affected popu-
lations, or among children affected in other ways by grief and trauma. Cohen,
Mannarino, and Starlon (2006) have described this work in their studies of child
traumatic grief, a mixture of unresolved grief symptoms and trauma symptoms.
They used modularized approaches to deal with grief problem symptoms, and
trauma-focused CBT for trauma symptoms. They also provided program compo-
nents for parents or caregivers to assist them to deal with their children’s needs
and possibly their own difficulties.
The model of complicated grief therapy used by Shear for adults, the modifica-
tions developed for complicated grief and substance use, and the modularized
formats for children have all been shaped by research and applied as clinically
relevant to the assessed phenomena of complicated grief and any associated
comorbidities. The therapeutic descriptions suggest both adaptation of frame-
works and responsiveness to particular patterns in these complex sets of comor-
bid phenomena and grief problems. The clinical programs described are forming
and developing the evidence base. Such evolution is of value, as are the cur-
rently available trials and conceptualizations. A more comprehensive research
 CG and other psychiatric disorders  139
framework could also build on emerging literature on Internet-based therapies
(Stroebe et al., 2005; Wagner & Maercker, 2008) and the potential role of pharma-
cotherapies. The various models with children and their families will also require
further research development (Cohen & Mannarino, 2008). The progressive evo-
lution, translation, and implementation of such scientifically developed programs
is critically important to meet real-world needs for those affected by complicated
grief, PTSD, and other comorbidities. This is the more so when the difficulties of
retaining people in some of the trials highlights the barriers to care delivery and
the potential benefits that evaluation of the translated program can add to research
and knowledge development. This is clear with the model of Pynoos et al. (2007),
which has translated screening and interventions into program models for trauma
and grief, taking into account developmental issues. As in all psychotherapies,
the context, the significant others, and being attuned to the person’s experience
of grief and trauma should inform therapeutic management. Genuineness, empa-
thy, and warmth are still central, particularly for those who have experienced the
tragic loss of a primary attachment figure in traumatic circumstances. The role of
consoling, human touch, and healing rituals also need to be further researched.
The processes of grieving are core human adaptations, and should be valued,
respected, and understood. They are shaped in their expression by enduring
cultural processes. Their core biological underpinnings are being progressively
explored. The problems, “diseases” of grief, are also being described. However,
the underlying power and “magic” of human bonds in times of loss and grief
should be seen for what they are: the core of family, the fabric of society.

References
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental
disorders (4th edn., text revision). Washington, DC: Author.
Anderson, W., Arnold, R., Angus, D., & Bryce, C. (2008). Posttraumatic stress and compli-
cated grief in family members of patients in the intensive care unit. Journal of General
Internal Medicine, 23, 1871–1876.
Benedek, D. & Ursano, R. (2006). Mass violent death and military communities: Domains
of response in military operations, disaster and terrorism. In Rynearson, E. (Ed.),
Violent death: Resilience and intervention beyond the crisis (pp. 295–310). New York:
Routledge.
Boelen, P., van de Schoot, R., van den Hout, M., de Keijser, J., & van den Bout, J. (2010).
Prolonged grief disorder, depression and posttraumatic stress disorder as distinguish-
able syndromes. Journal of Affective Disorders, 125, 374–378.
Brent, D., Melhelm, N., Donohoe, M., & Walker, M. (2009). The incidence and course of
depression in bereaved youth 21 months after the loss of a parent to suicide, accident or
a sudden natural death. American Journal of Psychiatry, 166, 786–794.
Brent, D., Perdper, J., Moritz, G., Allman, C., Friend, A., Schweers, J., Roth, C., Balach, L.,
& Harrington, K. (1992). Psychiatric effects of exposure to suicide among the friends
and acquaintances of adolescent suicide victims. Journal of the American Academy of
Child and Adolescent Psychiatry, 31, 629–639.
Burnett, P., Middleton, W., Raphael, B., & Martinek, N. (1997). Measuring core bereave-
ment phenomena. Psychological Medicine, 24, 411–421.
140  Beverley Raphael et al.
Byrne, G., & Raphael, B. (1993). A longitudinal study of bereavement phenomena in
recently widowed elderly men. Psychological Medicine, 27, 49–57.
Cohen, J., & Mannarino, A. (2008). Disseminating and implementing trauma-focused CBT
in community settings. Trauma, Violence, Abuse, 9, 214–226.
Cohen, J., Mannarino, A., & Starlon, V. (2006). A pilot study of modified cognitive behav-
ioural therapy for childhood traumatic grief (CBT-CTG). Journal of the American
Academy of Child and Adolescent Psychiatry, 45, 1465–1473.
Darwin, C. (1872/1998). The expression of emotions in men and animals (3rd edn.).
London: HarperCollins.
Fiegelman, W., Jordan, J., & Gorman, B. (2009). How they died, time since loss, and
bereavement outcomes. Omega, 58, 251–273.
Freud, S. (1917). Mourning and melancholia. Standard Edition, 14, 243–258.
Galea, S., Ahern, J., Resnick, H., Kilpatrick, D., Bucuvalas, M., Gold, J., & Vlahov, D.
(2002). Psychological sequelae of the September 11 terrorist attacks in New York City.
New England Journal of Medicine, 346, 982–987.
Golden, A., & Dalgleish, T. (2010). Is prolonged grief distinct from bereavement-related
posttraumatic stress? Psychiatry Research, 178, 336–341.
Hensley, P. (2006). Treatment of bereavement-related depression and traumatic grief.
Journal of Affective Disorders, 92, 117–124.
Kristensen, P., Weisaeth, L., & Heir, T. (2009). Psychiatric disorders among disaster
bereaved: An interview study of individuals directly or not directly exposed to the 2004
tsunami. Depression and Anxiety, 26, 1127–1133.
Melhelm, N., Walker, M., Moritz, G., & Brent, D. (2008). Antecedents and sequelae of
sudden parental death in offspring and surviving caregivers. Archives of Paediatric and
Adolescent Medicine, 162, 403–410.
Middleton, W., Moylan, A., Raphael, B., & Martinek, N. (1998). A longitudinal study
comparing bereavement phenomena in recently bereaved spouses, adults, children and
parents. Australian and New Zealand Journal of Psychiatry, 32, 235–241.
Morina, N., Rudari, V., Bleichhardt, G., & Prigerson, H. (2009). Prolonged grief disorder,
depression, and posttraumatic stress disorder among bereaved Kosovar civilian war
survivors: A preliminary investigation. International Journal of Social Psychiatry, 56,
288–297.
Neria, Y., Gross, R., Litz, B., Maguen, S., Insel, B., Seirmarco, G., et al. (2007). Prevalence
and psychological correlates of complicated grief among bereaved adults 2.5–3.5 years
after September 11th attacks. Journal of Traumatic Stress, 20, 251–262.
Neria, Y., & Litz, B. (2003). Bereavement by traumatic means: The complex synergy of
trauma and grief. Journal of Loss and Trauma, 9, 73–87.
Neria, Y., Olfson, M., Gameroff, M., Wickramaratne, P., Gross, R., Pilowsjy, D., et al.
(2008). The mental health consequences of disaster-related loss: Findings from primary
care one year after the 9/11 terrorist attacks. Psychiatry, 71, 339–348.
Norris, F., & Wind, L. (2009). The experience of disaster: Trauma, loss, adversities, and
community effects. In Neria, Y., Galea, S., & Norris, F. (Eds.) Mental health and disas-
ters (pp. 29–44). London: Cambridge University Press.
Prigerson, H., Maciejewski, P., Newson, J., Reynolds, C., Bierhals, A., Miller, M., &
Doman, J. (1995). Inventory of Complicated Grief: A scale to measure maladaptive
symptoms of loss. Psychiatry Research, 59, 65–79.
Pynoos, R., Frederick, C., Nada, K., Arroyo, W., Steinberg, A., Eth, S., et al. (1987). Life
threat and posttraumatic stress in school-aged children. Archives of General Psychiatry,
44, 1057–1063.
 CG and other psychiatric disorders  141
Pynoos, R., Nada, K., Frederick, C., Ginda, L., & Stuber, M. (1987). Grief reactions in
school-aged children following a sniper attack at school. Israeli Journal of Psychology
and Related Sciences, 24, 53–63.
Pynoos, R., Steinberg, A., & Brymer, M. (2007). Children and disasters: Public mental
health approaches. In Ursano, R. J., Fullerton, C. S., & Weisaeth, L. (Eds.) Textbook of
disaster psychiatry (pp. 48–68). Cambridge: Cambridge University Press.
Raphael, B. (1977). Preventive intervention with the recently bereaved. Archives of
General Psychiatry, 34, 1450–1459.
Raphael, B. (1983). The anatomy of bereavement. New York: Basic Books.
Raphael, B., Martinek, N., & Wooding, S. (2004). Assessing traumatic bereavement. In
Wilson, J., & Keane, T. (Eds.), Assessing psychological trauma and PTSD (2nd edn.,
pp. 492–510). London: Guilford Press.
Raphael, B., & Minkov, C. (1999). Abnormal grief. Current Opinion in Psychiatry, 12,
99–102.
Rynearson, E. (2006). Violent death: Resilience and intervention beyond the crisis. New
York: Routledge.
Rynearson, E., & McCreery, J. (1993). Bereavement after homicide: A synergism of trauma
and loss. American Journal of Psychiatry, 150, 258–261.
Shear, K., Frank, E., Foa, E., Cherry, C., Reynolds, C., Vander-Bilt, J., & Masters, S.
(2001). Traumatic grief treatment: A pilot study. American Journal of Psychiatry, 158,
1506–1508.
Shear, K., Frank, E., Houck, P., & Reynolds, C. (2005). Treatment of complicated grief:
A randomised controlled trial. Journal of the American Medical Association, 293,
2601–2608
Shear, K., Jackson, C., Essock, S., Donahue, S., & Felton, C. (2006). Screening for com-
plicated grief among project liberty service recipients 18 months after September 11,
2001. Psychiatric Services, 57, 1291–1297.
Shear, K., Monk, T., Houck, P., Melhem, N., Frank, E., Reynolds, C., & Sillowash, R.
(2007). An attachment-based model of complicated grief including the role of avoid-
ance. European Archives of Psychiatry and Clinical Neuroscience, 257, 453–461.
Simon, N., Shear, K., Fagiolini, A., Frank, E., Zalta, A., Thompson, E., et al. (2008).
Impact of concurrent naturalistic pharmacotherapy on psychotherapy of complicated
grief. Psychiatry Research, 159, 31–36.
Stroebe, W., Schut, H., & Stroebe, M. (2005). Grief work, disclosure and counselling: Do
they help the bereaved? Clinical Psychology Review, 235, 395–414.
Wagner, B., & Maercker, A. (2008). An internet-based cognitive–behavioural interven-
tion for complicated grief: A pilot study. Giornale Italiano di Medicina del Lavoro ed
Ergonomica, 30, B47–B53.
Zhang, B., El-Jawahri, A., & Prigerson, H. (2006). Update on bereavement research:
Evidence-based guidelines for the diagnosis and treatment of complicated grief.
Palliative Care Reviews, 9, 1188–1203.
Zinzow, H., Rheingold, A., Hawkins, A., Saunders, B., & Kilpatrick, G. (2009). Losing a
loved one to homicide: Prevalence and mental health correlates in a national sample of
young adults. Journal of Traumatic Stress, 22, 20–27.
Zuckoff, A., Shear, K., Frank, E., Dales, D., Seligman, K., & Sillowash, R. (2006). Treating
complicated grief and substance use disorder: A pilot study. Journal of Substance Abuse
Treatment, 30, 205–211.
Part IV

Contemporary research on
risk factors, processes, and
mechanisms
11 Prospective risk factors for
complicated grief
A review of the empirical literature
Laurie A. Burke and Robert A. Neimeyer

Grieving is a natural response to the loss of a loved one, one that is repeatedly
experienced by most individuals during their lifetimes. Although grief is ubiqui-
tous, research shows that responses to loss vary among grievers. Some individu-
als respond resiliently, by experiencing little in the way of psychological distress
(Bonanno & Kaltman, 2001), others experience acute grief for as long as 1–2
years (Bonanno & Mancini, 2006), and still others experience severe, debilitating,
and sometimes life-threatening grief for a protracted length of time – a condition
known as complicated grief (CG; Prigerson, Frank, et al., 1995) or prolonged
grief disorder (PGD; Prigerson et al., 2009). Therefore, because of the increased
precision with which we can identify the distinct characteristics of CG (Holland,
Neimeyer, Boelen, & Prigerson, 2009), better scales with which to measure it
(Prigerson, Frank, et al., 1995), and improved therapies with which to treat it
(Shear, Frank, Houch, & Reynolds, 2005), isolating prospective risk factors is cru-
cial. Our primary goal in this chapter is to identify empirically supported factors
that predict subsequent susceptibility to the full range of responses to loss, from
common to complicated grief, that merit further scientific and clinical attention.

CG as a distinct risk
Although depression and other forms of general psychopathology are important
components of bereavement distress (Bonanno & Mancini, 2006), some research-
ers maintain that grief-specific distress can be expressed on a continuum of
responses to loss. On one end is resilience, such that psychological equilibrium
is regained fairly quickly after the loss (Bonanno & Kaltman, 2001). The middle
range reflects a common response in which grievers suffer moderate distress (e.g.,
shock, anguish, sadness), but over time are able to adapt to the loss. The most
serious expression – CG – appears at the far end of the spectrum, and is character-
ized by a state of protracted grieving, reflected in profound separation distress,
emotionally disconcerting and invasive memories of the deceased, emptiness
and meaninglessness, an inability to accept the loss, and considerable difficulty
continuing to live life in the absence of the loved one (Holland et al., 2009). Other
researchers view CG as a distinct entity from normative grief, so that measured
symptom counts that exceed a normed cut-off score on a scale assessing CG (e.g.,
146  Laurie A. Burke and Robert A. Neimeyer
the Inventory of Complicated Grief, ICG; Prigerson, Maciejewski, et al., 1995)
are considered to be categorically different from lower ones and, thus, constitute
the presence of a discrete disorder (i.e., CG or prolonged grief disorder; Prigerson,
Frank, et al., 1995; Prigerson et al., 2009).
CG has demonstrated construct validity as a condition that predicts serious
medical and psychological outcomes, beyond those predicted by depression,
PTSD, or anxiety (Lichtenthal, Cruess, & Prigerson, 2004). Historically, most
studies have investigated samples of older Caucasian widows, with middle to
upper socioeconomic status, bereaved by natural deaths, who have a normative
response to their loss. However, recent studies have assessed CG in a variety of
samples (e.g., multiple races, Goldsmith, Morrison, Vanderwerker, & Prigerson,
2008; parents, Keesee, Currier, & Neimeyer, 2008; survivors of terrorism, Shear,
Jackson, Essock, Donahue, & Felton, 2006; African Americans bereaved by hom-
icide, McDevitt-Murphy, Neimeyer, Burke, & Williams, 2012), in which CG may
be more prevalent than originally thought and predicted by specific risk factors.

The present review

Although earlier reviews have been informative (see Sanders, 1988; M. Stroebe,
Schut, & Stroebe, 2007; W. Stroebe & Schut, 2001), they necessarily failed to
include the burgeoning recent literature on intense and prolonged grief. Lobb and
colleagues (2010) reviewed studies on CG, and grouped risk factors into catego-
ries associated with childhood, issues of dependency, caregiving, cognitive and
behavioral conceptualizations, traumatic death, and serious mental illness. They
found that insecure attachment, excessive dependency, negative interpretations of
grief reactions, a lack of meaning making, a lack of preparedness for the death,
perception that the loved one suffered while dying, low social support, caregiver
burden, and a history of psychopathology exacerbate grief. Though useful, Lobb
et al.’s (2010) review and others combined genuinely prospective predictors, per
se, with contemporaneous predictors that may actually represent correlates or
consequences of CG. Thus, there remains a need for an empirical review of risk
factors associated with grief-specific distress in the recent literature, with fine-
grained reporting of results from studies that explored the relation of risk factors
to common grief and CG.
We limited our review to independent variables that (a) preceded the loss
(e.g., kinship relationship to deceased), (b) were related to the death itself (e.g.,
cause of death), (c) were static at the time of the loss or during bereavement
(e.g., demographic factors, such as race or gender), or (d) were measured at a
minimum of two time points (e.g., Time 1, social support, predicted Time 2, CG),
excluding cognitive–behavioral factors (e.g., rumination, meaning making) that
could be correlates or consequences of CG, except when their predictive power
was explored in longitudinal studies. We further restricted dependent variables
to assessments of common grief and CG, excluding other negative and positive
outcomes (e.g., depression and posttraumatic growth, respectively). Our rationale
for including common grief in our review was that (a) prior to the creation of a
 Prospective risk factors for CG  147
reliable, standardized, and validated measurement of protracted and maladaptive
grief responses (e.g., ICG; Prigerson, Maciejewski, et al., 1995) researchers were
reliant upon less specialized scales to assess grief in survivors, and (b) contempo-
rary research suggests that grief can be evaluated on a continuum, ranging from
low-level normative grief to a severe grief disorder (Holland et al., 2009). As a
result, assessment of severity of grief across its full range can draw upon measures
of both normative and complicated grief symptomatology.

Procedure
We accessed articles using the PsycINFO and PsycARTICLES online databases
by using the search terms loss, death, grief, complicated grief, bereavement,
mourning, risk factor, and predict. We also used book chapters, our library of
grief-related articles, and the reference lists of other germane studies that emerged
in our search. Studies we included were (a) empirical, (b) quantitative, and (c)
published in English, in a peer-reviewed journal, between 1980 and 2010 (and
earlier seminal work). Although the genesis of CG cannot be firmly established
from non-experimental studies, in order to strengthen inferences related to causal-
ity, we limited our analysis of independent variables to stable factors (e.g., age,
ethnicity) found in cross-sectional studies, or to longitudinal studies measuring
independent variables at one time point that predict later grief. Because people’s
mood states could affect their report of psychological factors associated with the
loss, variables assessing coping behaviors or other cognitive, emotional, or social
processes were included only when studies that examined them used a truly pro-
spective design.
Data for risk factors were analyzed by recording (a) the number of studies that
examined each factor, (b) the number that found it to be a statistically significant
predictor, (c) its relation to grief, and (d) its grouping into one of six categories,
as described below.

Results

Initial analyses
Using 43 studies to explore risk factors of grief, we found that 16 studies measured
only CG using the ICG (Prigerson, Maciejewski, et al., 1995), or its revised ver-
sion, the ICG-R (Prigerson & Jacobs, 2001), 21 measured only grief more gener-
ally using a scale or items designed to measure more normative responses to loss
such as the Core Bereavement Items (Burnett, Middleton, Raphael, & Martinek,
1997), and six measured both using a version of the ICG and at least one other
scale measuring normative responses to loss. A total of 60 distinct independent
variables met our inclusion criteria for risk factors. Of these, 37 risk factors were
statistically significant in predicting grief or CG in at least one study. Risk factors
were collated into the following distinct categories: survivor’s background (e.g.,
gender), death- and bereavement-related (e.g., cause of death), relationship to
148  Laurie A. Burke and Robert A. Neimeyer
the deceased (e.g., kinship), intrapersonal (e.g., attachment style), religion/belief
(e.g., worldview), and interpersonal (e.g., social support).
When analyzed in groups, the death- and bereavement-related, relation to
deceased, and intrapersonal categories had the most statistically significant risk
factors (n = 8 in each), followed by survivor’s background (n = 5), and interper-
sonal, and religion/belief (n = 4 in each). When analyzed individually, 14 factors
emerged as strong indicators of CG (see Table 11.1). These were ranked in order
of the ratio of number of studies finding a given variable significant relative to
the number that explored the variable. In these terms, being female emerged
as the most prominent risk factor for CG, followed by being a spouse or parent
(especially a mother) of the deceased, violent death, low levels of social support,
the deceased’s age (both younger and older), younger age of the bereaved, sud-
denness/unexpectedness of the death, being non-Caucasian, anxious, avoidant, or
having insecure attachment style, discovering or identifying the body (in cases of
violent death), high pre-death marital dependence, high levels of neuroticism, less
education, prior losses, lower income, problematic relationship with the deceased,
recency of the death, and lack of family cohesion.

Subsequent analyses
To increase confidence in our results, we also considered the literature using even
more stringent criteria: confirmed risk factors of CG were each explored in at
least three studies and were found statistically significant more than 50% of the
time. Six such confirmed risk factors emerged: (1) low social support, (2) anxious/
avoidant/insecure attachment style, (3) discovering or identifying the body (in
cases of violent death), (4) being the spouse or parent of the deceased, (5) high
pre-death marital dependence, and (6) high neuroticism. Thirty-two variables
were identified as potential risk factors (explored in fewer than three studies or
found to be statistically significant less than half of the time). These included
being non-Caucasian, younger age of the bereaved, being female, less educa-
tion, low income, violent death, sudden/unexpected death, perception of death
as preventable, prior losses, lack of anticipatory grieving, searching for meaning,
less importance of religion, regular church attendance, lack of spiritual beliefs,
prior mental health counseling, pre-existing psychological condition, lack of
technological connectedness (no use of email, Internet, cell phone), little time
spent talking about the loss, frequent pre-death contact with the deceased, belief
in professional counseling, subjectively close relationship with the deceased,
problematic relationship with the deceased, recency of the death, lack of family
cohesion, deceased’s age (both younger and older), deceased’s gender opposite
of bereaved’s, good pre-death health of deceased, length of illness of deceased
(too long or too short), and negative cognitions related to self, life, the future, and
threatening interpretations of one’s own grief. Although the scope of this chap-
ter precludes reporting on every risk factor in every study, the following review
describes a sampling of risk factors nested within categories.
 Prospective risk factors for CG  149
Table 11.1 Top risk factors of common grief and complicated grief in order of number of
studies in which they were explored

Number Number Percentage Rank

of of studies of studies according Risk
studies statistically statistically to factor
Risk factors explored significant significant percentage typea
Younger age of bereaved 20 (5) 21 10 P
Being female 19 (8) 42 7 b
P
Violent death 17 (6) 35 8 P
Deceased’s age (both 12 (5) 42 7b P
younger/older)
Sudden, unexpected death 12 (5) 42 7b P
Less education 12 (2) 17 11b P
Spouse or parent of 11 (7) 64 5 C
deceased (especially a
mother)
Being non-Caucasian 8 (4) 50 6 P
Low social support 7 (6) 86 2 C
Low income, experienced 6 (2) 33 9c P
prior losses, problematic
relationship with deceased,
recency of loss
Lack of family cohesion 6 (1) 17 11 P
Anxious/avoidant/insecure 5 (4) 80 3 C
attachment style
Found, saw, or identified 3 (3) 100 1 C
the body in cases of violent
death, or issues related to
death notification
High pre-death marital 3 (2) 67 4c C
dependency, high
neuroticism

Notes
a C, confirmed risk factor; P, potential risk factor.
b Risk factors share ranking.
c Ranking represents multiple risk factors.

Survivor’s background

Gender1

Nineteen out of 43 (44%) studies explored the role of gender in grieving. Eight
(42%) found that it was significant. For example, Lang and Gottlieb’s (1993)
150  Laurie A. Burke and Robert A. Neimeyer
study of 57 parents bereaved of infants found that mothers suffered more than
fathers in terms of grief. Spooren, Henderick, and Jannes (2000) found in their
sample of 85 mothers and fathers bereaved by motor vehicle accidents that men
and women did not differ in terms of their general psychological distress. Gender
did, however, predict CG, with women suffering greater complications. Likewise,
Prigerson et al.’s (2002) study found higher rates of CG among 151 female
Pakistani psychiatric patients than among male Pakistani psychiatric patients. In
Keesee et al.’s (2008) study of 157 parents, mothers reported more common grief
than fathers but not more CG. More complex interactions of gender with other
variables have also occasionally been reported, as in Callahan’s (2000) study of
210 people bereaved by suicide in which women who found their loved one’s
body had more grief. However, other studies have found gender to be unrelated
to grief, as in Momartin, Silove, Manicavasagar, and Steel’s (2004) evaluation
of 126 Bosnian refugees in Australia. Nonetheless, when gender differences are
observed, as they often are, evidence indicates that women are more susceptible
to intense and complicated grief reactions than men.

Race

Half of the studies (four out of eight) exploring race reported significant results.
Goldsmith et al. (2008) investigated two samples – 316 bereaved individuals and
222 cancer patients and their caregivers – whereas Neimeyer, Baldwin, and Gillies
(2006) studied a sample of 506 young adults, both finding that African Americans
experienced more grief than Caucasian Americans. Tarakeshwar, Hansen,
Kochman, and Sikkema (2005) compared groups in a sample of 252 HIV-infected
grievers and found that minorities (African Americans and Hispanics) reported
more grief than Caucasians. Likewise, Laurie and Neimeyer’s (2008) study of
1,672 bereaved college students found that being African American predicted
CG, even when controlling for other variables (e.g., length of bereavement and
cause of death). Evaluating grieving parents (n = 52) and spouses (n = 90) in the
United States and the People’s Republic of China (PRC) longitudinally at 4 and
18 months post loss, Bonanno, Papa, Lalande, Zhang, and Noll (2005) found that
initially the PRC sample had higher grief than the U.S. sample, but later Chinese
participants had lower grief than their American counterparts. This suggests the
need to evaluate ethnic variations in bereavement beyond the narrow spectrum of
North American culture and across a longer period, to determine whether certain
ethnic groups are at greater risk of prolonged grief disorder and, if so, what might
account for this effect. Still, it is worth emphasizing that 50% of the studies found
that race is not a risk factor of CG. For example, Carr (2004) compared African
Americans (n = 33 widowed persons and 12 controls) and Caucasians (n = 177
widowed persons and 75 controls) in the Changing Lives of Older Couples
(CLOC) study and found no difference in levels of yearning or grief, just as Cruz
and colleagues (2007) found no differences in African Americans (n = 19) and
Caucasians (n = 19) presenting for CG therapy.
 Prospective risk factors for CG  151

Death- and bereavement-related factors

Cause of death2

Many studies (18) examined this risk factor, and over a third of them (7; 39%)
found cause of death to be related to subsequent grief. Most used cross-sectional
designs, except where noted. Prigerson et al.’s (2002) investigation found that
violent death (murder vs. illness, accident, and drowning) did not predict CG
in 151 bereaved psychiatric patients. However, of those studies in which cause
of death was a predictor, violent death was consistently found to produce more
intense and complicated grief than death due to illness. Cleiren (1993) examined
this factor over time and found that unnatural deaths (suicide or motor vehicle
accident [MVA] vs. extended illness) led to greater grief in 309 parents and
spouses, and that suicide bereaved were the most preoccupied with their loss.
At Time 1 (T1; 4 months post loss), families bereaved by MVAs had more grief
than those bereaved by suicide or illness, but at T2 (14 months) cause of death
was no longer a risk factor. Gamino, Sewell, and Easterling (2000) compared
85 people bereaved by illness, homicide, suicide, and accident, and found that
traumatic deaths produced more grief. Likewise, Currier, Holland, Coleman,
and Neimeyer’s (2007) cross-sectional investigation of 1,723 bereaved college
students indicated that people bereaved by violent death (accident, suicide, and
homicide) had more severe grief than those experiencing a loss through natural,
anticipated death or natural, sudden death. Specifically, in terms of CG, they found
no statistically significant difference among accident, suicide, and homicide, but
scores were higher for violent deaths than for natural, anticipated deaths, and
homicide and accident deaths produced more CG than did natural sudden deaths.
In terms of common grief, they found that homicide produced substantially higher
scores than all other types of deaths. Looking at both common grief and CG,
Keesee et al. (2008) found higher grief in 94 violently bereaved parents than in 63
parents bereaved by other means. Finally, Momartin et al.’s (2004) examination
of 126 Bosnian refugees indicated that the traumatic loss of a family member was
the strongest risk factor for CG.

Peri-event variables

One hundred percent of the small number of cross-sectional studies (three) exam-
ining peri-death variables found a relation with grief. To illustrate, Spooren et
al. (2000) assessed the support that 85 parents bereaved by MVAs received after
the death, and found that dissatisfaction with material help and with informa-
tion given about the event predicted CG. With 540 suicidally bereaved parents,
Feigelman, Jordan, and Gorman (2009) found that survivors who saw or found
the body had significantly greater grief than those who did not view the body
prior to the funeral. In fact, discovering the body proved to be the strongest risk
factor of grief. In another suicide study (n = 210), seeing the body at the scene
152  Laurie A. Burke and Robert A. Neimeyer
of the death intensified grief, as did being the one to find it  – especially for
women (Callahan, 2000). However, stepwise analyses revealed that finding the
body was not more grief producing than simply seeing the body, and viewing
the deceased’s body at the funeral did not increase grief. Nor was the specific
weapon type or suicide method (e.g., hanging) associated with grief outcomes,
even when comparing the use of guns (the most common method; 47%) with
seven other methods.

Relationship to the deceased

Kinship

Fully two thirds of studies of kinship (e.g., spouse, parent, child) demonstrated a
link to intensified grief (e.g., Boelen, van den Bout, & van den Hout, 2003). For
example, Laurie and Neimeyer’s (2008) cross-sectional sample of 1,670 bereaved
college students reported a main effect for kinship in predicting CG, such that
students bereaved of immediate family had more grief than those bereaved of
more distant relationships. In Cleiren’s (1993) longitudinal study (n = 309), kin-
ship proved the strongest predictor of grief, explaining 15% of the variance in T2
(14 months post loss) scores, such that parents and spouses grieved more severely
than did adult children or siblings. Even when the ages of both the bereaved and
the deceased child were controlled, grief was higher for mothers at 4 and 14
months post loss, and recovery was slower. Differences among kinship categories
are sometimes observed as well. For example, Prigerson and colleagues’ (2002)
cross-sectional examination found that spouses and parents (n = 151) were far
more likely (22 and 11 times, respectively) to have CG than other kinship types.
Bonanno et al.’s (2005) longitudinal assessment found no differences in spouse
(n = 90) and parent (n = 52) grieving at T1 (4 months); but, at T2 (14 months), par-
ents’ scores were higher than spouses’. Occasionally studies qualify this general
trend linking kinship with higher risk of intense grief, as in the finding by van
der Houwen and colleagues (2010) in their longitudinal study of 195 bereaved
individuals, which showed that partner loss predicted emotional loneliness, but
kinship did not predict grief more generally.

Marital dependency

Two out of three longitudinal studies found a relation between the mourner’s pre-
loss dependency upon his or her spouse and subsequent grief. However, both
studies with significant results used the CLOC data (n = 205 and 210 widowed
persons), so that this finding stands in need of replication. Bonanno et al. (2002)
found that pre-loss spousal dependency was associated with subsequent chronic
grieving as opposed to resilience, and Carr (2004) found that spousal dependency
was a risk factor for despair, a specific dimension of grief. Cleiren’s (1993) study
of 309 survivors of MVA loss, on the other hand, yielded null findings.
 Prospective risk factors for CG  153

Intrapersonal factors

Attachment style (avoidant, anxious, insecure)

Although too infrequently studied, attachment styles were associated with grief
in three out of four of the longitudinal studies we reviewed. For example, van
der Houwen et al.’s (2010) final statistical model in their study of 195 bereaved
individuals indicated that avoidant but not anxious attachment predicted higher
levels of CG. Using the CLOC data (n = 103), Brown, Nesse, House, and Utz
(2009) found that pre-loss insecure attachment style and grief were related at 6,
24, and 48 months. In two studies of 219 bereaved parents, Wijngaards-de Meij
and colleagues (2007a, 2007b) showed that avoidant and anxious attachment
styles explained 13% of the variance in CG scores. On the other hand, results
are not fully consistent. Bonanno et al. (2002) examined avoidant and dismissive
attachment in a study of 205 conjugally bereaved people and found that both were
unrelated to grief.

Neuroticism

Similarly, two out of three studies linked neuroticism with grief. For instance, in
their final regression analysis of their longitudinal study of 195 grievers, van der
Houwen et al. (2010) found a statistically significant relation between neuroticism
and CG. In Wijngaards-de Meij et al.’s (2007a) study with 219 bereaved parents,
attachment coupled with neuroticism explained 22% of the variance in CG scores,
with neuroticism alone accounting for 18% of the total variance. Yet Bonanno and
colleagues (2002) found no association between neuroticism and grief in their
longitudinal study of 205 elderly spouses.

Searching for meaning and sense making

Coleman and Neimeyer (2010) used the CLOC study’s prospective design to show
that engaging in a search for meaning predicted both concurrent and prospective
grief in a sample of bereaved spouses (n = 250). Specifically, those who struggled
to make sense of the loss 6 and 18 months post loss had higher subsequent grief
scores fully 4 years after the death. Interestingly, however, sense making, when it
did occur, emerged as a strong positive predictor of subsequent well-being (e.g.,
interest, excitement, accomplishment), rather than an inverse predictor of grief
symptomatology per se.

Negative cognitions

Including a sample of 97 mourners in their longitudinal study, Boelen, van den

Bout, and van den Hout (2006) found that bereavement-associated negative
cognitions at T1 (1–4 months post loss) related to the griever’s self, life, future,
154  Laurie A. Burke and Robert A. Neimeyer
or threatening interpretations of grief (e.g., “If I fully realized what the death
of ____ means, I would go crazy”) each individually predicted CG at T2 (7–10
months post loss), even after controlling for background variables and CG symp-
tomatology at T1. Additionally, the future subscale (e.g., “In the future, I will
never become really happy any more”) was the only subset of negative cognitions
at T1 that enduringly predicted CG at T3 (16–19 months post loss), with an inter-
action of threatening interpretations × avoidance of the reality of the loss likewise
prospectively predicting CG at T3 in hierarchical multiple regressions.

Religion/belief factors

Importance of religion

The few prospective studies conducted on the importance of religion yielded

inconsistent evidence for its relation to grief. For example, Brown et al. (2009)
used the CLOC data to monitor grief in 103 spouses. They found that assign-
ing greater importance to religious/spiritual beliefs pre-loss predicted less grief
at 6 and 18 months post loss, although high church attendance was unrelated.
Conversely, when Kersting et al. (2007) compared grief in 62 women who had
terminated their pregnancy with 65 women who birthed a full-term baby, they
found that those who placed higher import on faith grieved the hardest, possibly
reflecting guilt or regret about their decision to abort the fetus.

Spiritual belief or worldview

These risk factors produced equivocal results across two studies. In Bonanno
et al.’s (2002) prospective study (n = 205) bereaved spouses who were the most
resilient also had greater acceptance and believed in a just world. However, no
connection between chronic grief and a dysfunctional worldview was discovered.
Yet Easterling, Gamino, Sewell, and Stirman (2000) found in their cross-sectional
study of 85 bereaved individuals that spiritual beliefs about one’s relationship
with God or events that increase belief in God’s existence were related to less
grief.

Interpersonal factors

Social support

Six out of seven (86%) longitudinal studies found that low levels of social sup-
port predicted intensified grief (e.g., Bonanno et al., 2002). Vanderwerker and
Prigerson (2004) prospectively examined 293 older people and found that higher
social support at 6 months post loss forecast less CG near the end of the first year.
However, Gamino, Sewell, and Easterling’s (1998) cross-sectional study found
that grief and social support were unrelated in a sample of 74 mourners.
 Prospective risk factors for CG  155

Discussion
The multidimensional nature of adaptation to loss poses challenges to the iden-
tification of risk factors predicting prolonged and intense grieving. One specific
challenge concerns the basic understanding of CG as a construct. According to
those who conceptualize CG as existing on a continuum, differences between
grievers whose response to loss warrants treatment and those whose does not are
reflected in the duration and intensity of symptoms and levels of impairment, not
in distinctive symptoms. Nevertheless, lacking a genuine cut-point where grief
responses are considered in need of treatment, researchers and clinicians must
make personal or consensual judgments about a given griever’s level of impair-
ment and distress. On the other hand, finer discrimination of distress that spans the
range of grief responses provides sensitivity that may be lost in models that insist
on bifurcation of high-/low-distress respondents.
Likewise, the contrasting view held by some researchers that common grief
and CG are symptomatically different carries implications for understanding of
the grieving process. On the one hand, assessing grief in this way may blur vari-
ability in responses that represent different points on the same continuum. This
view also carries the potential for social and personal stigma in suggesting that
some individuals are grieving in a diagnosably disordered manner. Conversely,
this model’s clear identification of cases of CG could simplify communication
among mental health professionals and more readily specify who is or is not in
need of treatment.
Viewing grief in dimensional terms, we systematically sought out studies that
explored antecedents and predictors of CG, and were limited only by the types of
factors explored in the primary studies. We identified more consistent prospec-
tive predictors of intense grieving, as well as those that were potential factors in
forecasting grief outcomes. Risk factors that emerged as most salient included
low levels of social support, avoidant/anxious/insecure attachment style, discov-
ering the body (in cases of violent death) or dissatisfaction with death notifica-
tion, being a spouse or a parent of the deceased, high levels of pre-death marital
dependency, and high levels of neuroticism. Inasmuch as CG is conceptualized
as an attachment-based disorder, with symptomatology indicative of separation
distress and preoccupation with the deceased, it is understandable that mourners
who are vulnerable to feeling abandoned and alone, who suffer from excessive
anxiety or obsession, and who lose a security-enhancing or care-providing rela-
tionship, under conditions of minimal support, and perhaps in circumstances that
leave them struggling with posttraumatic imagery, would be especially prone to
the development of CG.
In addition to these primary conclusions, studies further suggest that being
young, being non-Caucasian, having less education, little income, prior losses,
or losing a child of any age to a violent, sudden death tends to predict pro-
longed and intense grief. Unfortunately, few of those potential risk factors are
modifiable in the context of therapy. This highlights the importance of studying
156  Laurie A. Burke and Robert A. Neimeyer
those predictors of poor outcome that in principle are modifiable, as intervention
could focus usefully on strengthening social integration of the bereaved (Burke,
Neimeyer, & McDevitt-Murphy, 2010), facilitating the use of their spiritual or
secular philosophies as a psychological resource (Park & Halifax, 2011), chal-
lenging their dysfunctional interpretations or predictions about themselves and
the future (Boelen et al., 2006), joining them in their quest for meaning in a
senseless loss (Neimeyer, Burke, Mackay, & van Dyke-Stringer, 2010), and
strengthening their continuing bond with the deceased so as to enhance their
attachment security (Field & Wogrin, 2011). Fortunately, empirically informed
therapies that pursue such goals are currently being developed (Neimeyer, Harris,
Winokuer, & Thornton, 2011).

Limitations to this review

Generally speaking, research on risk factors is rife with complexity. For instance,
because we sought to include only truly prospective factors in our review, poten-
tially modifiable variables included only in correlational studies were excluded
(although static variables in the same studies might be included). Other challenges
included collating the variety of measures and items used for assessing grief,
interpreting poorly defined variables, and deciphering vague reporting of results,
all of which required some level of subjective judgment.
Moreover, some variables represented in this review are likely to have equal
relevance in predicting grief and more general symptomatology (e.g., depression,
suicidality), whereas other factors may differentially predict grief and non-grief
outcomes in ways that could not be addressed in this review. For instance, some
studies have shown that men suffer more depression than women following
spousal loss (Stroebe & Stroebe, 1983), in contrast to findings reviewed above
implicating female gender with greater grieving. Furthermore, the search for uni-
versal risk factors also can be complicated by the different cultural, economic, and
political contexts in which studies are conducted. For example, concern for health
care disparities experienced by different ethnicities could make investigation of
racial differences in CG a high priority in the United States, whereas in some
European countries such practices could be considered inappropriate.
To strengthen causal inferences, we excluded risk factors that were likely to
change as a result of the loss, or that were simply correlated with grief, focusing
instead on longitudinal designs and stable independent variables (e.g., gender
of the mourner). These exclusions meant that most studies on coping strategies,
cognitions, or meaning making in bereavement – factors that are perhaps the most
malleable and amenable to therapeutic intervention – were given little attention,
simply because the majority of research to date on these factors utilized cross-
sectional designs. We were similarly limited by the paucity of studies exploring
mediators or moderators of grief,3 and by the equivocal findings across studies,
which probably result from variations in methodological standards and measure-
ment tools. Use of state-of-the-science measures and consideration of moderator
variables could yield more precise findings in future research. Finally, we were
 Prospective risk factors for CG  157
further restricted by our exclusion of qualitative studies that may have suggested
a richer understanding of the CG experience.
Although we have made a concerted attempt to isolate and categorize pro-
spective risk factors that predict an intense grieving process in order to make
definitive statements about who is susceptible to CG, a meta-analytical review of
risk factor effect sizes would be desirable. Perhaps most informative of all would
be one that includes other relevant bereavement outcomes (e.g., depression, post-
traumatic stress), both longitudinal and cross-sectional, including samples diverse
in age, ethnicity, and nationality. A more comprehensive analysis along these lines
would provide a broader picture of why some people are more likely than others
to develop intense and sustained difficulties in bereavement, as opposed to grief
symptoms alone.

Clinical implications and future directions for research

A systematic empirical review of this sort is valuable for many reasons. First,
although CG is concerning in itself, it is often not the final outcome in mourning
gone awry; rather, CG can function as a risk factor for even more dire psycho-
logical and physical health problems. For example, CG subsequently predicts
cardiovascular illness (Prigerson et al., 1997), suicidality (Latham & Prigerson,
2004), substance abuse, depression, anxiety, and overall life disruption (Ott, 2003;
Shear et al., 2011).
Moreover, identification of CG predictors enables clinicians to provide proac-
tive assistance, especially to individuals facing expected deaths (e.g., of a loved
one with cancer), or to those at special risk in the wake of a widespread trauma
or disaster. For example, Rando (1983) found in her study of 54 parents bereaved
by their child’s cancer that, when anticipatory grieving was done prior to the
death, less disordered grieving occurred afterward. Additionally, identification
of modifiable risk factors can guide the development of relevant secondary or
tertiary interventions. For instance, armed with the knowledge that poor social
support poses risks to bereavement adaptation (Burke et al., 2010), health care
professionals in end-of-life contexts could assess families’ social support before
bereavement begins, with an eye toward preventive intervention.
Finally, the argument for a continuing focus on risk factors of CG in bereave-
ment research is, perhaps, made more evident in terms of the tremendous need
for more widespread understanding of this life-vitiating condition. As a whole,
medical and mental health professionals, clergy, and society alike are woefully
uninformed about CG, its predictors, or its treatment. Disseminating knowledge
to frontline workers and to the grievers themselves could be a first step toward
prevention and treatment of a life-limiting response to loss.

Acknowledgment
The authors gratefully acknowledge the invaluable help of Natalie L. Davis in
this work.
158  Laurie A. Burke and Robert A. Neimeyer

Notes
1 Comparisons of mothers versus fathers were reported in the gender category rather
than under kinship, which compared several relationship types (i.e., parents, siblings,
spouses).
2 See Chapter 20 in this volume for commentary on the definition and categorization
of violent versus non-violent death loss.
3 An exception to this was Boelen and colleagues’ (2006) study that examined nega-
tive thoughts/beliefs, cognitive–behavioral avoidance strategies, and interactions
between the two.

References
Boelen, P.  A., van den Bout, J., & van den Hout, M. A. (2003). The role of negative
interpretations of grief reactions in emotional problems after bereavement. Journal of
Behavior Therapy and Experimental Psychiatry, 34, 225–238.
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2006). Negative cognitions and
avoidance in emotional problems after bereavement: A prospective study. Behavior
Research and Therapy, 44, 1657–1672.
Bonanno, G.  A., & Kaltman, S. (2001). The varieties of grief experience. Clinical
Psychology Review, 21, 705–734.
Bonanno, G. A., & Mancini, A.  D. (2006). Bereavement-related depression and PTSD:
Evaluating interventions. In Barbanel, L., & Sternberg, R.  J. (Eds.), Psychological
interventions in times of crisis (pp. 37–55). New York: Springer.
Bonanno, G. A., Papa, A., Lalande, K., Zhang, N., & Noll, J. G. (2005). Grief process-
ing and deliberate grief avoidance: A prospective comparison of bereaved spouses and
parents in the United States and the People’s Republic of China. Journal of Counseling
and Clinical Psychology, 73(1), 86–98.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring, M., Sonnega, J.,
et al. (2002). Resilience to loss and chronic grief. Journal of Personality and Social
Psychology, 83, 1150–1164.
Brown, S.  L., Nesse, R.  M., House, J.  S., & Utz, R.  L. (2009). Religion and emotional
compensation: Results from a prospective study of widowhood. Society for Personality
and Social Psychology, 30, 1165–1174.
Burke, L.  A., Neimeyer, R.  A., & McDevitt-Murphy, M.  E. (2010). African American
homicide bereavement: Aspects of social support that predict complicated grief, PTSD
and depression. Omega, 61, 1–24.
Burnett, P., Middleton, W., Raphael, B., & Martinek, N. (1997). Measuring core bereave-
ment phenomena. Psychological Medicine, 27, 49–57.
Callahan, J. (2000). Predictors and correlates of bereavement in suicide support group
participants. Suicide and Life Threatening Behavior, 30, 104–124.
Carr, D. S. (2004). African American/Caucasian differences in psychological adjustment to
spousal loss among older adults. Research on Aging, 26, 591–622.
Cleiren, M. (1993). Bereavement and adaptation: A comparative study of the aftermath of
death. Washington, DC: Hemisphere.
Coleman, R. A., & Neimeyer, R. A. (2010). Measuring meaning: Searching for and making
sense of spousal loss in later life. Death Studies, 34, 804–834.
Cruz, M., Scott, J., Houck, P., Reynolds, C.  F. III., Frank, E., & Shear, M.  K. (2007).
Clinical presentation and treatment outcome of African Americans with complicated
grief. Psychiatric Services, 58, 700–702.
 Prospective risk factors for CG  159
Currier, J. M., Holland, J., Coleman, R., & Neimeyer, R. A. (2007). Bereavement follow-
ing violent death: An assault on life and meaning. In Stevenson, R., & Cox, G. (Eds.),
Perspectives on violence and violent death (pp. 175–200). Amityville, NY: Baywood.
Easterling, L. W., Gamino, L. A., Sewell, K. W., & Stirman, L. S. (2000). Spiritual experi-
ence, church attendance, and bereavement. Journal of Pastoral Care, 54, 263–275.
Feigelman, W., Jordan, J. R., & Gorman, B. S. (2009). How they died, time since loss, and
bereavement outcomes. Omega: Journal of Death and Dying, 58, 251–273.
Field, N. P., & Wogrin, C. (2011). The changing bond in therapy for unresolved loss. In
Neimeyer, R. A., Harris, D., Winokuer, H., & Thornton, G. (Eds.), Grief and bereave-
ment in contemporary society (pp. 37–46). New York: Routledge.
Gamino, L. A., Sewell, K. W., & Easterling, L. W. (1998). Scott & White Grief Study: An
empirical test of predictors of intensified mourning. Death Studies, 22, 333–355.
Gamino, L. A., Sewell, K. W., & Easterling, L. W. (2000). Scott & White grief study phase
2: Toward an adaptive model of grief. Death Studies, 24, 633–660.
Goldsmith, B., Morrison, R.  S., Vanderwerker, L.  C., & Prigerson, H. (2008). Elevated
rates of prolonged grief disorder in African Americans. Death Studies, 32, 352–365.
Holland, J.  M., Neimeyer, R. A., Boelen, P. A., & Prigerson, H.  G. (2009). The under-
lying structure of grief. Journal of Psychopathology and Behavioral Assessment, 31,
190–201.
van der Houwen, K., Stroebe, M., Stroebe, W., Schut, H., van den Bout, J., & Wijngaards-de
Meij, L. (2010). Risk factors for bereavement outcome: A multivariate approach. Death
Studies, 34, 195–220.
Keesee, N.  J., Currier, J.  M., & Neimeyer, R. A. (2008). Predictors of grief following
the death of one’s child: The contribution of finding meaning. Journal of Clinical
Psychology, 64, 1–19.
Kersting, A., Kroker, K., Steinhard, J., Ludorff., K., Wesselmann., U., & Ohrmann, P.
(2007). Complicated grief after traumatic loss: A 14-month follow-up study. European
Archive of Psychiatry Clinical Neuroscience, 257, 437–443.
Lang, A., & Gottlieb, L. (1993). Parental grief reactions and marital intimacy following
infant death. Death Studies, 17, 233–255.
Latham, A., & Prigerson, H. (2004). Suicidality and bereavement. Suicide and Life
Threatening Behavior, 34, 350–362.
Laurie, A., & Neimeyer, R. A. (2008). African Americans and bereavement: Grief as a
function of ethnicity. Omega, 57, 173–193.
Lichtenthal, W. G., Cruess, D. G., & Prigerson, H. G. (2004). A case for establishing com-
plicated grief as a distinct mental disorder in DSM-V. Clinical Psychology Review, 24,
637–662.
Lobb, E. A., Kristjanson, L. J., Aoun, S. M., Monterosso, L., Halkett, G. K. B., & Davies,
A. (2010). Predictors of complicated grief: A systematic review of empirical studies.
Death Studies, 34, 673–698.
McDevitt-Murphy, M. E., Neimeyer, R. A., Burke, L. A., & Williams, J. L. (2012). Assessing
the toll of traumatic loss: Psychological symptoms in African Americans bereaved by
homicide. Psychological Trauma: Theory, Research, and Policy, 4, 303–311.
Momartin, S., Silove, D., Manicavasagar, V., & Steel, Z. (2004). Complicated grief in
Bosnian refugees. Comprehensive Psychiatry, 45, 475–482.
Neimeyer, R. A., Baldwin, S. A., & Gillies, J. (2006). Continuing bonds and reconstructing
meaning: Mitigating complications in bereavement. Death Studies, 30, 715–738.
Neimeyer, R. A., Burke, L., Mackay, M., & van Dyke-Stringer, J. (2010). Grief therapy and
the reconstruction of meaning: From principles to practice. Journal of Contemporary
Psychotherapy, 40, 73–83.
160  Laurie A. Burke and Robert A. Neimeyer
Neimeyer, R.  A., Harris, D., Winokuer, H., & Thornton, G. (Eds.). (2011). Grief and
bereavement in contemporary society: Bridging research and practice. New York:
Routledge.
Ott, C. H. (2003). The impact of complicated grief on mental and physical health at various
points in the bereavement process. Death Studies, 27, 249–272.
Park, C., & Halifax, J. (2011). Religion and spirituality in adjusting to bereavement. In
Neimeyer, R. A., Harris, D., Winokuer, H., & Thornton, G. (Eds.), Grief and bereave-
ment in contemporary society (pp. 355–364). New York: Routledge.
Prigerson, H., Ahmed, I., Silverman, G.  K., Saxena, A.  K., Maciejewski, P.  K., Jacobs,
et al. (2002). Rates of risks of complicated grief among psychiatric clinic patients in
Karachi, Pakistan. Death Studies, 26, 781–792.
Prigerson, H.  G., Beirhals, A.  J., Kasl, S. V., Reynolds, C.  F., Shear, K., Day, N., et al.
(1997). Traumatic grief as a risk factor for mental and physical morbidity. American
Journal of Psychiatry, 154, 616–623.
Prigerson, H. G., Frank, E., Kasl, S., Reynolds, C., Anderson, B., Zubenko, G. S., et al.
(1995). Complicated grief and bereavement related depression as distinct disorders.
American Journal of Psychiatry, 152, 22–30.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine, 6(8), 1–12.
Prigerson, H. G., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale,
consensus criteria, and a preliminary empirical test. In Stroebe, M. S., Hansson, R. O.,
Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research (pp.  613–645).
Washington, DC: American Psychological Association.
Prigerson, H. G., Maciejewski, P., Reynolds, C. F., Beirhals, A. J., et al. (1995). Inventory
of Complicated Grief: A scale to measure maladaptive symptoms of loss. Psychiatry
Research, 59, 65–79.
Rando, T. A. (1983). An investigation of grief and adaptation in parents whose children
have died of cancer. Journal of Pediatric Psychology, 8, 3–20.
Sanders, C. M. (1988). Potential risk factors in bereavement outcome. Journal of Social
Issues, 44, 97–111.
Shear, K., Frank, E., Houch, P. R., & Reynolds, C. F. (2005). Treatment of complicated
grief: A randomized controlled trial. Journal of the American Medical Association, 293,
2601–2608.
Shear, M.  K., Jackson, C.  T., Essock, S.  M., Donahue, S.  A., & Felton, C.  J. (2006).
Screening for complicated grief among Project Liberty service recipients 18 months
after September 11, 2001. Psychiatric Services, 57, 1291–1297.
Spooren, D. J., Henderick, H., & Jannes, C. (2000). Survey description of stress of parents
bereaved from a child killed in a traffic accident. Omega, 42, 171–185.
Stroebe, M., Schut., H., & Stroebe, W. (2007). Health outcomes in bereavement. The
Lancet, 370, 1960–1073.
Stroebe, M. S., & Stroebe, W. (1983). Who suffers more? Sex differences in health risks of
the widowed. Psychological Bulletin, 93, 279–301.
Stroebe, W., & Schut, H. (2001). Risk factors in bereavement outcome: A methodological
and empirical review. In Stroebe, M.  S., Hansson, R. O., Stroebe, W., & Schut, H.
(Eds.), Handbook of bereavement research: Consequences, coping and care (pp. 349–
371). Washington, DC: American Psychological Association.
Tarakeshwar, N., Hansen, N., Kochman, A., & Sikkema, K. J. (2005). Gender, ethnicity and
spiritual coping among bereaved HIV-positive individuals. Mental Health, Religion, &
Culture, 8, 109–125.
 Prospective risk factors for CG  161
Vanderwerker, L.  C., & Prigerson, H.  G. (2004). Social support and technological con-
nectedness as protective factors in bereavement. Journal of Loss and Trauma, 9, 45–57.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., & Heijden,
P. G. M. (2007). Neuroticism and attachment insecurity as predictors of bereavement
outcome. Journal of Research and Personality, 41, 498–505.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., & Heijden,
P. G. M. (2007). Patterns of attachment and parents’ adjustment to the death of their
child. Personality and Social Psychology Bulletin, 33, 537.
12 Repetitive thought
Rumination in complicated grief
Edward R. Watkins and Michelle L. Moulds

Repetitive thought (RT) is defined as the ‘process of thinking attentively, repeti-

tively or frequently about one’s self and one’s world’ and has been proposed to
form ‘the core of a number of different models of adjustment and maladjustment’
(Segerstrom, Stanton, Alden, & Shortridge, 2003, p. 3). Repetition is a property
that is common to a range of thought processes that have implications for self-
regulation, psychopathology, and mental and physical health, including worry,
rumination, cognitive processing, rehearsal, reflection, and problem-solving.
Watkins (2008) argued that organizing our knowledge of these different processes
around the generic construct of RT provides a means to integrate a number of
disparate literatures, without being limited by preconceptions or different termi-
nology. Importantly, RT encompasses processes that are relevant to both normal
and pathological grief.
Thus, RT includes depressive rumination, which is defined as ‘passively
and repetitively focusing on one’s symptoms of distress and the circumstances
surrounding these symptoms’ (Nolen-Hoeksema, McBride, & Larson, 1997,
p. 855). Rumination prospectively predicts the onset and maintenance of depres-
sion (Nolen-Hoeksema, 2000), and is hypothesized to be a maladaptive process
that prolongs grief reactions (Bonanno, Papa, Lalande, Zhang, & Noll, 2005; M.
Stroebe et al., 2007). RT may also occur within more adaptive processes, such as
cognitive processing; that is, actively thinking about a stressor, the thoughts and
feelings it evokes, and its implications for one’s life and future (Bower, Kemeny,
Taylor, & Fahey, 1998; Greenberg, 1995). Cognitive processing accounts hypoth-
esize that, in response to a stressful experience, people think repetitively about the
event in order to integrate and incorporate it into their existing meaning structures
(e.g. Janoff-Bulman, 1992).
As this comparison highlights, a key question within the study of normal and
pathological grief is whether RT about loss has unconstructive versus construc-
tive consequences. We review the relevant theories and evidence concerning
the different consequences of RT following bereavement. We then propose an
integrative solution to this question that we frame within a recent theory that
hypothesizes that the consequences of RT vary according to its content (positive
versus negative) and the way in which individuals engage in RT (abstract versus
concrete; Watkins, 2008).
 Rumination in CG  163

Repetitive thought and grief: theory

The leading cognitive conceptualization of RT is based on control theory
approaches that propose that rumination is triggered by a discrepancy in goal
progress. Within this account, RT serves the function of facilitating progress
towards the desired goal by increasing awareness and focus on the unresolved
issue (Martin & Tesser, 1996). RT continues either until the goal is met or until the
individual disengages from and abandons the goal. Increased focus on the discrep-
ancy will be constructive if the individual makes progress towards the goal, but
will only highlight the discrepancy and increase distress if the discrepancy cannot
be resolved (e.g. if the goal is unrealistic/unattainable and cannot be altered). This
process also applies to discrepancies that result from events that do not fit into an
individual’s mental models of the world. Moreover, the discrepancies that initiate
RT are often a consequence of loss events. Since bereavement is one of the most
common and yet most personally significant loss events, this theory proposes that
RT will be a common response to bereavement.
The grief literature also suggests that RT is an important process in griev-
ing, with a similar debate about whether it is constructive or unconstructive.
Historically, ‘grief work’ models have argued that it is necessary to work through
the negative emotions, thoughts, and memories associated with loss in order to
facilitate adjustment (e.g. Bowlby, 1980). These approaches by definition impli-
cate RT about the deceased and the loss in the process of recovery. These theories
have been challenged for having unspecified operationalizations of ‘working
through’ and because there is limited evidence that confronting feelings and grief-
related memories in a non-structured way is adaptive (e.g. Bonanno & Kaltman,
1999; W. Stroebe, Schut, & M. Stroebe, 2005). Nonetheless, more recent cogni-
tive processing accounts (e.g. Bower et al., 1998; Lepore, Silver, Wortman, &
Wayment, 1996) and cognitive–behavioural accounts of grieving (e.g. Boelen,
van den Hout, & van den Bout, 2006b) propose a role for RT in integrating the
loss into existing mental structures.
More specifically, Bonanno and colleagues have postulated the grief work as
rumination hypothesis, which proposes that grief processing involves pathologi-
cal RT about the loss that exacerbates rather than ameliorates grief-related distress
(e.g. Bonanno, Papa, & O’Neill, 2002; Bonanno et al., 2005). In the dual-process
model, M. Stroebe and Schut (1999) propose that grief processing, presumably
including RT about the loss, may be adaptive, but only when balanced with
avoidance and other restoration-coping responses. By implication, this account
suggests that there is an optimal amount of RT: too little will prevent effective
adjustment, whereas too much will exacerbate distress.
We propose that, rather than adopt an all-or-nothing view of the consequences
of RT, it better reflects the evidence on the effects of RT (Watkins, 2008), as well
as evidence that RT is both a normal response within non-pathological grief and
a characteristic symptom of complicated grief (CG), to hypothesize that RT in
the aftermath of loss can have constructive or unconstructive consequences. In
the initial months following bereavement, nearly all individuals, including those
164  Edward R. Watkins and Michelle L. Moulds
who do not develop chronic difficulties, report rumination and daily thoughts
about the deceased (Bonanno & Kaltman, 2001; Bonanno, Wortman, & Neese,
2004). Nonetheless, the majority of bereaved individuals (85–90%) adapt to their
bereavement in the initial months, such that they no longer experience distress
about the loss (Bonanno & Kaltman, 2001; Bonanno et al., 2004). Thus, for the
majority of people who respond adaptively to bereavement, RT experienced in
the initial months of normal grief cannot have a powerful unconstructive effect:
it is an epiphenomenon with little impact on outcome, or it has a beneficial effect
on outcome, or it may be the result of another process. Nonetheless, elevated
levels of RT about the loss have been identified in the 10–15% of individuals with
chronic grief symptoms (e.g. Bonanno et al., 2004), that is, diagnosed with CG
(also prolonged grief or traumatic grief). CG is characterized by a persistent sense
of yearning for the deceased, difficulty accepting or believing the loss, bitterness,
lack of trust, and loss of perceived meaning in life. To receive a CG diagnosis,
symptoms must be present for at least 6 months and cause functional impairment
(e.g. Lichtenthal, Cruess, & Prigerson, 2004; Prigerson et al., 2009). Yearning
for or preoccupation with the deceased is a key symptom of CG (Prigerson et al.,
2009).

Repetitive thought and grief: evidence

Unconstructive effects of RT following bereavement

There is growing evidence that RT can have negative consequences for adjust-
ment to a bereavement. The strongest evidence comes from longitudinal prospec-
tive studies in which RT at an initial assessment predicted negative outcomes
such as depression at a subsequent assessment. Unfortunately, no studies have
directly examined whether RT prospectively predicts specific CG symptoms such
as yearning. Evidence that RT prospectively predicts depression is consistent
with the hypothesis that RT is involved in contributing to chronic, dysfunctional
patterns of grief, although it needs to be treated cautiously given evidence that
CG is distinctive and distinguishable from depression (e.g. Bonanno et al., 2004;
Prigerson et al., 2009).
Nolen-Hoeksema, Parker, and Larson (1994) examined 253 adults who lost
a family member to a terminal illness, and found that individuals with a more
depressive ruminative style (using the Response Styles Questionnaire, RSQ) at
1 month post loss were more depressed 6 months later, controlling for baseline
depression and demographic variables. Similarly, Nolen-Hoeksema et al. (1997)
found that, in 30 gay men whose partners had recently died from AIDS, codings
of RT from interview transcripts prospectively predicted levels of distress 1 year
later, although this effect was lost after controlling for baseline symptoms. Lepore
et al. (1996) found that high levels of RT in the first 3 weeks after the death of an
infant child predicted mothers’ later depression, but only for those mothers who
perceived social constraints about talking about their loss.
Bonanno et al. (2005) prospectively examined measures of grief processing,
 Rumination in CG  165
deliberate grief-related avoidance, and distress at 4 and 18 months post-bereave-
ment in individuals from the USA and China who had experienced the death of
a spouse or child. The grief-processing measure included an assessment of the
frequency of RT about the deceased. Consistent with the grief work as rumination
hypothesis, independent of 4-month post-bereavement distress, grief processing
and deliberate avoidance at 4 months predicted distress 18 months later in the
U.S. sample. Similarly, Bonanno et al. (2004) found that, for widows who were
not already depressed before their bereavement, more frequent talking and think-
ing about their loss 6 months post-bereavement predicted elevated depression at
18 months post-bereavement. However, neither of these studies included a pure
index of RT, so it is unknown which element of the grief-processing construct pre-
dicted increasing symptoms. Moreover, there is a question concerning the selec-
tion of the timing of the follow-up assessments. It is possible that beneficial grief
processing, and, by extension, constructive forms of RT, might be more common
in the first 1–4 months of bereavement.
Finally, several cross-sectional studies have examined the relationship between
RT, coping responses, and symptoms following bereavement. These studies are
limited by their correlational nature, leaving unresolved the direction of the causal
relationship between RT and symptoms. In a Japanese study of parents who had
experienced the loss of a child, RT was significantly associated with the onset of
major depression, after controlling for demographic variables (Ito et al., 2003).
Michael and Snyder (2005) found that increased RT was associated with reduced
well-being in bereaved students. Davis, Lehman, Wortman, Silver, and Thompson
(1995) reported that RT following uncontrollable and traumatic events, such as
sudden infant death, was associated with a greater level of distress.
In a prospective study, Boelen, van den Bout, and van den Hout (2006a)
examined the relationship between cognitive and behavioural avoidance and the
symptoms of CG. The avoidance measure included items that assessed RT (e.g. ‘I
ruminate about the question why he/she died’). Although avoidance was associ-
ated with concurrent CG symptoms, it did not predict subsequent CG symptoms
after baseline symptoms were controlled. However, since the measure assessed
other aspects of avoidance as well as RT, we cannot be confident that the pattern
of findings reflects assessment of RT.
Also relevant is the extensive evidence that RT in response to loss or trauma
events can have unconstructive consequences (Watkins, 2008). The response
to bereavement theoretically fits within this broader class of loss and traumatic
events, and we propose that the same mechanisms and processes should be
relevant. This evidence includes the finding that RT about a traumatic event
predicts the persistence of posttraumatic stress disorder (PTSD) in prospective
longitudinal studies from 6 months to 3 years later (e.g. Ehlers, Mayou, &
Bryant, 1998; Mayou, Ehlers, & Bryant, 2002). Taken together, the theoretical
overlap between cognitive–behavioural accounts of PTSD and CG (e.g. Ehlers,
2006) and evidence of the link between RT and poor psychological adjustment
are consistent with the hypothesis that RT may be unconstructive following
bereavement.
166  Edward R. Watkins and Michelle L. Moulds

Constructive effects of RT
Only a few studies have demonstrated that RT following bereavement has con-
structive consequences. Nonetheless, we need to be careful that the ‘absence of
evidence’ is not falsely interpreted as ‘evidence of absence’, especially given the
wider evidence that RT can in fact have constructive consequences, including
recovery from traumatic events, and reducing anxiety and depression (Watkins,
2008).
First, in a prospective study that examined the health outcomes of HIV-
seropositive men who had experienced an AIDS-related bereavement, RT was
associated with finding more meaning in the loss over the next 2–3 years. In
turn, this outcome was associated with better immune responses and reduced
AIDs-related mortality over a 7-year follow-up (Bower et al., 1998). Finding
meaning was operationalized as a major shift in values, priorities, or perspectives
in response to the loss, and included developing new personal growth goals, an
enhanced sense of living in the present, and the development of new perspec-
tives (e.g. views such as life is precious). These changes are examples of ‘finding
benefit’, which is defined as positive appraisals about the meaning of the event.
In two prospective longitudinal studies that examined responses to bereavement,
Davis, Nolen-Hoeksema, and Larson (1998) and Stein, Folkman, Trabasso, and
Richards (1997) reported that finding benefit predicted better psychological
adjustment and more adaptive responses to loss. Combined with the results of
Bower et al.’s (1998) study, these findings suggest that RT following bereavement
may be adaptive when it focuses on positive benefits or values learnt as a result
of the loss.
Second, cross-sectional studies have provided evidence that the extent to
which individuals engage in RT after a traumatic or stressful event is positively
associated with more posttraumatic growth, as indexed by self-reported increases
in relating to others, discovering new possibilities and personal strength, and
increased appreciation of life. Tedeschi and Calhoun (2004) found that RT (defined
as automatic or deliberate thinking about the traumatic event) immediately after
a child’s death was associated with posttraumatic growth in bereaved parents,
whereas RT that occurred later after the death was not.
Third, there is extensive evidence from experimental and longitudinal pro-
spective studies that RT can lead to constructive consequences in response to
similar events (e.g. loss, trauma) and on symptom clusters related to CG, such as
depression (see Watkins, 2008). For example, RT prospectively predicted reduced
levels of depression in several longitudinal studies (e.g. Treynor, Gonzalez, &
Nolen-Hoeksema, 2003; Yamada, Nagayama, Tsutiyama, Kitamura, & Furukawa,
2003). Moreover, experimental studies have suggested that certain variants of
RT can have constructive consequences in response to loss and trauma events.
These studies have manipulated whether participants think repetitively in either
an abstract (thinking about the causes, meanings, implications, and ends of an
event and why it occurred, e.g. ‘Why do you feel this way?’) or a concrete way
(imagining the concrete, sensory, and contextual details of what is happening in a
situation and focusing on the process and means of how it occurred, e.g. ‘How do
 Rumination in CG  167
you feel moment by moment?’). Relative to manipulations to engage in abstract
RT, manipulations that instructed participants to engage in concrete RT produced
faster recovery of negative affect and reduced intrusions after a previous nega-
tive induction (Ehring, Szeimies, & Schaffrick, 2009; Watkins, 2004). Similarly,
individuals who were trained to think about emotional events in a concrete way
had reduced emotional reactivity to a subsequent experimental stressor (failure)
relative to those trained to think in an abstract way (Watkins, Moberly, & Moulds,
2008). These results suggest that there are more constructive forms of RT, char-
acterized by a concrete thinking style, in contrast to unconstructive forms of RT,
characterized by an abstract style. Given the broad applicability of theories that
address responses to stressful situations, it is reasonable to assume that similar
constructive consequences of RT apply following bereavement.

Accounting for constructive versus unconstructive effects

of RT

Moderators of the consequences of RT

Reviewing the literature on RT, Watkins (2008) noted that a number of proper-
ties could account for the distinct helpful versus unhelpful consequences of RT:
(a) the valence (positive versus negative) of thought content, with more positive
thought content leading to more constructive consequences (e.g. ‘finding benefit’,
Bower et al., 1998; Segerstrom et al., 2003); (b) the context in which RT occurs
(e.g. intrapersonal context such as mood and self-beliefs, such that more nega-
tive beliefs about the self produce more unconstructive RT; situational context,
such as difficult life events, such that more stressful events produce more uncon-
structive RT); (c) the style of thinking adopted, with abstract RT having more
unconstructive consequences than concrete RT, at least when focused on negative
content. Each of these factors is relevant to responses to bereavement. Bonanno
and Kaltman (1999, 2001) reviewed evidence relevant to the issue of how positive
thought content can be beneficial in the aftermath of grief, and noted that the
context of the bereavement may be important. Moreover, Bonanno et al. (2002)
hypothesized that focusing on concrete aspects of the self may be one means to
maintain identity continuity and, thus, maintain resilience following bereavement.

A control theory account of RT

These moderators are predicted from an elaboration of the control theory account
of RT (Watkins, 2008). Within control theory, representations of events, behav-
iours, or goals can be arranged in a hierarchy of means and ends in which subor-
dinate, concrete goals, or means, serve to achieve the more abstract, superordinate
goals, or ends. A particular level in the hierarchy may be functionally and opera-
tionally superordinate at any moment in time, reflecting whether the individual is
focusing attention and awareness on a more abstract or concrete level, and thereby
representing reference values (goals, expectations) and perceptual signals (e.g.
events and actions in the world) in a more abstract or concrete manner (Carver
168  Edward R. Watkins and Michelle L. Moulds
& Scheier, 1982; Vallacher & Wegner, 1987; Watkins, 2008). Abstract process-
ing involves representing the reasons why an action is performed or an event
occurred, its meaning and implications, and elements that are common across
multiple situations. Concrete processing represents how an event unfolded and
how a behaviour is performed, and reflects contextual and sensory details that are
specific to a particular situation. Thus, following an unexpected death of a loved
one, abstract processing may include existential questions (‘Why me? Why did it
happen to them?’), global negative self-evaluations (‘I am weak’), and negative
generalizations (‘Life is unfair’), whereas processing at a concrete level would
focus on recalling the details of the final days of the deceased, specific memories
of his or her life, and plans for what to do next.
Effective self-regulation requires flexible and balanced coordination between
the different levels within the goal hierarchy, such that the superordinate level of
control adaptively varies in response to situational and task demands. Depending
on context, a level of control that is too abstract or too concrete, or that fails to
link abstract levels to concrete levels, is hypothesized to be detrimental (Carver
& Scheier, 1998).
Within control theory, abstract processing (a) insulates an individual from the
specific context, making him or her less responsive to the environment, (b) gener-
ates inferences across different situations beyond available data, enabling transfer
of learning from one situation to another, and (c) represents events and actions in
terms of their implications for more abstract goals (i.e. more personally meaning-
ful ends). When there is extensive, relevant procedural knowledge that specifies
the links between the abstract ends and the concrete means necessary to achieve
them (e.g. when an activity or situation is familiar), these effects make abstract
processing more adaptive, because they make the individual less distractible and
impulsive, enable consistency and stability of goal pursuit across time, and afford
useful generalizations about the world, along with effective action. However,
abstract processing will be unconstructive when there is no extensive procedural
knowledge, such as during unexpected, difficult, or stressful circumstances (e.g.
the loss of a loved one). In such situations, abstract processing makes the indi-
vidual less responsive to the environment and situational change, and provides
fewer specific guides to action and problem-solving. In addition, because abstract
representations are linked to personal identity, goals represented at an abstract
level are harder to relinquish when progress is difficult, increasing the risk of
being trapped in chronic, unhelpful RT. Under these circumstances, concrete
processing is hypothesized to be constructive. Consistent with this, experimental
manipulations of processing style have demonstrated that concrete processing
facilitates more effective problem-solving than the abstract style (Watkins &
Baracaia, 2002; Watkins & Moulds, 2005).

RT and autobiographical memory

Moreover, abstract RT results in the recall of other related events, such that the
individual is no longer focused on a discrete event but rather on a number of
related difficult events or a category of similar events (Watkins & Teasdale, 2001).
 Rumination in CG  169
Theoretical accounts suggest that it will be easier to organize and make one single
event coherent rather than to process multiple events simultaneously. Specifically,
multiple memories interfere with the processing of one another and consume
more central executive resources (Foa & Kozak, 1986) and, further, include more
disparate material that does not easily fit into the sequence that is necessary to
create a coherent story; coherent story making is hypothesized to be essential to
effectively working through upsetting events (Pennebaker & Seagal, 1999; Smyth,
True, & Souto, 2001). Thus, it is hypothesized that abstract thinking would make
it difficult to emotionally process upsetting events in an effective way, consistent
with experimental findings (e.g. Ehring et al. 2009; Watkins, 2004). This process
is especially pertinent to CG because it could result in poor integration of the
loss with existing autobiographical knowledge about the self and one’s relation-
ship with the deceased person, which is hypothesized to contribute to difficulties
in acknowledging the loss, as well as associated searching behaviours that are
characteristic of CG (Boelen et al., 2006b; Ehlers, 2006).

RT as avoidance
Abstract RT is also hypothesized to have maladaptive effects because it is concep-
tualized as a form of avoidance. Rumination is conceptualized as an escape and
avoidance behaviour that has been negatively reinforced by the removal of aver-
sive experience or because it has perceived or actual functions (Martell, Addis,
& Jacobson, 2001; Nolen-Hoeksema, Wisco, & Lyubomirsky, 2008; Watkins et
al., 2007). Hypothesized and clinically observed functions of rumination include
(a) avoiding the risk of failure/humiliation by thinking about rather than imple-
menting behaviour, (b) attempting to problem solve or to understand current
problems but without a concrete plan of action, (c) avoiding and minimizing
criticism by anticipating potential negative responses from others, (d) controlling
unwanted feelings, and (e) avoiding unwanted attributes by motivating oneself
(e.g. ‘keeping me on my toes’). These functions parallel those that have been
hypothesized to maintain and reinforce pathological worry (Borkovec & Roemer,
1995). Following bereavement, similar avoidant functions of RT might include
(a) attempts to understand unexpected loss, (b) focusing on the anger felt at the
deceased to minimize feelings of guilt (or vice versa), and (c) concerns that not
thinking about the deceased indicates a lack of respect and love for them, or might
lead to forgetting them, such that RT serves the function of avoiding being an
‘uncaring’ person. This functional analytic approach to RT suggests that an idi-
osyncratic assessment and treatment plan is required for each individual patient.
Critically, it has been proposed that abstract RT (e.g. worry), by distancing an
individual from specific details and increasing verbal–conceptual thinking at the
expense of emotionally vivid imagery, may avoid intense affect and/or reduce
physiological arousal (Borkovec, Ray, & Stöber, 1998; Stöber & Borkovec,
2002). Likewise, individuals with PTSD engage in RT about their trauma and its
sequelae but do so in an abstract, vague way (e.g. ‘why did this happen to me?’).
Such thinking about the causes and consequences of a trauma avoids direct reliv-
ing of the event and re-experiencing of the distress associated with recall, which
170  Edward R. Watkins and Michelle L. Moulds
in turn prevents the successful emotional processing that is critical to recovery
(Foa & Kozak, 1986). Consistent with an avoidant conceptualization of rumina-
tion, in the context of depression, there is evidence that rumination is associated
with measures of avoidance (Cribb, Moulds, & Carter, 2006; Moulds, Kandris,
Starr, & Wong, 2007).
Recent conceptualisations have argued that RT following bereavement also
serves to avoid recalling painful memories and experiencing the emotions associ-
ated with them (Boelen et al., 2006a; M. Stroebe et al., 2007). As noted above,
focusing on the reasons and implications of the loss would take an individual
away from direct contact with specific memories of the event, which may be nega-
tively reinforced in the short term by reducing negative affect but in the longer
term would prevent effective habituation to such memories. Despite the concep-
tualization of RT as avoidance in the context of grief, there is currently limited
empirical support for this hypothesis. Boelen et al. (2006a) found that items that
indexed RT (e.g. ‘I keep on pondering about who is to blame for the loss’) were
correlated with items that assessed behavioural avoidance in patients with CG.
However, we need to be cautious when interpreting correlations between RT and
avoidance because we cannot determine that rumination functions as avoidance;
the correlation could reflect a common factor such as a passive coping style, or
even that increased avoidance leads to more RT. Finally, the findings of Bonanno
et al. (2005) are of note here. Consistent with avoidance being problematic in
resolving grief but inconsistent with RT functioning as avoidance, Bonanno et
al. (2005) reported that grief processing (including assessment of RT) and grief
avoidance were uncorrelated but that each process independently prospectively
predicted distress.

Summary of hypotheses and relevant evidence

Thus, for both theoretical and empirical reasons, we hypothesize that there are
specific and distinct subtypes of RT within grief, with distinct functional conse-
quences. In particular, we hypothesize that concrete RT during situations of stress,
such as following bereavement, is adaptive because it improves problem solving,
focuses attention on the immediate environment, facilitates abandoning unachiev-
able goals or unrealistic expectations, and aids habituation. We hypothesize that
adaptive variants of RT will be characterized by finding benefit following a loss or
processing grief-related events in a concrete way. We also hypothesize that there is
a more pathological variant of RT characterized by an abstract focus on the mean-
ings or implications of the bereavement, which contributes to CG. Moreover, we
hypothesize that the normal process of grieving that is experienced by the major-
ity of individuals is characterized by the adaptive variant of RT. Consistent with
this hypothesis, there is extensive evidence that individuals tend (by default) to
adopt more abstract processing, yet, when faced with difficult, novel, or complex
situations, move towards more concrete processing (Watkins, 2008). Further, the
lack of evidence that unsystematic grief work, expressive writing, or counselling
is beneficial for the majority of bereaved individuals (W. Stroebe et al., 2005)
 Rumination in CG  171
is consistent with normal grief already involving adaptive responses to the loss.
Because concrete RT facilitates coping and adaptation, and reduces the discrep-
ancy between the loss and the individuals’ existing mental models, we further
hypothesize that it is necessarily self-limiting and thus short-lived.
In contrast, we hypothesize that the pathological forms of grieving (exemplified
by CG) involve maladaptive abstract RT, which (because it prevents the resolu-
tion of the discrepancy between the loss and the individual’s prior mental models)
is self-perpetuating and prolongs distress. Thus, we hypothesize that constructive
and unconstructive forms of RT can be distinguished on the basis of their degree
of abstract versus concrete processing. Preliminary support for this hypothesis
comes from findings that (a) healthy adjustment over the first 2 years following a
spouse’s death was associated with self-evaluations focused on concrete actions
and behaviours rather than on abstract evaluations focused on character traits
(Bauer & Bonanno, 2001), (b) narratives from the first month of bereavement
that focused on concrete goals and plans predicted healthier outcomes 1 year later
(Stein et al., 1997), and (c) bereaved individuals who report not searching for
meaning (i.e. not adopting an abstract style) are more resilient and have better
outcomes (Bonanno et al., 2004; Davis, Wortman, Lehman, & Silver, 2000).
Relevant to this analysis, we note that several of the measures employed to
index RT following bereavement are more consistent with the conceptualization
of abstract RT (e.g. RSQ, Nolen-Hoeksema et al., 1994; Boelen et al., 2006a),
which could account for the asymmetry of findings towards RT having uncon-
structive outcomes. Other measures index only frequency of thoughts about the
deceased (e.g. Bonanno et al., 2005), which cannot discriminate between adaptive
and maladaptive RT, and this indicates the value of employing more sensitive
measures.
A key test of the proposed hypothesis would be to examine whether a measure
of RT that focused on concrete processing of the loss (e.g. recalling the detailed
sensory and emotional aspects without dwelling on its meaning: thinking about
how rather why it happened) predicts better adjustment in a longitudinal study.
Further, this account predicts that those who show more normal grief reactions
would focus on the loss in a more concrete way than those who become stuck in
more protracted grief. The hypothesis could be further tested by adapting previous
experimental studies that have examined the effect of abstract versus concrete
manipulations in response to distressing events to the response to the death of a
loved one (e.g. through an expressive writing design). Similarly, given evidence
that training dysphoric individuals to repeatedly practise thinking in a concrete
way reduces rumination and depression relative to a no-training control condition
(Watkins, Baeyens, & Read, 2009), the current analysis predicts that such a treat-
ment approach may be beneficial for individuals with CG.

Treatment implications
Our hypothesis that concrete processing of memories may promote an adap-
tive grief response accords with recent treatment developments in the CG field.
172  Edward R. Watkins and Michelle L. Moulds
Although treatments for CG have not directly addressed RT about the loss per se,
recently developed exposure-based approaches are relevant to our argument. Two
recent randomized controlled trials indicated that CBT therapies that involved
imaginal exposure – in which patients repeatedly relive the story of their loved
one in order to process the content of grief-related memories – had significantly
better outcomes for patients with CG than interpersonal psychotherapy (Shear,
Frank, Houck, & Reynolds, 2005) or supportive counselling (Boelen, de Keijser,
van den Hout, & van den Bout, 2007). Given that imaginal exposure involves the
repeated and direct step-by-step reliving of the detail of distressing memories and
their associated affect (rather than RT about why the distressing event occurred
and what it means), these outcomes are consistent with the hypothesis that con-
crete RT may facilitate adjustment after the loss of a loved one (although we note
that reliving was not the only element in these treatment packages).
Our hypothesis suggests that directly training individuals with CG to be more
concrete (Watkins et al., 2009) or shift to more adaptive forms of RT (Watkins et
al., 2007) could be effective treatments in CG in their own right or as adjuncts to
imaginal exposure. A trial of concrete training versus a control condition would
provide a test of proof-of-principle for the causal role of processing style on the
maintenance of grief-related symptoms.

Implications for CG
Our review has focused on evidence regarding the role of RT in grief, given the
lack of direct evidence in CG. Nonetheless, the evidence reviewed and hypoth-
eses proposed have a number of implications for CG. First, the wider literature on
RT suggests the hypothesis that abstract RT will be involved in the development
and maintenance of CG, and that abstract RT about the bereavement may con-
tribute to an ongoing preoccupation with the deceased. As noted earlier, abstract
processing leads to RT that is less constructive, more prolonged, and harder to
abandon. Second, abstract RT soon after the bereavement, relative to more con-
crete RT, may indicate an individual at risk for developing CG. Third, treatment
interventions that explicitly and directly target abstract RT (rumination), whether
through direct concreteness training, structured exposure, or problem solving, are
hypothesized to be more effective at treating CG.

References
Bauer, J., & Bonanno, G. A. (2001). Being and doing well (for the most part): Adaptive
patterns of narrative self-evaluation during bereavement. Journal of Personality, 69,
451–482.
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2006a) Negative cognitions and
avoidance in emotional problems after bereavement: A prospective study. Behaviour
Research and Therapy, 44, 1657–1672.
Boelen P. A., van den Hout, M. A., & van den Bout, J. (2006b). A cognitive–behavioral
conceptualization of complicated grief. Clinical Psychology: Science and Practice, 13,
109–128.
 Rumination in CG  173
Boelen, P. A., de Keijser, J., van den Hout, M. A., & van den Bout, J. (2007). Treatment of
complicated grief: A comparison between cognitive–behavioral therapy and supportive
counselling. Journal of Consulting and Clinical Psychology, 75, 277–284.
Bonanno, G. A., & Kaltman, S. (1999). Toward an integrative perspective on bereavement.
Psychological Bulletin, 125, 760–776.
Bonanno, G.  A., & Kaltman, S. (2001). The varieties of grief experience. Clinical
Psychology Review, 21, 705–734.
Bonanno, G., Papa, A., & O’Neill, K. (2002). Loss and human resilience. Applied and
Preventative Psychology, 10, 193–206.
Bonanno, G., Papa, A., Lalande, L., Zhang, N., & Noll, J. (2005). Grief processing and
deliberate grief avoidance: A prospective comparison of bereaved spouses and parents
in the United States and the People’s Republic of China. Journal of Consulting and
Clinical Psychology, 73, 86–98.
Bonanno, G. A., Wortman, C. B., & Neese, R. M. (2004). Prospective patterns of resilience
and maladjustment during widowhood. Psychology and Aging, 19, 260–271.
Borkovec, T. D., Ray, W. J., & Stöber, J. (1998). Worry: A cognitive phenomenon intimately
linked to affective, physiological, and interpersonal behavioral processes. Cognitive
Therapy and Research, 22, 561–576.
Borkovec, T. D. & Roemer L. (1995) Perceived functions of worry among generalized anx-
iety disorder subjects: distraction from more emotionally distressing topics? Journal of
Behaviour Therapy and Experimental Psychiatry, 26, 25–30.
Bower, J. E., Kemeny, M. E., Taylor, S. E., & Fahey, J. L. (1998). Cognitive processing,
discovery of meaning, CD4 decline, and AIDS-related mortality among bereaved HIV-
seropositive men. Journal of Consulting and Clinical Psychology, 66, 979–986.
Bowlby, J. (1980) Attachment and loss, vol. 3: Loss: Sadness and depression. London:
Hogarth Press.
Carver, C.  S., & Scheier, M.  F. (1982). Control-theory: A useful conceptual-framework
for personality-social, clinical, and health psychology. Psychological Bulletin, 92,
111–135.
Carver, C.  S., & Scheier, M.  F. (1998). On the self-regulation of behavior. Cambridge:
Cambridge University Press.
Cribb, G., Moulds, M. L., & Carter, S. (2006). Rumination and experiential avoidance in
depression. Behaviour Change, 23, 165–176.
Davis, C. G., Lehman, D. R., Wortman, C. B., Silver, R. C., & Thompson, S. C. (1995).
The undoing of traumatic life events. Personality and Social Psychology Bulletin, 21,
109–124.
Davis, C. G., Nolen-Hoeksema, S., & Larson, J. (1998). Making sense of loss and benefit-
ing from the experience: Two construals of meaning. Journal of Personality and Social
Psychology, 75, 561–574.
Davis, C. G., Wortman, C. B., Lehman, D. R., & Silver, R. C. (2000). Searching for mean-
ing in loss: Are clinical assumptions correct? Death Studies, 24, 497–540.
Ehlers, A. (2006). Understanding and treating complicated grief: What can we learn from
post-traumatic stress disorder. Clinical Psychology Science and Practice, 13, 135–140.
Ehlers, A., Mayou, R. A., & Bryant, B. (1998). Psychological predictors of chronic posttrau-
matic stress disorder after motor vehicle accidents. Journal of Abnormal Psychology,
107, 508–519.
Ehring, T., Szeimies, A.-K., & Schaffrick, C. (2009). An experimental analogue study into
the role of abstract thinking in trauma-related rumination. Behaviour Research and
Therapy, 47, 284–293.
174  Edward R. Watkins and Michelle L. Moulds
Foa, E., & Kozak, M.  J. (1986). Emotional processing of fear: Exposure to corrective
information. Psychological Bulletin, 99, 20–35.
Greenberg, M. A. (1995). Cognitive processing of traumas: The role of intrusive thoughts
and reappraisals. Journal of Applied Social Psychology, 25, 1262–1296.
Ito, T., Tomita, T., Hasui, C., Otsuka, A., Katayama, Y.,, Kawamura, Y., et al. (2003). The
link between response styles and major depression and anxiety disorders after child
loss. Comprehensive Psychiatry, 44, 396–403.
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma.
New York: Free Press.
Lepore, S. J., Silver, R. C., Wortman, C. B., & Wayment, H. A. (1996). Social constraints,
intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of
Personality and Social Psychology, 70, 271–282.
Lichtenthal, W. G., Cruess, D. G., & Prigerson, H. G. (2004). A case for establishing com-
plicated grief as a distinct mental disorder in DSM-V. Clinical Psychology Review, 24,
637–662.
Martell, C. R., Addis, M. E., & Jacobson, N. S. (2001). Depression in context: Strategies
for guided action. New York: Norton.
Martin, L.  L., & Tesser, A. (1996). Some ruminative thoughts. In Wyer, R.  S. (Ed.),
Advances in social cognition, Vol. 9: Ruminative thoughts (pp.  1–47). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Mayou, R. A., Ehlers, A., & Bryant, B. (2002). Posttraumatic stress disorder after motor
vehicle accidents: 3-year follow-up of a prospective longitudinal study. Behaviour
Research and Therapy, 40, 665–675.
Michael, S., & Snyder C. (2005) Getting unstuck: The roles of hope, finding meaning, and
rumination in the adjustment to bereavement among college students. Death Studies,
29, 435–458.
Moulds, M. L., Kandris, E., Starr, S., & Wong, A. C. M. (2007). The relationship between
rumination, avoidance and depression in a non-clinical sample. Behaviour Research
and Therapy, 45, 251–261.
Nolen-Hoeksema, S. (2000). The role of rumination in depressive disorders and mixed
anxiety/depressive symptoms. Journal of Abnormal Psychology, 109, 504–511.
Nolen-Hoeksema, S., McBride, A., & Larson, J. (1997). Rumination and psychological
distress amongst bereaved partners. Journal of Personality and Social Psychology, 72,
855–862.
Nolen-Hoeksema, S., Parker, L. E., & Larson, J. (1994). Ruminative coping with depressed
mood following loss. Journal of Personality and Social Psychology, 67, 92–104.
Nolen-Hoeksema, S., Wisco, B.  E., & Lyubomirsky, S. (2008). Rethinking rumination.
Perspectives on Psychological Science, 3, 400–424.
Pennebaker, J. W., & Seagal, J. D. (1999). Forming a story: The health benefits of narrative.
Journal of Clinical Psychology, 55, 1243–1254.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Med 6(8): e1000121.
Segerstrom, S. C., Stanton, A. L., Alden, L. E., & Shortridge, B. E. (2003). A multidimen-
sional structure for repetitive thought: What’s on your mind, and how, and how much?
Journal of Personality and Social Psychology, 85, 909–921.
Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. (2005). Treatment of complicated
grief: A randomized controlled trial. Journal of the American Medical Association, 293,
2601–2608.
 Rumination in CG  175
Smyth, J., True, N., & Souto, J. (2001). Effects of writing about traumatic experiences:
The necessity for narrative structuring. Journal of Social and Clinical Psychology, 20,
161–172.
Stein, N., Folkman, S., Trabasso, T., & Richards, T. A. (1997). Appraisal and goal pro-
cesses as predictors of psychological well-being in bereaved caregivers. Journal of
Personality and Social Psychology, 72, 872–884.
Stöber, J., & Borkovec, T. D. (2002). Reduced concreteness of worry in generalized anxiety
disorder: Findings from a therapy study. Cognitive Therapy and Research, 26, 89–96.
Stroebe, M., Boelen, P. A., van den Hout, M., Stroebe, W., Salemink, E., & van den Bout,
J. (2007). Ruminative coping as avoidance: a reinterpretation of its function in adjust-
ment to bereavement. European Archives of Psychiatry and Clinical Neuroscience,
257, 462–472.
Stroebe, M., & Schut, H. (1999). The dual-process model of coping with bereavement:
Rationale and description. Death Studies, 23, 174–184.
Stroebe, W., Schut, H., & Stroebe M. (2005) Grief work, disclosure and counselling: Do
they help the bereaved? Clinical Psychology Review, 25, 395–314.
Tedeschi, R. G. & Calhoun, L. G. (2004). Posttraumatic growth: Conceptual foundations
and empirical evidence. Psychological Inquiry, 15, 1–18.
Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A
psychometric analysis. Cognitive Therapy and Research, 27, 247–259.
Vallacher, R. R., & Wegner, D. M. (1987). What do people think they’re doing? Action
identification and human behavior. Psychological Review, 94, 3–15.
Watkins, E. (2004). Adaptive and maladaptive ruminative self-focus during emotional pro-
cessing. Behaviour Research and Therapy, 42, 1037–1052.
Watkins, E. R. (2008). Constructive and unconstructive repetitive thought. Psychological
Bulletin, 134, 163–206.
Watkins, E., & Baracaia, S. (2002). Rumination and social problem-solving in depression.
Behaviour Research and Therapy, 40, 1179–1189.
Watkins, E. R., Baeyens, C. B., & Read, R. (2009). Concreteness training reduces dyspho-
ria: Proof-of-principle for repeated cognitive bias modification in depression. Journal
of Abnormal Psychology, 118, 55–64.
Watkins, E R.., Moberly, N. J., & Moulds, M. L. (2008). Processing mode causally influ-
ences emotional reactivity: Distinct effects of abstract versus concrete construal on
emotional response. Emotion, 8, 364–378.
Watkins, E., & Moulds, M. (2005). Distinct modes of ruminative self-focus: Impact of
abstract versus concrete rumination on problem solving in depression. Emotion, 5,
319–328.
Watkins, E., Scott, J., Wingrove, J., Rimes, K., Bathurst, N., Steiner, H., et al. (2007).
Rumination-focused cognitive–behaviour therapy for residual depression: A case
series. Behaviour Research and Therapy, 45, 2144–2154.
Watkins, E., & Teasdale, J.  D. (2001). Rumination and overgeneral memory in depres-
sion: Effects of self-focus and analytic thinking. Journal of Abnormal Psychology, 110,
353–357.
Yamada, K., Nagayama, H., Tsutiyama, K., Kitamura, T., & Furukawa, T. (2003). Coping
behavior in depressed patients: A longitudinal study. Psychiatry Research, 121,
169–177.
13 Autobiographical memory
processes in complicated grief
Ann-Marie J. Golden

What is autobiographical memory (AM) and overgeneral

memory (OGM) bias and its importance?
When we talk of memory we all know what we mean: we lay down memories of
our everyday experiences and then later on we attempt to recall instances from
this databank. According to Baddeley (1997) the memory system consists of three
closely interlinked stages of processing: first encoding, an initial and highly selec-
tive processing of new information; second storage, a limited capacity retention
of encoded information over time; and third retrieval, the attempts to access the
encoded/stored information. Memory is usually thought to consist of two ‘stores’:
semantic (i.e. general factual information) and episodic (i.e. information about
past personal life experiences). In this chapter we are only interested in the stor-
age and retrieval of autobiographical ‘episodic’ memories. These are essential
for self-descriptions, emotions, and the phenomenology of an individual across
time and life-time experience, which contributes to an individual’s sense of self
(Conway & Pleydell-Pearce, 2000) and inevitably carries related physical, social,
and emotional stamps.
So why is it of any value to understand autobiographical memory? Because
it is related to our major goals in life, is embroidered with powerful affective
states, and impacts our personal meanings. Bluck, Alea, Haberman, and Rubin
(2005) suggested that autobiographical memory serves three broad functions: a
directive function, because we use our past experiences as a reference for dis-
entangling current problems and as a guide for confronting new challenges; a
social function, because memories help us with development, maintenance, and
nurturing of social relationships by providing material for social interactions; and
a self-representative function, because personal memories help us to generate and
sustain a self-identity that is coherent over a period of time. An additional, adap-
tive function is that the retrieval of positive personal experiences may be used as
a maintenance tool of desirable moods or to modify unwanted moods (Williams,
Conway, & Cohen, 2008).
In this domain of memory and affect, retrieval of personal memories using
cue words has been investigated for many decades (Teasdale & Fogarty, 1979),
resulting in the robust conclusion that mood-related memory biases (whether
natural or induced) can contribute to and/or maintain affective disorder (e.g. the
 Autobiographical memory processes and CG  177
more depressed an individual, the slower his or her latency, and the greater dif-
ficulty he or she has, in the recall of positive memories). Williams and Broadbent
(1986) also noted that when they asked depressed and suicidal individuals to
retrieve personal memories, to a series of emotion-related cue words as part of
the Autobiographical Memory Test (AMT), they tended to be general rather than
specific memories (in comparison with normal controls). This finding shifted
the research to include the degree of specificity, not just the speed or content of
retrieval with which events are recollected.
Since this prototypical study, and over the last two decades, research has
shown that individuals suffering from a range of clinical conditions such as
depression (Kuyken & Dalgleish, 1995) and posttraumatic stress disorder (PTSD)
(McNally et al., 1995) generally have relative difficulties retrieving specific auto-
biographical memories. This has been called an overgeneral memory (OGM) bias
(for reviews see Moore & Zoellner, 2007; Sumner, Griffith, & Mineka, 2010;
Williams et al., 2007) – that is, the tendency to recollect the past in terms of regu-
larities across multiple experiences as opposed to specific events. OGM is also a
characteristic of complicated grief (CG; Golden, Dalgleish & Mackintosh, 2007):
a debilitating condition arising when, following the death of a significant other, an
individual presents with a range of symptoms that cause significant impairment
in day-to-day functioning for 6 months or more, including yearning, pining, guilt,
bitterness or anger, intrusive distress, preoccupation with and thoughts relating to
the deceased, and a difficulty in moving on (e.g. Prigerson & Maciejewski, 2005).
This chapter focuses on the mechanisms involved in OGM bias and studies that
address autobiographical memory and CG.
Why is this bias towards overgeneralizing memories so important to CG and
why does it warrant further attention? So far we know that OGM is a trait-like
characteristic that may serve as a vulnerability factor for depression (Williams
et al., 2007) and plays a role in the maintenance of depression (Gibbs & Rude,
2004; Williams et al., 2006) and range of disorders (and as such OGM is an excel-
lent example of a transdiagnostic process; Harvey, Watkins, Mansell, & Shafran,
2004). In addition to this, research has shown that individuals who exhibit this
OGM bias:

1 have impaired problem solving and increased cognitive reactivity to changes

in mood (i.e. easier reactivation of patterns of negative thinking) (Williams,
Barnhofer, Crane, & Beck, 2005);
2 exhibit hopelessness and a diminished ability to imagine future events
(Dickson & Bates, 2006);
3 have ruminative tendencies (Ramponi, Barnard, & Nimmo-Smith, 2004);
4 are at risk of developing symptomatology following stressful life events
(Bryant, Sutherland, & Guthrie, 2007).

Therefore an impaired ability to retrieve specific memories from one’s past may
impede integration of an existing situation and thus delay recovery (see Ehlers &
Clark, 2000; Williams et al., 2006).
178  Ann-Marie J. Golden

The methodology and limitations of the Autobiographical

Memory Test (AMT)
The OGM bias is usually assessed using the AMT whereby cue words that vary
in emotional valence (e.g. negative, positive, and neutral) are presented and par-
ticipants are instructed to retrieve an event that the given word reminds them of.
The event could be important or trivial, recent or from a long time ago, but it
should be a specific event (i.e. something that happened at a particular time on a
particular day and lasted for a day or less). Participants are typically allowed 30
seconds to 1 minute for each cue word for retrieval of a specific memory. They
are given an example of what is meant by specific (to the word enjoy, it would be
inappropriate to say ‘I always enjoy a good party’ because that does not mention a
particular time, but it would be fine to say ‘Jane’s party last Friday’), followed by
three practice trials to ensure that they have understood the instructions.
Memories generated are coded according to the criteria laid down by Williams
and Dritschel (1991). Non-specific memories include extended memories (events
that lasted for longer periods of time) and categoric memories (events that
occurred repeatedly over a period of time). Failure to recall, or giving opinions
about the cue instead of a memory, is classed as ‘no memory’.
No methodology comes without limitations and the same can be said for the
AMT. Williams et al. (2007) reviewed these limitations in detail and emphasized
that there are five methodological constraints that may limit the conclusions drawn
from AMT studies: the number of words used as cues; general intelligence; con-
founding of depression and trauma history; retrospective assessment of trauma;
and scarce longitudinal studies.

The psychological mechanisms in OGM bias in relation to

CG research
Several mechanisms have been proposed to explain OGM bias, such as truncated
search patterns in clinical groups (Williams & Dritschel, 1988); hierarchical auto-
biographical memory structures (Conway, 1996) with specific (and potentially
aversive) memories at the lowest level; and poor executive control (Dalgleish et al.,
2007). These have been integrated within the CaRFAX (capture and rumination,
functional avoidance, and executive control) model (Williams, 2006). In terms of
CG the most relevant processes are the three factors (rumination, functional avoid-
ance, and executive control) of the CaRFAX model; the more recent immunity
hypothesis (Golden et al., 2007); emerging grief theories (discussed elsewhere in
detail: Boelen, van den Hout, & van den Bout, 2006; Shear & Shair, 2005); and the
importance of the self in autobiographical memories (Conway & Pleydell-Pearce,
2000). These will be the subject of further discussion in this chapter.

Rumination and OGM bias

Rumination consists of the repetitive and intrusive revisiting of the causes, con-
sequences, and symptoms of one’s negative affective state (Nolen-Hoeksema,
 Autobiographical memory processes and CG  179
1991). There is persuasive evidence for a close link between rumination and
OGM bias. Watkins and Teasdale (2001) specifically identified that it is the
analytical aspect of self-focused rumination that is associated with OGM bias.
Dalgleish et al. (2003) further hypothesized that self-analytical processing may
be activated by negative self-schemata. Grief researchers have emphasized that
negative cognitions and avoidance, that is, rumination, play an important role in
the development and maintenance of CG (Boelen et al., 2006) and also mediate
the effect of risk factors in bereaved individuals (van der Houwen et al., 2010).

Functional avoidance and OGM bias

According to Williams, Stiles, and Shapiro (1999), trauma-exposed individuals
develop a reduced capability to access specific memories as a form of ‘functional
avoidance’ (Williams et al., 2007) of the distress associated with remembering
the specific details of their traumatic experiences. They suggest that this avoid-
ance operates through the truncation of an effortful, hierarchical search of the
self-memory system (SMS), so-called generative retrieval (Conway & Pleydell-
Pearce, 2000), at the level of categorical autobiographical descriptors, which
gives rise to reduced recall of specific memories. Instances of trauma memory
recollection generalize to the whole domain of autobiographical recollection,
resulting in the classic reduced specificity effect (OGM bias) on the AMT.
Therefore individuals who have been involved in traumas show this classic OGM
bias on the AMT.
Kuyken and Brewin (1995) showed a significant association between OGM
bias and trauma and confirmed that OGM bias not only is related to a clinical
condition such as depression but is also a function of people’s exposure to trauma
and to previous related distress. This association between a history of trauma and
OGM bias has now been replicated several times and following traumatic experi-
ences other than abuse (e.g. Dalgleish et al., 2003). These findings suggest that
OGM bias in trauma-exposed samples plays a key role in the onset and main-
tenance of posttraumatic stress and it is not simply a cognitive epiphenomenon
associated with the acute clinical state. In terms of CG this is further examined in
the ‘immunity hypothesis’ (Golden et al., 2007), which is discussed later.

Executive control and OGM bias

Autobiographical knowledge is thought to be mediated through executive con-
trol processes (Baddeley, 1986; Conway & Pleydell-Pearce, 2000). Executive
resources are assumed to be required for keeping focused on task, inhibiting
incorrect responses, and searching downwards through the hypothetical hierar-
chical database (e.g. Dalgleish et al., 2007; Dalgleish, Rolfe, Golden, Dunn, &
Barnard, 2008). Considering that memory retrieval in its earlier stages is effortful
and uses executive capacity, deficits in executive processing resources are also
thought to be involved in OGM bias (Dalgleish et al., 2007). Williams (1996)
proposed that OGM bias results when working memory is compromised during
‘mnemonic interlock’ (when the search process is cut short at an upper, general
180  Ann-Marie J. Golden
level). Some experimental studies have demonstrated that cognitive deficits and
reduction in working memory capacity are linked with difficulties retrieving spe-
cific memories (e.g. Birch & Davidson, 2007). A number of studies have indicated
that difficulties in specific retrieval may arise as a result of depletion of cognitive
capacity caused by intrusive cognitions that are associated with clinical disorders
(Kuyken & Brewin, 1995). Research in CG awaits studies looking at executive
resources and OGM bias.

Immunity and OGM bias

Experiencing intrusive recollections about traumatic events is a key feature of
clinical states such as acute stress disorder, adjustment disorder, and posttraumatic
stress disorder (American Psychiatric Association, 1994). Distressing intrusions
also play a part in CG; however, in CG the dominant emotion is sadness, pining,
and yearning, whereas in PTSD it is fear and horror (Dalgleish & Power, 2004).
There is an interesting paradox here since trauma victims, although exhibiting
an OGM bias on the AMT trauma, report intrusions full of detailed and specific
memories about their trauma (Brewin, Dalgleish, & Joseph, 1996; McNally,
2003). If OGM bias is a form of affect regulation then it seems to be acting some-
what inefficiently on individuals’ day-to-day intrusive memories of the trauma, as
these remain highly specific.
This raises the question whether memories associated with the source of the
person’s distress are somehow not susceptible, that is, are immune and resist-
ant to the OGM bias (hence ‘immunity hypothesis’). One explanation for this
paradoxical ‘immunity’ may be that, because these ‘recurrent intrusive’ memories
are explicitly concerned with the source of person’s distress, they are sufficiently
prepotent to be directly accessed in the autobiographical database (Conway &
Pleydell-Pearce, 2000) in a way that does not require the effortful, hierarchical
search of the autobiographical memory system. Perhaps this search, or ‘genera-
tive retrieval’, is circumvented and memories related to the loss and bereavement
(or trauma) are being retrieved by a very direct route.
Golden et al. (2007) hypothesized that if, because of their intrusive and recur-
rent nature, memories related to the source of an individual’s distress are habitu-
ally directly retrieved as a function of their pre-priming/prepotency, then it should
be possible to demonstrate this using the cue word AMT task. This was important
in order to elucidate whether OGM bias as a form of affect regulation is indeed
impotent in this way with regard to memories tied to the source of an individual’s
distress, indicating that reducing specificity is a somewhat ineffective method of
psychological defence  – less a form of ‘functional avoidance’ (Williams et al.,
2007, p. 122) than a form of dysfunctional avoidance.
Golden et al. (2007) explained that, in terms of PTSD, one could constrain
the search parameters of the AMT and ask participants to respond only with
trauma-related memories, that is, use a trauma-only AMT (see Figure 13.1). If
trauma memories were directly retrieved, this constraint would have produced
significant attenuation or elimination of the usual OGM bias associated with the
 Autobiographical memory processes and CG  181

POST TRAUMATIC STRESS DISORDER (PTSD)

AMT  -­‐LIFETIME AMT  – TRAUMA  ONLY

OGM AMT
bias specificity

COMPLICATED GRIEF (CG)

AMT  – OWN  LIFE BMT  – ANOTHER’S  LIFE

OGM BMT-­‐ BMT-­‐

bias DECEASED   LIVING  
PERSON PERSON

AMT OGM
specificity bias

Figure 13.1 OGM bias in PTSD and CG: the comparisons between Autobiographical
Memory Test (AMT) and Biographical Memory Test (BMT).

task. However, there is a clear methodological confound associated with using

a trauma-only AMT. If the search parameters of the AMT were restricted to a
narrowly constrained set of experiences such as traumas only, then individuals
would have fewer opportunities for any generative, hierarchical search process
to become truncated. First, such a search would require less navigation of the
memory system and, second, establishing such a search set would additionally
prime a number of specific events (i.e. the relevant trauma experiences). This
would have inevitably lead to high levels of specificity in terms of the memories
recalled relative to an unrestricted AMT.
The standard and trauma-only AMT are, therefore, not comparable like-for-
like, and CG provided a really good vehicle to investigate this interesting question
about OGM bias. In their day-to-day life, individuals with CG experience intrusive
memories as part of their diagnostic symptomatology. These intrusive memories
are linked to the aetiology of the individual’s distress (i.e. memories about the
deceased). As such these memories about the deceased are sourced from a rich
database of specific events: the lifetime of the deceased person, as opposed to
one or more discrete traumas (Dalgleish & Power, 2004). Intrusive memories are
experienced as involuntary and highly affect-laden (Boelen et al., 2006; Dalgleish
& Power, 2004). In individuals with CG, this characteristic provided an excellent
opportunity to examine whether recollection targeted at the source of a person’s
distress becomes markedly less or more specific when the relevant memories are
prompted using the cue word methodology of the AMT.
As shown in Figure 13.1, individuals with CG retrieved memories about their
own life (standard AMT) as well as memories from the life of the deceased person
(called BMT-Deceased). It is possible that differences between these conditions
 182  Ann-Marie J. Golden
could simply be a result of comparison between an autobiographical and bio-
graphical version of the test. To control for this, individuals with CG also retrieved
memories from the lifetime of someone who had not died: someone whom they
knew as well and as long as the deceased (called BMT-Living). It was expected
that the standard OGM bias would be present since, according to Williams et al.
(2007), the overgenerality mode develops as a way of avoiding distress and it then
generalizes to other conditions. The introduction of BMT-Deceased and BMT-
Living (Golden et al., 2007) was the first time that the cue word methodology has
been extended to interrogate memories from the lifetime of someone other than
the self, and it reflected a growing trend to examine memory specificity outside
the strict boundaries of the standard AMT (e.g. Rottenberg, Hildner, & Gotlib,
2006).
In Figure 13.2 the findings of the study by Golden et al. (2007) show that for
both positive and negative cues the standard AMT condition was as predicted:
participants with CG were less specific than the controls (i.e. showing OGM bias).
This finding was reversed on the BMT-Deceased: the CG group retrieved far more
specific memories about the life of the person who died than did the bereaved con-
trols. It seems that because these memories are about the source of distress, and
are somehow immune, individuals are actually spared from the OGM bias. On the
BMT-Living there was the same pattern of OGM bias as found on the standard
AMT. Therefore this finding was not simply a function of BMT-Deceased being
about somebody else. These interactions and relevant paired comparisons were
statistically significant for the negative cues only. The pattern for positive cues
was similar but the effects were not significant.

0.95 AM
Specificity
Proportions of Specific Memories

0.9

0.85

Control Group
0.8
CG Group

0.75

0.7

0.65

0.6
AM BMT- BMT-Living AM BMT- BMT-Living
T Deceased T Deceased
Positive Cues Negative Cues

Figure 13.2 Proportions of specific memories retrieved to positive and negative cues

across the two groups (error bars are + 1 standard error). (Adapted from
Golden, Dalgleish, & Mackintosh, 2007.)
 Autobiographical memory processes and CG  183
A number of other studies have looked at CG and OGM. Boelen, Huntjens, van
Deursen, and van den Hout (2010) investigated the specificity of autobiographical
memories and symptomatology of CG, depression, and PTSD in 109 bereaved
participants with no formal diagnosis of CG using standard AMT and trait AMT
(McNally et al., 1995) (i.e. participants are presented with five positive and five
negative trait adjectives and asked to recall a specific memory of when they
displayed each trait). An important difference from previous studies is that the
AMT tasks were conducted over the telephone and participants were given only
30 seconds to produce a specific memory. Their findings confirmed that the pres-
ence of OGM bias was significantly associated with symptom levels of CG but
not with depression and PTSD. They also identified that only the symptom levels
of CG and PTSD were associated with preferential retrieval of specific memories
that were related to the loss or lost person on the standard AMT, whereas all three
symptom measures were associated with preferential retrieval of loss-related spe-
cific memories on the trait AMT. These findings on a trait AMT are in accord with
Golden et al.’s (2007) immunity hypothesis. Maccallum and Bryant (2008) and
Boelen et al. (2010) showed that memories retrieved in the trait AMT were associ-
ated more with loss-related distress (these memories included the deceased person
and were part of the self-relevant autobiographical information) than memories
that were retrieved in the standard AMT (these memories included the lost person
within more general autobiographical knowledge). These additional studies have
further highlighted the importance of OGM bias in CG.

CG research studies and self and OGM bias

Whereas the CaRFAX model primarily focused on the quality of autobiographi-
cal remembering, Conway and Pleydell-Pearce (2000) proposed the self memory
system (SMS), which focuses more on the content of remembering. This places
an individual’s sense of self at the core of memory construction and retrieval.
Autobiographical memories are produced to serve the goals of the self; in other
words ‘the working self’ is a set of active goals that reflect the concerns of the
dominant self concept. So, following the death of a loved person, self-identity
is somewhat reconstructed and new life goals need to be found (e.g. Neimeyer,
1998). The degree to which a bereaved individual is able to reconstruct him- or
herself in the absence of the loved one, and to find a new purpose and meaning
in life, is assumed to be connected with recovery (Neimeyer, Baldwin, & Gillies,
2006). The SMS in relation to CG would predict that, given that there are strong
feelings of preoccupation and persistent yearning, the individual’s self-identity
would be related to the deceased, so the death of a loved person would be a self-
defining event. Recently Boelen (2009) showed that individuals who exhibited a
more severe form of CG also experienced loss as more central to their everyday
inferences, their self-identity, and their perception of their own life story.
The loss of a loved one can be a challenge to one’s identity because the pro-
cessing of such loss may actively disrupt attachments with others and may also
stimulate changes in life roles as well as life goals that are already established.
Maccallum and Bryant (2008) examined the links and any distinctive patterns
184  Ann-Marie J. Golden
between autobiographical memory and self-identity. They tested self-defining
memories in 20 participants with CG and 20 control participants and found
that participants with CG provided more self-defining memories involving the
deceased. Although both CG and non-CG groups reported a loved one’s death as
a self-defining moment, the non-CG group showed more evidence of finding ben-
efit in their memory narratives and experienced less negative emotion on recall.
Interestingly, fewer than half of participants with and without CG reported the
death as a self-defining memory. Maccallum and Bryant (2008) suggest that this
may reflect lack of integration into the autobiographical memory database due to
preferential retrieval of loss-related memories or avoidance of the reality of the
death; they further suggest that individuals who have successfully adapted to the
loss may never had loss/death as their self-constructing pivotal event.
Reported goals are also related to personal memories and to an individual’s self-
construct (e.g. Moberly & MacLeod, 2006). Maccallum and Bryant (2010), using
the standard AMT, set out to examine the relationship between autobiographical
memory and personal goals in 24 individuals with CG and 21 bereaved controls
without CG. Their results showed that individuals with CG exhibited impaired
retrieval of specific autobiographical memories in response to both positive and
negative cues. In addition they found that individuals with CG were more likely
to retrieve memories related to their loss and there was a positive relationship
between the proportion of loss-related memories recalled and the proportion of
grief-related goals held by individuals after controlling for symptom level. They
showed that autobiographical retrieval in CG is shaped by individual goals.
To recap, CG and the psychological mechanisms involved in OGM bias in
relation to rumination and executive resources are yet to be empirically explored.
Based on other clinical disorders they are likely to play an important role and it
is possible that rumination may impact on retrieval style through associations
with depression and symptoms of severe grief. OGM bias, as in the retrieval of
categorical memories, is very likely to contribute to complications in emotionally
processing and integrating the experience of death/loss. It may also have implica-
tions for pre-existing knowledge structures about the self, as this is emphasized as
a core component of grief resolution. Golden et al. (2007) demonstrated standard
OGM bias in a sample with CG and also revealed that retrieval of memories that
are related to the source of an individual’s distress (BMT-Deceased) are immune
to OGM bias – an effect that has been replicated. This finding questions theoreti-
cal accounts of OGM bias as a form of affect regulation. Maccallum and Bryant
(2008, 2010) showed that self-identity and goals are important for individuals
with CG.

The clinical relevance of OGM bias

Should OGM bias be targeted in therapy? Considering that this bias is associated
with problem-solving difficulties, an intervention that aims to help individuals to
access their store of specific memories may prove beneficial. In terms of bereaved
individuals and/or those with current CG diagnosis, this is particularly relevant
 Autobiographical memory processes and CG  185
since they should be able to retrieve memories that are not related to the deceased.
This may help them, in turn, retrieve other positive specific memories and manage
the issues that living without the deceased may cause.
There are some studies showing that OGM bias may be experimentally modifi-
able in depressed individuals. Watkins, Teasdale, and Williams (2000) showed
that decentring and distraction methods reduced OGM bias in depressed indi-
viduals. Watkins and Teasdale (2001) showed that once the rumination levels are
reduced then the OGM bias is also reduced, and this leads to retrieval of more
specific autobiographical memories.
Through CG resolution theories, the literature has emphasized that the emo-
tional processing of the loss is highly important (Boelen et al., 2006; Shear &
Shair, 2005) and individuals with CG exhibiting OGM bias may have their recov-
ery further hindered by delaying the emotional processing of loss. So the most
obvious clinical implication is to incorporate techniques or strategies within exist-
ing cognitive therapy that may increase the specificity of memory retrieval. For
example, preliminary findings from a 4-week group-based intervention entitled
MEmory Specificity Training (MEST; Raes, Williams, & Hermans, 2009) have
shown that as memory specificity increases individuals become better at problem
solving and additionally their rumination levels decrease.
There is also preliminary evidence from a treatment study conducted by
Williams, Teasdale, Segal, and Soulsby (2000) showing that mindfulness-based
cognitive therapy (MBCT; Segal, Williams, and Teasdale, 2002) reduces OGM
bias in individuals who are not currently depressed but have had at least two
previous episodes of depression. Van der Houwen et al. (2010) suggested that
mindfulness could be added to various aspects of grief therapy. Mindfulness  –
paying attention on purpose and non-judgmentally to what is happening around
and within us – when applied to grief facilitates turning towards the grief process
and being connected to all the processes that are unfolding. Through mindful
attention, grieving individuals can discover and further process that grief is not
their identity – that is, grief is not who they are. Kumar (2005) described in detail
different aspects of mindfulness that can be used when working with the bereaved,
in an interactive book entitled Grieving Mindfully. Additionally, there are aspects
of compassion-focused therapy (CFT; Gilbert, 2009) for individuals who experi-
ence self-criticism and shame since they struggle to feel reassurance or safety and
these could be further examined in bereaved individuals with CG.
Recently, Maccallum and Bryant (2011) found that OGM bias can be modi-
fied by a 10-week CBT programme in treatment-seeking individuals with CG.
As symptoms of CG reduced after CBT, positive cue word memory specificity
increased. However changes in memory specificity to negative cue words were
not associated with a reduction in CG symptomatology. Further investigations
could be tailored to add some of the aspects that specifically target OGM bias.
However, the findings by Golden et al. (2007) suggest that, for a bereaved
individual with CG, being able to retrieve specific memories that are related to the
loss of the loved one does not reflect an adaptive or a functional response. This
ongoing focus on the individual’s loss and his or her recall of specific memories
186  Ann-Marie J. Golden
related to the lost person prolongs avoidance of the reality of the loss. Training
individuals with CG to retrieve specific memories related to their life and the lives
of others may, by incorporating other aspects of their life, prove to be more useful
and supportive of emotional processing of their loss.
Maccallum and Bryant (2010) showed that there is a significant relationship
between grief-related goals and retrieval of grief-related memories; when the cur-
rent working self is focused on grief and loss goals, this may reinforce dysfunc-
tional grief processes. They also suggested that individuals with CG would benefit
from the development of an alternative goal focus. Boelen’s (2009) findings also
emphasized that bereaved individuals with more severe symptomatology of CG
experienced their loss as more central to their self-identity and life story. So intro-
ducing alternative but still life-relevant practical activities may be a way forward
without making closure the main goal to be achieved.

Future research
It is clear that factors that seem to prompt and probe autobiographical memories
and in turn prolong negative affective states following loss of a loved individual
need further examination. There are many issues that the future studies will
hopefully address, and conducting longitudinal studies and focusing on issues
surrounding comorbidity are paramount. Specifically it is important to address to
what degree OGM bias is modifiable within the bereaved population using MEST,
CBT, MBCT, and CFT. Some other avenues of examination may be:

• whether the rumination, avoidance, and executive resources are differentially

involved in the OGM bias in CG;
• what difference (if any) instructing participants to retrieve memories that
are either loss or non-loss related and/or to set limits to the content (self/
deceased/living) of a participant’s recall may make to the outcome;
• assessment of memory content and levels of specificity with a longer post-
treatment interval to overcome recency effects;
• comparing the content, rather than simply ascertaining the presence, of
responses on the memory tasks, in particular the BMT-Deceased, with the
content of the intrusions that participants with CG experience in day-to-day
life.

This chapter has shown that OGM bias is present in individuals with compli-
cated grief and that a way forward would be to incorporate targeting this OGM
bias as part of their treatment package. This would, in turn, enable bereaved indi-
viduals to incorporate loss into their current schema; that is, eventually to accept
the loss and be able to deal with the future challenges that life may bring about.
There are, as with everything, other issues that may be considered in relation to
bereavement (e.g. cultural values, attachment), and these are discussed elsewhere
in this book.
 Autobiographical memory processes and CG  187

References
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4th edn.). Washington, DC: American Psychiatric Association.
Baddeley, A. D. (1986). Working memory. Oxford: Clarendon Press.
Baddeley, A. D. (1997). Human memory: Theory and practice (revised edn.). Hove, UK:
Psychology Press.
Birch, L.  S., & Davidson, K.  M. (2007). Specificity of autobiographical memory in
depressed older adults and its relationship with working memory and IQ. British
Journal or Clinical Psychology, 46, 175–186.
Bluck, S., Alea, N., Haberman, T., & Rubin, D. C. (2005). A tale of three functions: The
self-reported uses of autobiographical memory. Social Cognition, 23, 91–117.
Boelen, P. A. (2009). The centrality of a loss and its role in emotional problems among
bereaved people. Behaviour Research and Therapy, 47, 616–622.
Boelen, P. A., van den Hout, M. A., & van den Bout, J. (2006). A cognitive–behavioral
conceptualization of complicated grief. Clinical Psychology-Science and Practice, 13,
109–128.
Boelen, P. A., Huntjens, R.  J.  C., van Deursen, D.  S., & van den Hout, M. A. (2010).
Autobiographical memory specificity and symptoms of complicated grief, depression,
and posttraumatic stress disorder following loss. Journal of Behaviour Therapy and
Experimental Psychiatry, 41, 331–337.
Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of post-
traumatic stress disorder. Psychological Review, 103, 670–686.
Bryant, R. A., Sutherland, K., & Guthrie, R. M. (2007). Impaired specific autobiographical
memory as a risk factor for posttraumatic stress after trauma. Journal of Abnormal
Psychology, 116, 837–841.
Conway, M.A (1996). Autobiographical memories and autobiographical knowledge.
In Rubin, D.  C. (Ed.), Remembering our past: Studies in autobiographical memory
(pp. 67–93). Cambridge: Cambridge University Press.
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical
memories in the self-memory system. Psychological Review, 107, 261–288.
Dalgleish, T., & Power, M. J. (2004). Emotion-specific and emotion-non-specific compo-
nents of Posttraumatic Stress Disorder (PTSD): Implications for a taxonomy of related
psychopathology. Behavior Research and Therapy, 42, 1069–1088.
Dalgleish, T., Rolfe, J., Golden, A.-M., Dunn, B., & Barnard, P. J. (2008). Reduced auto-
biographical memory specificity and posttraumatic stress: Exploring the contributions
of impaired executive control and affect regulation. Journal of Abnormal Psychology,
117, 236–241.
Dalgleish, T., Tchanturia, K., Serpell, L., Hems, S., Yiend, J., de Silva, P., & Treasure,
J. (2003). Self-reported parental abuse relates to autobiographical memory style in
patients with eating disorders. Emotion, 3, 211–222.
Dalgleish, T., Williams, J. M. G., Golden, A.-M., Perkins, N., Barrett, L. F., Barnard, P. J.,
et al. (2007). Reduced specificity of autobiographical memory and depression: The role
of executive control. Journal of Experimental Psychology: General, 136, 23–42.
Dickson, J. M., & Bates, G. W. (2006). Autobiographical memories and views of the future:
In relation to dysphoria. International Journal of Psychology, 41, 107–116.
Ehlers, A., & Clark, D.  M. (2000). A cognitive model of posttraumatic stress disorder.
Behaviour Research and Therapy, 38, 319–345.
Gibbs, B. R., & Rude, S. S. (2004). Overgeneral autobiographical memory as depression
vulnerability. Cognitive Therapy and Research, 28, 511–526.
188  Ann-Marie J. Golden
Gilbert, P. (2009). Introducing compassion-focused therapy. Advances in Psychiatric
Treatment, 15, 199–208.
Golden, A-M., Dalgleish, T., & Mackintosh, B. (2007). Levels of specificity of autobio-
graphical memories and of biographical memories of the deceased in bereaved indi-
viduals with and without complicated grief. Journal of Abnormal Psychology, 116,
786–795.
Harvey, A., Watkins, E., Mansell, W., & Shafran, R. (2004). Cognitive behavioural pro-
cesses across psychological disorders: A transdiagnostic approach to research and
treatment. Oxford: Oxford University Press.
van der Houwen, K., Stroebe, M., Stroebe, W., Schut, H., van den Bout, J., & Wijngaards-de
Meij, L. (2010). Risk factors for bereavement outcome: A multivariate approach. Death
Studies, 34, 195–220.
Kumar, S. M. (2005). Grieving mindfully: A compassionate spiritual guide to coping with
loss. Oakland, CA: New Harbinger Publications.
Kuyken, W., & Brewin, C. R. (1995). Autobiographical memory functioning in depression
and reports of early abuse. Journal of Abnormal Psychology, 104, 585–591.
Kuyken, W., & Dalgleish, T. (1995). Autobiographical memory and depression. British
Journal of Clinical Psychology, 33, 89–92.
Maccallum, F., & Bryant, R.  A. (2008). Self-defining memories in complicated grief.
Behaviour Research and Therapy, 46, 1311–1315.
Maccallum, F., & Bryant, R. A. (2010). Impaired autobiographical memory in complicated
grief. Behaviour Research and Therapy, 48, 328–334.
Maccallum, F. & Bryant, R.  A. (2011). Autobiographical memory following cogni-
tive behaviour therapy for complicated grief. Journal of Behaviour Therapy and
Experimental Psychiatry, 42, 26–31.
McNally, R. J. (2003). Remembering trauma. Cambridge, MA: Harvard University Press.
McNally, R. J., Lasko, N. B., Macklin, M. L., & Pitman, R. K. (1995). Autobiographical
memory disturbance in combat-related post-traumatic stress disorder. Behaviour
Research and Therapy, 33, 619–630.
Moberly, N. J., & MacLeod, A. K. (2006). Goal pursuit, goal self-concordance, and the
accessibility of autobiographical knowledge. Memory, 14, 901–915.
Moore, S. A., & Zoellner, L. A. (2007). Overgeneral autobiographical memory and trau-
matic events: An evaluative review. Psychological Bulletin, 133, 419–437.
Neimeyer, R. A. (1998). Lessons of loss: A guide to coping. New York: McGraw Hill.
Neimeyer, R. A., Baldwin, S. A., & Gillies, J. (2006). Continuing bonds and reconstructing
meaning: Mitigating complications in bereavement. Death Studies, 30, 715–738.
Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of
depressive episodes. Journal of Abnormal Psychology, 100, 569–582.
Prigerson, H.  G., & Maciejewski, P.  K. (2005). A call for sound empirical testing and
evaluation of criteria for Complicated Grief proposed for DSM-V. Omega-Journal of
Death and Dying, 52, 9–19.
Raes, F., Williams, J.  M.  G., & Hermans, D. (2009). Reducing cognitive vulnerability
to depression: A preliminary investigation of Memory Specificity Training (MEST)
in inpatients with depressive symptomatology. Journal of Behaviour Therapy and
Experimental Psychiatry, 40, 24–38.
Ramponi, C., Barnard, P., & Nimmo-Smith, I. (2004). Recollection deficits in dysphoric
mood: An effect of schematic models and executive mode? Memory, 12, 655–670.
Rottenberg, J., Hildner, J. C., & Gotlib, I. H. (2006). Idiographic autobiographical memo-
ries in major depressive disorder. Cognition & Emotion, 20, 114–128.
 Autobiographical memory processes and CG  189
Segal, Z. V., Williams, J. M. G., & Teasdale, J. D. (2002). Mindfulness-based cognitive
therapy for depression: A new approach for preventing relapse. New York: Guilford
Press.
Shear, K. M., & Shair, H. (2005). Attachment, loss, and complicated grief. Developmental
Psychobiology, 47, 253–267.
Sumner, J. A., Griffith, J. W., & Mineka, S. (2010). Overgeneral autobiographical memory
as a predictor of the course of depression: A meta-analysis. Behaviour Research and
Therapy, 48, 614–625.
Teasdale, J. D., & Fogarty, S. J. (1979). Differential effects of induced mood on retrieval
of pleasant and unpleasant memories from episodic memory. Journal of Abnormal
Psychology, 88, 248–257.
Watkins, E., & Teasdale, J.  D. (2001). Rumination and overgeneral memory in depres-
sion: Effects of self-focus and analytic thinking. Journal of Abnormal Psychology, 110,
353–357.
Watkins, E., Teasdale, J. D., & Williams, R. M. (2000). Decentring and distraction reduce
over general autobiographical memory in depression. Psychological Medicine, 30,
911–920.
Williams, H.  L., Conway, M.  A., & Cohen, G. (2008). Autobiographical memory. In
Cohen, G., & Conway, M. A. (Eds.), Memory in the real world (3rd edn., pp. 21–90).
Hove, UK: Psychology Press.
Williams, J.  M.  G. (1996). Depression and the specificity of autobiographical memory.
In Rubin, D.  C. (Ed.), Remembering our past: Studies in autobiographical memory
(pp. 244–267). New York: Cambridge University Press.
Williams, J. M. G. (2006). Capture and rumination, functional avoidance, and executive
control (CaRFAX): Three processes that underlie overgeneral memory. Cognition and
Emotion, 20, 548–568.
Williams J. M. G., Barnhofer T., Crane C., & Beck A. T. (2005). Problem solving deterio-
rates following mood challenge in formerly depressed patients with a history of suicidal
ideation. Journal of Abnormal Psychology, 114, 421–431.
Williams, J.  M.  G., Barnhofer, T., Crane, C., Hermans, D., Raes, F., Watkins, E., &
Dalgleish, T. (2007). Autobiographical memory specificity and emotional disorder.
Psychological Bulletin, 113, 122–148.
Williams, J. M. G., & Broadbent, K. (1986). Autobiographical memory in suicide attempt-
ers. Journal of Abnormal Psychology, 95, 144–149.
Williams J. M. G., Chan S., Crane C., Barnhofer T., Eade J., & Healy H. (2006). Retrieval
of autobiographical memories: The mechanisms and consequences of truncated search.
Cognition & Emotion, 20, 351–382.
Williams, J. M. G., & Dritschel, B. H. (1988). Emotional disturbance and the specificity of
autobiographical memory. Cognition and Emotion, 2, 221–234.
Williams, J.  M.  G., & Dritschel, B. (1991). Categoric and extended autobiographical
memories. In Conway, M., Rubin, H., Spinnler, W., & Wagennar, W. (Eds.), Theoretical
perspectives on autobiographical memory (pp. 391–409). Dodrecht: Kluwer Academic
Publishers.
Williams, J.  M.  G., Stiles, W.  B., & Shapiro, D. (1999). Cognitive mechanisms in the
avoidance of painful and dangerous thoughts: Elaborating the assimilation model.
Cognitive Therapy and Research, 23, 285–306.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V., & Soulsby, J. (2000). Mindfulness-based
cognitive therapy reduces overgeneral autobiographical memory in formerly depressed
patients. Journal of Abnormal Psychology, 109, 150–155.
14 Attachment insecurities and
disordered patterns of grief
Mario Mikulincer and Phillip R. Shaver

As everyone with an interest in bereavement knows, Bowlby’s (1980) extension

of attachment theory into the realm of bereavement and grieving was a major
component of his now massively influential theoretical work. In countless arti-
cles and many books, including a recent integration of the literature by Bowlby’s
long-time colleague Parkes (2006), Bowlby’s ideas have been tested, challenged,
applied clinically, and extended. In the present chapter we offer a brief overview
of the theory and its relevance to complicated, or disordered, grieving; its tests in
our and other contemporary investigators’ studies; and the potential applications
of our research to clinical interventions for individuals suffering from compli-
cated or disordered grief.
We begin with an overview of attachment theory and explain the theory’s
perspective on loss, bereavement, and both normative and complicated grieving.
We then review empirical evidence demonstrating how attachment insecurities
are involved in complicated patterns of grief. At the end of the chapter we offer
a brief, but we hope seminal, research-based perspective on ways to prevent and
treat complicated grief.

Overview of attachment theory

Bowlby (1982) claimed that human beings are born with an innate psychobiologi-
cal system (which he called the attachment behavioral system) that motivates them
to seek proximity to supportive others (attachment figures) in times of need. He
also (Bowlby, 1973) delineated important individual differences in attachment-
system functioning, which he attributed mainly to the history of interactions with
attachment figures, beginning in infancy. Interactions with attachment figures
who are available, sensitive, and responsive in times of need promote the normal
functioning of the attachment system and encourage the development of a stable
sense of security and positive mental representations of self and others, which
Bowlby (1973) called internal working models. Unfortunately, when a person’s
attachment figures have not been reliably available, sensitive, and supportive,
negative working models of self (as not sufficiently lovable) and others (as unac-
cepting and unresponsive) are likely to have developed.
 Attachment insecurities and disordered grief  191
When studying individual differences in attachment-system functioning
in adults, attachment research has focused mainly on attachment orientations:
a person’s chronic pattern of relational expectations, emotions, and behaviors
resulting from the internalization and memory of a particular attachment history
(Fraley & Shaver, 2000). Specifically, many recent studies of adult attachment
(reviewed by Mikulincer & Shaver, 2007a) have found that individual differences
in attachment orientations can be measured along two continuous dimensions of
attachment anxiety and avoidance (Brennan, Clark, & Shaver, 1998), which are
roughly orthogonal. Attachment anxiety reflects the degree to which a person wor-
ries that relationship partners will not be available in times of need and is afraid
of being rejected or abandoned. Attachment-related avoidance reflects the extent
to which a person distrusts relationship partners’ goodwill and strives to maintain
independence and emotional distance from partners. People who score low on
both dimensions are said to be secure, or secure with respect to attachment. The
two dimensions can be measured with reliable and valid self-report scales (e.g.,
Brennan et al., 1998) and are associated in theoretically predictable ways with
many aspects of mental health and relationship quality (see Mikulincer & Shaver,
2007a, for a review).
We (Mikulincer & Shaver, 2007a) have proposed that a person’s location in
the two-dimensional space defined by attachment anxiety and avoidance reflects
both the person’s sense of attachment security and the ways in which he or she
deals with threats and stressors. People who score low on these dimensions are
generally secure and tend to employ constructive and effective affect-regulation
strategies. Those who score high on either attachment anxiety or avoidance, or
both (a condition called fearful avoidance), suffer from attachment insecurities
and tend to use secondary attachment strategies (Cassidy & Kobak, 1988) in
an effort to cope with threats, frustrations, losses, and insecurities. People who
score high on attachment anxiety rely on hyperactivating strategies: energetic
attempts to achieve support and love combined with a lack of confidence that
these resources will be provided and with feelings of anger and despair when
they are not provided (Cassidy & Kobak, 1988). In contrast, people who score
high on avoidant attachment tend to use deactivating strategies: trying not to seek
proximity to others when threatened, denying vulnerability and a need for other
people, and avoiding closeness and interdependence in relationships. Research
has consistently shown that attachment anxiety and avoidance can be viewed as
risk factors for emotional and behavioral problems (see Mikulincer & Shaver,
2007a, for an extensive review of results).

An attachment perspective on loss and grief

According to attachment theory, the loss of an attachment figure is a profound
blow that triggers intense distress and a predictable series of responses, which
Bowlby (1982) originally called protest, despair, and detachment. In childhood,
the initial response to separation from an attachment figure is protest: The child
192  Mario Mikulincer and Phillip R. Shaver
very actively resists separation by crying, calling, searching, and clinging in an
attempt to regain contact. If protest fails to restore proximity, as is obviously
the case following the death of an attachment figure, these vigorous reactions
eventually wane and anxiety and anger give way to pervasive despair, includ-
ing depressed mood, pained expressions, decreased appetite, and disturbed sleep.
Over time, despair usually gives way to what Bowlby (1982) called “detach-
ment”: apparent recovery and gradual renewal of interest in other activities and
new relationship partners.
Bowlby (1980) viewed adult romantic, or pair-bond, relationships as the
primary attachment bonds in adulthood and assumed that adults who lose their
long-term romantic, or marital, partner will undergo a series of reactions similar
to those observed in infants (see also Parkes & Weiss, 1983; Shaver & Fraley,
2008). Like infants, bereaved adults react with strong protest, anger, calling, and
yearning for reunion. When a person fully realizes that his or her partner will not
return, despair and disorganization can ensue, accompanied by sleeping and eating
disturbances, social withdrawal, intense sorrow, and loneliness (Weiss, 1991). In
the case of adult bereavement, Bowlby (1980) preferred to call the final phase of
grief responses “reorganization” rather than “detachment,” because adults often
transfer their proximity seeking, at least partly, to new relationship partners with-
out fully detaching from the lost partner. According to Bowlby (1980), adults
can rearrange their attachment representations so that the deceased can continue
to serve as a symbolic source of protection while new relationships with living
partners are formed on new foundations.
According to Bowlby (1980), “reorganization” is the optimal resolution of the
bereavement process. It involves two major psychological tasks: (1) accepting
the death of the lost partner, returning to mundane activities, and forming new
relationships, and (2) maintaining some kind of symbolic bond to the deceased
and integrating the lost relationship within a new reality. These tasks require an
“editing” of the hierarchy of attachment figures in a process that resembles the
replacement of parents by peers as primary attachment figures during adolescence
(e.g., Hazan & Zeifman, 1999; Weiss, 1991). Psychologically successful mourn-
ers can maintain a symbolic bond to the deceased while adjusting to real circum-
stances, and restore and even enhance their sense of security and well-being based
on both the continuing attachment bond with the deceased and new attachment
bonds with living companions.
Bowlby’s (1980) reasoning fits well with the various dual-process models of
bereavement (e.g., Rubin, 1991; Stroebe & Schut, 1999). For instance, Stroebe
and Schut (1999) viewed adjustment to loss as a dynamic oscillation between
loss orientation and restoration orientation. Loss orientation includes yearning,
rumination, separation distress, and reappraisal of the meaning and implications
of the loss. Restoration orientation includes attending to life changes, doing new
things, distracting oneself from grief, denying or suppressing grief, and form-
ing new relationships. In this model, oscillation between these two orientations
brings about a gradual reorganization of life and mind, such that the deceased is
integrated into the bereaved individual’s identity and he or she establishes new
relationships and finds new meanings in life.
 Attachment insecurities and disordered grief  193

Attachment insecurities and disordered patterns of grief

Beyond describing the normative processes of bereavement and coping with the
loss of a close relationship partner, Bowlby (1980) proposed a framework for
conceptualizing disordered patterns of mourning and suggested that attachment
insecurities can sometimes complicate grief. Anxiously attached individuals are
unwilling or unable to handle many life tasks on their own and to suppress or
inhibit painful feelings, thoughts, and memories related to a deceased partner,
thereby making attachment reorganization impossible. Avoidant individuals are
unwilling or unable to experience thoughts, feelings, and memories related to a
deceased partner, which makes it difficult to create meaning from the loss and
integrate the lost relationship into a new reality. In both cases, attachment-related
worries and defenses may interfere with attachment reorganization and compli-
cate the grief process.
Bowlby (1980) suggested that attachment insecurities contribute to two major
forms of disordered mourning: “chronic mourning” and “prolonged absence of
conscious grieving” (p. 138). Chronic mourning is characterized by overwhelm-
ing anxiety and sadness, prolonged difficulty in re-establishing normal function-
ing and forming new relationships, preoccupation with the deceased partner, and
maintenance of intense emotional bonds with the deceased. In contrast, prolonged
absence of grief is characterized by lack of overt expressions of sadness or dis-
tress, detachment from the deceased partner, and continuation of normal life
without major disruptions. Most clinicians agree with Bowlby’s conceptualiza-
tion of these two forms of disordered mourning, although they tend to label the
absence of grief “delayed grief,” “inhibited mourning,” or “absent mourning”
(see Stroebe, Hansson, Stroebe, & Schut, 2001). According to Bowlby (1980),
whereas attachment anxiety underlies chronic mourning, attachment-related
avoidance contributes to the absence of grief.

Anxious attachment and chronic mourning

Even when their relationship partners are alive, attachment-anxious people are
preoccupied with their responsiveness and supportiveness, likely to make intru-
sive demands for greater love and attention, and eager for support and reassurance
(Shaver, Schachner, & Mikulincer, 2005). Not surprisingly, therefore, when they
lose a relationship partner to death, they are likely to experience intense preoccu-
pation with the deceased, yearn inconsolably for the lost source of protection and
support, fail to accept the loss, and have difficulty establishing new relationships.
These are some of the core features of chronic mourning.
Another characteristic of chronic mourning is that the bereaved person is fre-
quently overwhelmed by painful memories of the deceased and is unable to cope
with or manage them effectively (e.g., Boelen, van den Hout, & van den Bout,
2006). This inability to control the flow of painful memories is one of the most
salient characteristics of attachment-anxious adults. For example, Mikulincer and
Orbach (1995) found that anxiously attached individuals had very fast access to
painful memories and that these memories spread like wildfire throughout their
194  Mario Mikulincer and Phillip R. Shaver
cognitive system. Using the Adult Attachment Interview, Roisman, Tsai, and
Chiang (2004) found similar signs of emotional dysregulation when anxiously
attached individuals were asked to talk about their emotional experiences (i.e.,
they exhibited facial expressions of sadness or anger even they were speaking
about neutral or positive childhood experiences). We believe that this is the
same kind of confusion and disorganization that occurs when anxiously attached
mourners are bombarded with intrusive images, feelings, and memories related to
the deceased person, although it is likely to be even more intense in the case of
bereavement.
Another characteristic of chronic mourning is the pervasive presence of nega-
tive beliefs about the self and the future (e.g., Boelen, van den Bout, & van den
Hout, 2003). According to Foa and Rothbaum (1998), these negative beliefs can
be particularly resistant to change when the loss confirms negative views of the
self, hopeless beliefs, and catastrophic cognitions that were present before the
loss. This is more likely to be the case when a person was already anxiously
attached. There is extensive evidence that anxiously attached individuals tend to
possess negative views of self, to exaggerate even fairly minor threats, and to hold
pessimistic beliefs about managing distress (see Mikulincer & Shaver, 2007a, for
a review).

Avoidant attachment and the absence of grief

Avoidant people habitually deal with stress and distress by denying attachment
needs, suppressing attachment-related thoughts and emotions, and inhibiting
unwanted urges to seek proximity or support (Mikulincer & Shaver, 2007a).
Following the death of an attachment figure, avoidant people are likely to use their
deactivating defenses to inhibit anxiety and despair, downplay the importance of
the loss, and try to steer clear of thoughts and memories focused on the deceased.
This is what Bowlby (1980) meant by the absence of grief. He considered this
to be a defensive reaction involving redirection of attention away from painful
thoughts and feelings (“defensive exclusion”) and the segregation or dissociation
of memories of the deceased which nevertheless continue to influence emotions
and behaviors without the individual’s awareness of their effects.
Bowlby (1980) thought the prolonged absence of grief could eventually lead
to difficulties in mental and physical health. He expected that people who fail to
mourn would have difficulties integrating losses meaningfully into their work-
ing models and schemas. Since a bereaved person is likely to have engaged in
many daily activities with the now-deceased partner, each of these activities or
the places where they occurred becomes an unwanted reminder of the loss and
a further source of either distress or the need to suppress thoughts and feelings.
Of course, the negative emotional and physical consequences of “absence of
grieving” are most likely to emerge in cases where the bereaved individual was
deeply attached to the lost partner. If an avoidant person was able to avoid close-
ness and interdependence with a partner while he or she was alive, he or she
may experience less anxiety and despair following the death of this partner even
 Attachment insecurities and disordered grief  195
without strenuous attachment deactivation. In such cases, the absence of grieving
may reflect a real absence of distress because there was no strong emotional bond
between the bereaved individual and the deceased. Compatible with this idea,
many people who show few signs of grief immediately after the death of a partner
do not exhibit heightened distress and emotional problems months or years later
(see Bonanno, 2001, for a review).
It may be difficult, in particular cases, to tell the difference between success-
ful but very active suppression, on one hand, and a true absence of anything to
suppress, on the other. For several years, our own research was unclear about the
existence of this difference. Now, however, Bowlby’s (1980) ideas about avoidant
people’s defensive suppression of memories, thoughts, and feelings concerning
separations and losses have been well supported in a series of experiments con-
ducted in our laboratories.
Fraley and Shaver (1997) asked participants to write about whatever thoughts
and feelings they were experiencing while being allowed to think about anything
except thoughts about their long-term relationship partner leaving them for
someone else. Findings indicated that avoidant attachment was associated with
both less frequent thoughts of loss following the suppression task and lower skin
conductance during the task, suggesting that avoidant people may be able to block
unwanted thoughts and to prevent the emotional arousal these thoughts might
otherwise cause. A subsequent study used functional magnetic resonance imaging
(fMRI) to observe brain processes during this same suppression task (Gillath,
Bunge, Shaver, Wendelken, & Mikulincer, 2005) and found that avoidant attach-
ment was associated with a particular pattern of activation and deactivation in
brain regions involved in suppression.
Avoidant people’s ways of dealing with losses are also manifested in what
Fraley, Garner, and Shaver (2000) called preemptive defenses, for example,
directing attention away from loss-related information, or encoding it in a shal-
low way. Fraley et al. (2000) asked participants to listen to a genuinely emotional
interview about the loss of a close relationship partner and then to recall details
of the interview, either soon after hearing it or at various delays ranging from half
an hour to 21 days. Whereas avoidant people initially encoded less loss-related
information, they did not differ from more secure participants in the rate of forget-
ting the encoded information. It therefore seems that avoidant people defend from
loss-related information preemptively, by blocking threatening material from
awareness before it is fully encoded.
Although these findings imply that avoidant people are effective in suppressing
painful memories and thoughts, Mikulincer, Dolev, and Shaver (2004) found that
avoidant people can nevertheless be disturbed by the unwanted resurgence of sup-
pressed thoughts. In two studies, participants were asked to think about a painful
relationship breakup, were either instructed or not instructed to suppress thoughts
about this separation, and then performed a Stroop color-naming task under con-
ditions of low or high cognitive load. In the Stroop task, participants were asked
to indicate the color in which words related to the painful separation (Study 1)
or negative and positive self-traits (Study 2) were printed. Longer color-naming
196  Mario Mikulincer and Phillip R. Shaver
latencies implied greater accessibility to suppressed thoughts of separation or to
self-related traits. In this way, Mikulincer et al. (2004) examined whether avoid-
ant defenses are capable of inhibiting the post-suppression rebound effect and the
intrusion of self-related worries (heightened access to negative self-traits) even
when other cognitive demands draw upon limited psychological resources.
Under low cognitive load, findings of the two studies revealed that avoidant
people were able to suppress thoughts related to the breakup and had poorer access
to such thoughts and enhanced access to positive self-representations in the Stroop
task following suppression. However, the effectiveness of avoidant defenses was
significantly impaired when a high cognitive load taxed the mental resources
needed to maintain thought suppression. Under a high cognitive load, avoidant
people exhibited greater automatic activation of thoughts of separation and nega-
tive self-traits following suppression. In other words, their defenses collapsed
when mental resources were too scarce to maintain them, and this collapse was
associated with a spread of activation from unwanted separation-related thoughts
to formerly suppressed negative self-representations. Studies showing that avoid-
ant people do sometimes experience strong negative emotions in response to
chronic, uncontrollable, and severely distressing events also suggest a breakdown
of defenses when the mental load includes not just cognitive processes but intense
stress as well (e.g., Berant, Mikulincer, & Shaver, 2008). In line with Bowlby’s
(1980) analysis, these findings indicated that avoidant individuals’ attempts to
suppress painful thoughts concerning a loss fail to eliminate the distress, and the
suppressed material can resurface in experience and action when high cognitive
or emotional demands are encountered.

Empirical evidence on the links between attachment insecurities and

complicated grief
A few studies have directly examined the associations between a history of inse-
cure attachment experiences in childhood (separation anxiety, childhood abuse,
parental death) and bereavement in adulthood and found that such a history
was associated with complicated patterns of grief (e.g., Silverman, Johnson, &
Prigerson, 2001; Vanderwerker, Jacobs, Parkes, & Prigerson, 2006). In addition,
adult attachment studies have consistently found that self-reports of attachment
anxiety are positively associated with complicated grief reactions (e.g., Field
& Sundin, 2001; Fraley & Bonanno, 2004; Jerga, Shaver, & Wilkinson, 2011;
Wayment & Vierthaler, 2002; Wijngaards-de Meij et al., 2007a, 2007b). For
example, Field and Sundin (2001) found that anxious attachment, assessed 10
months after the death of a spouse, predicted higher levels of psychological dis-
tress 14, 25, and 60 months after the loss, and Fraley and Bonanno (2004) found
that attachment anxiety assessed 4 months after the loss of a spouse predicted
higher levels of anxiety, depression, grief, trauma-related symptoms, and alcohol
consumption 18 months following the loss.
There is also evidence concerning the association between attachment anxi-
ety and difficulties in attachment reorganization. Field and Sundin (2001), for
 Attachment insecurities and disordered grief  197
example, found that higher scores in attachment anxiety were associated with
more positive thoughts about a deceased spouse 14 months after the loss, perhaps
reflecting a continuing emotional investment in an idealized figure. This kind of
idealization was also evident in Nager and De Vries’s (2004) qualitative study of
the contents of memorial websites created by adult daughters in memory of their
deceased mothers. Using the Continuing Bonds Scale, Waskowic and Chartier
(2003) found that, although secure and anxious people scored relatively high on
the scale, anxious people scored higher than their secure counterparts on rumina-
tion about and preoccupation with a lost spouse.
With regard to avoidant attachment, studies have generally found no significant
association between this attachment orientation and depression, grief, or distress
following the death of a relationship partner (Field & Sundin, 2001; Fraley &
Bonanno, 2004; Wayment & Vierthaler, 2002). However, Wayment and Vierthaler
(2002) found that avoidance was associated with higher levels of somatic symp-
toms following the death of spouse, and Wijngaards-de Meij et al. (2007a, 2007b)
found that avoidant people exhibited higher levels of grief and depression follow-
ing the death of their child than secure people.
Recently, Jerga et al. (2011) reviewed previous studies that failed to find sig-
nificant associations between avoidant attachment and complicated grief and con-
cluded that the failure might have occurred for four methodological reasons. First,
although past studies assessed grief symptoms, they failed to assess symptoms of
prolonged or complicated grief (Prigerson et al., 2009). Second, participants in
previous studies were aware, when they agreed to participate, that the study was
designed to investigate loss and grief, perhaps biasing the sample and causing
people who were struggling with their loss, or were using an avoidant strategy to
keep from thinking about it, to choose not to participate. Third, previous research
did not always control for the closeness or strength of the relationship with the
deceased, so avoidant people may have reported losses of a less significant nature
than the ones described by other participants. Fourth, previous studies often
involved small samples and may have had insufficient statistical power to detect
a link between avoidant attachment and complicated grief.
To begin to deal with these methodological problems, Jerga et al. (2011) con-
ducted a cross-sectional study with a large sample, measuring both manifestations
of typical grief and symptoms of prolonged grief, assessing attachment-related
avoidance both generally and with respect to the lost partner, disguising the nature
of the study in the beginning to avoid sampling biases, and controlling statistically
for relationship closeness and strength. The researchers found that global avoidant
attachment was positively associated with prolonged grief symptoms but not with
typical or normative grief symptoms. That is, people who are generally avoidant
in close relationships experience long-term difficulties adjusting to the death of a
relationship partner, even though they do not necessarily experience more intense
typical grief symptoms.
Interestingly, Jerga et al. (2011) also found that relationship-specific avoid-
ance was negatively associated with both typical and prolonged grief symptoms.
However, Jerga et al. (2011) showed that this association disappeared when
198  Mario Mikulincer and Phillip R. Shaver
measures of relationship closeness and strength were statistically controlled, sug-
gesting that avoidant individuals may maintain relatively weak and emotionally
distant relationships with the deceased, which in turn leaves them with less to
grieve about. In other words, it is not avoidant attachment per se that protects
avoidant individuals from grief symptoms; it is rather the weakness of the emo-
tional bonds they have to contend with when a relationship partner dies. Overall,
these findings emphasize that researchers should take seriously Bowlby’s (1980)
cautions about assessing grief responses among avoidant people, because “in all
studies except those using the most sophisticated of methods, it is easy to over-
look such people and to group them with those whose mourning is progressing in
a genuinely favorable way” (p. 211).

An attachment perspective on grief resolution and counseling

If attachment insecurities are risk factors for complicated grief, then the initiation,
maintenance, or restoration of a sense of attachment security should increase resil-
ience and facilitate attachment reorganization and a more optimal resolution of
the grief process. According to attachment theory, interactions with available and
supportive attachment figures impart a sense of safety, arouse positive emotions
(e.g., relief, love), and provide psychological resources for dealing with problems
and adversities (Bowlby, 1988). Hundreds of cross-sectional and longitudinal
studies reviewed by Mikulincer and Shaver (2007a) have found that secure indi-
viduals are likely to remain relatively unperturbed in times of stress, to recover
rapidly from periods of distress, and to experience relatively long stretches of
positive affectivity, which can increase resilience and mental health following the
loss of a relationship partner.
Bowlby (1988) also suggested that attachment security, which can be either
dispositional or rooted in a lost relationship, facilitates attachment reorganization
and makes healthy adjustment to loss more likely. Secure individuals can recall
and think about a lost partner without extreme difficulty, can acknowledge feel-
ings of love and grief, and can talk about the loss without being overwhelmed by
distress and pain (Stroebe, Schut, & Stroebe, 2005). In addition, secure individu-
als’ positive models of others allow them to continue to think positively about the
deceased, whereas their positive models of self allow them to cope with the loss
and begin to form new relationships. They can therefore be open to new relation-
ships and invest emotionally in new partners without defensively detaching from
the deceased partner.
Attachment research has shown that self-reports of dispositional attachment
security are associated with better emotional adjustment following the loss of
a relationship partner (e.g., Fraley & Bonanno, 2004; van Doorn, Kasl, Beery,
Jacobs, & Prigerson, 1998; Waskowic & Chartier, 2003; Wayment & Vierthaler,
2002). For example, Fraley and Bonanno (2004) found that people classified as
securely attached 4 months after the loss of a spouse reported relatively low levels
of bereavement-related anxiety, grief, depression, and posttraumatic distress con-
currently (at 4 months) and 18 months after the loss.
 Attachment insecurities and disordered grief  199
In some of our studies we examined effects of increased security on various
indicators of mental health by experimentally activating mental representations
of supportive attachment figures (e.g., Mikulincer, Hirschberger, Nachmias, &
Gillath, 2001). Although none of these studies examined the effects of momentary
activation of the sense of attachment security on the grief process, they provide
clues that might guide and enrich future bereavement studies. Using several
research techniques (which we refer to as “security priming”; Mikulincer &
Shaver, 2007b), including subliminal exposure to names of people designated
by participants as security-enhancing attachment figures and visualization of the
faces of security-enhancing attachment figures, we have consistently found that
security priming improved participants’ moods even in threatening contexts (e.g.,
Mikulincer et al., 2001) and mitigated cognitive symptoms of posttraumatic stress
disorder and eating disorders (Admoni, 2006; Mikulincer, Shaver, & Horesh,
2006).
In a recent laboratory experiment, we (Cassidy, Shaver, Mikulincer, & Lavy,
2009) took another step forward by showing that a momentary increase in the
sense of attachment security can mitigate hyperactivating and deactivating
affect-regulation strategies that are likely to complicate the process of griev-
ing. Participants completed a measure of dispositional attachment anxiety and
avoidance and wrote a description of an incident in which a close relationship
partner hurt them. Participants then completed a short computerized task in
which they were repeatedly exposed subliminally (for 22 milliseconds) to either a
security-enhancing prime word (love, affection) or a neutral prime (lamp, staple).
Immediately after the priming trials, participants were asked to think again about
the hurtful event they had described and to rate how they would react to such an
event if it happened in the future: how rejected they would feel, how they would
feel about themselves, and how well each of nine action descriptions fitted the
way they would react (various defensive and hostile reactions, various construc-
tive reactions, and crying).
In the neutral priming condition, the findings conformed to the usual deacti-
vating and hyperactivating strategies of avoidant and anxious people. Avoidance
scores were associated with less negative appraisals of the hurtful episode, less
intense feelings of rejection, less crying, and more defensive/hostile reactions.
In the neutral priming condition, attachment anxiety was associated with more
intense feelings of rejection, more crying, and more negative emotions. These
typical correlational findings were dramatically reduced in size (most approached
zero) in the security-priming condition. In other words, security priming reduced
the tendency of avoidant people to rely on deactivating defenses and the tendency
of anxious people to react hyperactively to hurtful episodes and psychological
pain.
These findings may be useful in planning future bereavement research and may
assist in developing interventions for complicated grief. When a person is already
secure, based on a history of supportive relationships, he or she can deal construc-
tively with the loss of a relationship partner. When a person has a history of unreli-
able, unpredictable attachment figures and is dispositionally anxious as a result,
200  Mario Mikulincer and Phillip R. Shaver
he or she is likely to intensify distress and cope badly with a loss. However, this
habitual hyperactivating response can be notably softened by a boost in felt secu-
rity accomplished through symbolic sources (as in our studies). We also believe
that actual interactions with a security provider (e.g., a counselor or therapist)
may also boost felt security and then soften habitual hyperactivating responses. In
contrast, people who have had to deal, historically, with cool, consistently unsup-
portive attachment figures and have developed a defensively avoidant pattern of
relating to others may be somewhat protected from the pain and distress usually
caused by a loss. However, this benefit may come at the expense of consider-
able problems in attachment reorganization and meaning construction following
a loss. This habitual protective armor can be at least temporarily softened by an
infusion of felt security, which seems to allow an avoidant person to become
more open to inner pain (Cassidy et al., 2009). This openness may provide an
opening for exploring the loss experience and learning more about the importance
of relationships and their relevance for finding meaning in life.
These ideas fit with Bowlby’s (1988) notion that therapeutic change begins
with the formation of a secure attachment bond with a therapist. Basically, thera-
pists need to provide what attachment theory calls a safe haven and secure base
for a client, which are prerequisites for boldly exploring painful experiences,
reworking mental representations of self and others, and learning new approaches
to behaving within relationships. These beneficial effects of enhanced attach-
ment security have already been documented in studies of treating clients with
mood disorders. For example, in a study based on data from the multi-site NIMH
(National Institute of Mental Health) Treatment of Depression Collaborative
Research Program, Zuroff and Blatt (2006) found that a client’s positive views of
his or her therapist’s sensitivity and supportiveness predicted relief from depres-
sion and maintenance of therapeutic benefits over an 18-month period. Future
research should examine whether the formation of a secure attachment bond with
a therapist has beneficial effects in therapeutic interventions with people suffering
from complicated grief.

References
Admoni, S. (2006). Attachment security and eating disorders. Unpublished doctoral dis-
sertation, Bar-Ilan University, Ramat Gan, Israel.
Berant, E., Mikulincer, M., & Shaver, P. R. (2008). Mothers’ attachment style, their mental
health, and their children’s emotional vulnerabilities: A seven-year study of children
with congenital heart disease. Journal of Personality, 76, 31–66.
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2003). The role of cognitive vari-
ables in psychological functioning after the death of a first degree relative. Behavior
Research and Therapy, 41, 1123–1136.
Boelen, P. A., van den Hout, M. A., & van den Bout, J. (2006). A cognitive–behavioral
conceptualization of complicated grief. Clinical Psychology: Science and Practice, 13,
109–128.
Bonanno, G. (2001). Grief and emotion: A social-functional perspective. In Stroebe, M.,
Stroebe, W., Hansson, R. O., & Schut, H. A.  W. (Eds.), Handbook of bereavement
 Attachment insecurities and disordered grief  201
research: Consequences, coping, and care (pp. 493–515). Washington, DC: American
Psychological Association.
Bowlby, J. (1973). Attachment and loss, vol. 2. Separation: Anxiety and anger. New York:
Basic Books.
Bowlby, J. (1980). Attachment and loss, vol. 3. Sadness and depression. New York: Basic
Books.
Bowlby, J. (1982). Attachment and loss, vol. 1. Attachment (2nd edn.). New York: Basic
Books.
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London:
Routledge.
Brennan, K. A., Clark, C.  L., & Shaver, P.  R. (1998). Self-report measurement of adult
romantic attachment: An integrative overview. In Simpson, J.  A., & Rholes, W.  S.
(Eds.), Attachment theory and close relationships (pp.  46–76). New York: Guilford
Press.
Cassidy, J., & Kobak, R. R. (1988). Avoidance and its relationship with other defensive
processes. In Belsky, J., & Nezworski, T. (Eds.), Clinical implications of attachment
(pp. 300–323). Hillsdale, NJ: Erlbaum.
Cassidy, J., Shaver, P.  R., Mikulincer, M., & Lavy, S. (2009). Experimentally induced
security influences responses to psychological pain. Journal of Social and Clinical
Psychology, 28, 463–478.
van Doorn, C., Kasl, S. V., Beery, L. C., Jacobs, S. C., & Prigerson, H. G. (1998). The
influence of marital quality and attachment styles on traumatic grief and depressive
symptoms. Journal of Nervous and Mental Disease, 186, 566–573.
Field, N. P., & Sundin, E. C. (2001). Attachment style in adjustment to conjugal bereave-
ment. Journal of Social and Personal Relationships, 18, 347–361.
Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitive–behavior
therapy for PTSD. New York: Guilford Press.
Fraley, R., & Bonanno, G.  A. (2004). Attachment and loss: A test of three competing
models on the association between attachment-related avoidance and adaptation to
bereavement. Personality and Social Psychology Bulletin, 30, 878–890.
Fraley, R. C., Garner, J. P., & Shaver, P. R. (2000). Adult attachment and the defensive
regulation of attention and memory: Examining the role of preemptive and postemptive
defensive processes. Journal of Personality and Social Psychology, 79, 816–826.
Fraley, R. C., & Shaver, P. R. (1997). Adult attachment and the suppression of unwanted
thoughts. Journal of Personality and Social Psychology, 73, 1080–1091.
Fraley, R. C., & Shaver, P. R. (2000). Adult romantic attachment: Theoretical developments,
emerging controversies, and unanswered questions. Review of General Psychology, 4,
132–154.
Gillath, O., Bunge, S.  A., Shaver, P.  R., Wendelken, C., & Mikulincer, M. (2005).
Attachment-style differences in the ability to suppress negative thoughts: Exploring the
neural correlates. Neuroimage, 28, 835–847.
Hazan, C., & Zeifman, D. (1999). Pair-bonds as attachments: Evaluating the evidence.
In Cassidy, J., & Shaver, P. R. (Eds.), Handbook of attachment: Theory, research, and
clinical applications (pp. 336–354). New York: Guilford.
Jerga, C., Shaver, P. R., & Wilkinson, R. B. (2011). Attachment insecurities and identifica-
tion of at-risk individuals following the death of a loved one. Journal of Social and
Personal Relationships, 28, 891–914.
Mikulincer, M., Dolev, T., & Shaver, P.  R. (2004). Attachment-related strategies during
thought-suppression: Ironic rebounds and vulnerable self-representations. Journal of
Personality and Social Psychology, 87, 940–956.
202  Mario Mikulincer and Phillip R. Shaver
Mikulincer, M., Hirschberger, G., Nachmias, O., & Gillath, O. (2001). The affective com-
ponent of the secure base schema: Affective priming with representations of attachment
security. Journal of Personality and Social Psychology, 81, 305–321.
Mikulincer, M., & Orbach, I. (1995). Attachment styles and repressive defensiveness: The
accessibility and architecture of affective memories. Journal of Personality and Social
Psychology, 68, 917–925.
Mikulincer, M., & Shaver, P. R. (2007a). Attachment in adulthood: Structure, dynamics,
and change. New York: Guilford Press.
Mikulincer, M., & Shaver, P. R. (2007b). Boosting attachment security to promote mental
health, prosocial values, and inter-group tolerance. Psychological Inquiry, 18, 139–156.
Mikulincer, M., Shaver, P. R., & Horesh, N. (2006). Attachment bases of emotion regula-
tion and posttraumatic adjustment. In Snyder, D. K., Simpson, J. A., & Hughes, J. N.
(Eds.), Emotion regulation in families: Pathways to dysfunction and health (pp. 77–99).
Washington, DC: American Psychological Association.
Nager, E. A., & de Vries, B. (2004). Memorializing on the World Wide Web: Patterns of
grief and attachment in adult daughters of deceased mothers. Omega, 49, 43–56.
Parkes, C.  M. (2006). Love and loss: The roots of grief and its complications. London:
Routledge.
Parkes, C. M., & Weiss, R. S. (1983). Recovery from bereavement. New York: Basic Books.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K.,
et al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed
for DSM-V and ICD-11. PLoS Med, 6(8), 20. Retrieved February 3. 2010, from http://
www.plosmedicine.org/article/info:doi%2F10.1371%2Fjournal.pmed.1000121.
Roisman, G. I., Tsai, J. L., & Chiang, K. H. S. (2004). The emotional integration of child-
hood experience: Physiological, facial expressive, and self-reported emotional response
during the Adult Attachment Interview. Developmental Psychology, 40, 776–789.
Rubin, S. S. (1991). Adult child loss and the two-track model of bereavement. Omega, 24,
183–202.
Shaver, P.  R., & Fraley, R.  C. (2008). Attachment, loss, and grief: Bowlby’s views and
current controversies. In Cassidy, J., & Shaver, P. R. (Eds.), Handbook of attachment:
Theory, research, and clinical applications (2nd edn., pp. 48–77). New York: Guilford
Press.
Shaver, P.  R., Schachner, D. A., & Mikulincer, M. (2005). Attachment style, excessive
reassurance seeking, relationship processes, and depression. Personality and Social
Psychology Bulletin, 31, 1–17.
Silverman, G. K., Johnson, J. G., & Prigerson, H. G. (2001). Preliminary explorations of
the effects of prior trauma and loss on risk for psychiatric disorders in recently widowed
people. Israel Journal of Psychiatry and Related Sciences, 38, 202–215.
Stroebe, M., Hansson, R. O., Stroebe, W., & Schut, H. A. W. (2001) (Eds.) Handbook of
bereavement research: Consequences, coping, and care. Washington, DC: American
Psychological Association.
Stroebe, M., & Schut, H. A. W. (1999). The dual process model of coping with bereave-
ment: Rationale and description. Death Studies, 23, 1–28.
Stroebe, M., Schut. H. A. W., & Stroebe, W. (2005). Attachment in coping with bereave-
ment: A theoretical integration. Review of General Psychology, 9, 48–66.
Vanderwerker, L. C., Jacobs, S. C., Parkes, C. M., & Prigerson, H. G. (2006). An explora-
tion of associations between separation anxiety in childhood and complicated grief in
later-life. Journal of Nervous and Mental Diseases, 194, 121–123.
 Attachment insecurities and disordered grief  203
Waskowic, T. D., & Chartier, B. M. (2003). Attachment and the experience of grief follow-
ing the loss of a spouse. Omega, 47, 77–91.
Wayment, H. A., & Vierthaler, J. (2002). Attachment style and bereavement reactions.
Journal of Loss and Trauma, 7, 129–149.
Weiss, R. S. (1991). The attachment bond in childhood and adulthood. In Parkes, C. M.,
Stevenson-Hinde, J., & Marris, P. (Eds.), Attachment across the life cycle (pp. 66–76).
London: Tavistock.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., van der
Heijden, P. G., & Dijkstra, I. (2007). Neuroticism and attachment insecurity as predic-
tors of bereavement outcome. Journal of Research in Personality, 41, 498–505.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., van der
Heijden, P. G., & Dijkstra, I. (2007). Patterns of attachment and parents’ adjustment
to the death of their child. Personality and Social Psychology Bulletin, 33, 537–548.
Zuroff, D. C., & Blatt, S. J. (2006). The therapeutic relationship in the brief treatment of
depression: Contributions to clinical improvement and enhanced adaptive capacities.
Journal of Consulting and Clinical Psychology, 74, 199–206.
15 Physiological mechanisms and the
neurobiology of complicated grief
Mary-Frances O’Connor

Complicated grief (CG) is a disorder of significant impact, as described in other

chapters of this book. In addition to the mental anguish accompanying this dis-
order, new evidence suggests that concurrent physiological changes occur and
these could impact physical health. This chapter will begin by reviewing selec-
tive theories that incorporate physiological aspects of general bereavement (i.e.,
attachment theory and physiological co-regulation), and will then review theories
that incorporate physiological changes in CG specifically. Next, the empirical
evidence for the health effects of CG and physiological mechanisms are reviewed.
A discussion of the advantages of studying the physiology and neurobiology of
CG follows. Health outcomes are one piece of evidence of physiological mecha-
nisms in CG. However, the use of physiology in research designs to distinguish
aspects of psychological phenomena will be discussed below as a useful reason
to include physiology in research studies of CG. For example, the underlying
aspects of the body’s stress response hint at distinctions between CG and post-
traumatic stress disorder (PTSD), or simply CG and non-complicated grief (non-
CG). Finally, the chapter ends with a summary of the research that has begun,
and, importantly, demonstrates the gaps in knowledge of the neurobiological and
physiological aspect of CG.
In an attempt to be clear regarding the operationalization of CG, each of the
studies reviewed has included a description of how the investigators measured
CG. In addition, the studies reviewed have not included participants with comor-
bid major depression, unless explicitly stated.

Models and theories of general bereavement

In addition to documenting empirical evidence of a physiological component to
CG, developing theories about why there would be a physiological component
to CG will assist the field in moving forward. These theories will inform what
hypotheses are developed and tested, and ultimately may lead to an understanding
of how we may improve treatment for CG. This section will review attachment
theory, and the consequences of separation: a disorganized protest response to
the withdrawal of the attachment figure, and an organized physiological stress
response.
 Physiological mechanisms of CG  205
One theory that general bereavement research has found very useful is attach-
ment theory (see Chapter 14). This theory states that the bonds between parent
and child, and between romantic partners, are a product of behavioral condition-
ing whereby an association is developed between the attachment figure and (1) a
reduction in distress and (2) the generation of pleasure (Bowlby, 1980). This con-
ditioning explains a variety of behaviors, such as the maintenance of close proxim-
ity between bonded individuals, the development of mental schemas, or working
models, that provide comfort during absence of the attachment figure, and distress
that is generated upon separation from the attachment figure (for a review of these
behaviors in adults, see Hazan, Gur-Yaish, & Campa, 2004). Attachment theory
has specific predictions for bereavement: the process of bereavement includes a
gradual extinction of this conditioning, in which the regulatory benefits conferred
by mental representations of the attachment figure diminish slowly over time.
Bowlby (1980) described the end point of successful mourning as a psychological
reorganization of one’s thoughts and feelings about a deceased attachment figure
(for review, see Sbarra & Hazan, 2008).
Added to the original attachment theory is an emphasis on the role of the
attachment figure in physiological, as well as psychological, regulation. Thus,
repeated social contact with a particular person results in a conditioned response
whereby this attachment figure is reliably associated with a state of psychological
security and physiological calm (Depue & Morrone-Strupinsky, 2005). Much of
the original work on physiological co-regulation came from a series of studies by
Myron Hofer (1984). These studies were designed to isolate different aspects of
what exactly was lost when a rat pup was separated from its mother. For example,
warmth and milk are two very different aspects of the loss. Separate experiments
were conducted to test the impact of the different aspects that were lost, and to
determine which physiological systems were impacted in the pup by the loss of
the individual aspects. Hofer theorized that the diverse responses to loss could be
understood in terms of the removal of “maternal regulators” (p. 12), which were
physiological. He drew analogies to human separation and bereavement, indicat-
ing that human bereavement also included the loss of physiological regulators,
rather than only psychological stress.
Sbarra and Hazan (2008) theorized that the response to separation (or bereave-
ment) in fact has two unrelated (though usually co-occurring) physiological
components. First, there is a general stress response (termed organized by Sbarra
and Hazan). Second, there is an attachment-specific stress response (termed dis-
organized by Sbarra and Hazan) driven by the loss of the rewarding aspects of
attachment. These two aspects of response will be discussed, followed by their
relationship to CG.

General stress response

First, bereavement provokes a general stress response (termed organized by
Sbarra and Hazan). This is the physiological stress response that psychologists
refer to as the “fight-or-flight” response. When an event triggers a fear reaction,
206  Mary-Frances O’Connor
accompanying changes allow the person to be physically prepared to fight or flee
from the thing that has triggered the fear. This reaction, however, occurs in a wide
range of circumstances that feel threatening, even when there is not a clear “thing”
that can be fought or fled. The physical reactions happen in physiological systems
that allow for increased blood flow to muscles, increased attention to the environ-
ment, and increased metabolism. The systems include the cardiovascular system,
which shows increased heart rate and increased chemicals used by the cardiovas-
cular system, called catecholamines (e.g., epinephrine). The systems also include
the hypothalamic–pituitary–adrenal (HPA) axis, from which the stress hormone
cortisol is made. Cortisol helps to metabolize food quickly into blood sugar, to
provide increased energy to the body.
General bereavement research has demonstrated increases in catecholamines
and cortisol in the early stages of bereavement (Gerra et al., 2003; Hofer, Wolff,
Friedman, & Mason, 1972; Jacobs et al., 1987; Jacobs, Mason, et al., 1986), and
higher heart rate and blood pressure in the first 2 weeks after the loss (Buckley et
al., 2011). However, this general physiological stress response to bereavement is
not distinct from the response to other stressful life events (e.g., stress of job loss,
stress associated with man-made disasters).

Attachment-specific stress response

In addition to the general stress response, there is an attachment-specific stress
response (termed disorganized by Sbarra and Hazan) driven by the loss of the
rewarding aspects of attachment (Insel, 2003; Kovacs, Sarnyai, & Szabo, 1998;
Panksepp, Knutson, & Burgdorf, 2002). Physiological systems respond to the
removal of the conditioned pleasure and soothing associated with the attach-
ment figure. Sbarra and Hazan (2008) use the term co-regulation to describe
the physiological aspect of the feelings of security that an attachment figure
provides.
The physiological systems responsible for this attachment-specific stress
response include the dopamine system (Depue & Morrone-Strupinsky, 2005),
the opioid system (Nelson & Panksepp, 1998; Panksepp, Nelson, & Bekkedal,
1997), and the oxytocin system (Lim & Young, 2006; Young & Wang, 2004).
The dopamine system is important in the experience of motivation to seek our
rewards, both wanting and, quite literally, moving toward a desired object.
Dopamine is one of the neurotransmitters that is fundamental in conditioning, in
associating the experience of reward with specific objects (Berridge, Robinson,
& Aldridge, 2009). In the present discussion, this conditioning specifically cre-
ates the attachment to a particular figure. Dopamine is important in pursuing
rewards, and opioids are important in the enjoyment of those rewards (Depue &
Morrone-Strupinsky, 2005). Opioids are another group of neurochemicals made
endogenously by the body, and they are also released under a variety of social
interactions, including gentle physical touch. Oxytocin is a neurohormone impor-
tant in birthing and nursing in all mammals, but in humans it has also been linked
to suppression of anxiety during psychosocial stress and to the enhancement of
 Physiological mechanisms of CG  207
trust (Heinrichs, Baumgartner, Kirschbaum, & Ehlert, 2003; Kosfeld, Heinrichs,
Zak, Fischbacher, & Fehr, 2005).
Data about the effect of bereavement or separation on these attachment-specific
systems are primarily from rodent and primate research, although empirical data
in human bereavement have documented an increase in β-endorphins (one type of
opioids) (Gerra et al., 2003) in the first weeks after the loss.

Models and theories of CG

It is important to pause at this point and reflect on the fact that the models and
theories described so far apply to bereavement generally. These are not models
that specifically explain why some individuals develop CG in response to the
death of a loved one and others adjust resiliently.
A biopsychosocial model of CG has been put forward by Shear and Shair
(2005). They posit first that the symptoms of acute grief result from a temporary
failure of biobehavioral regulatory functions resulting from the mental representa-
tion of the deceased person, much like what has been described above. In most
cases, this acute grief resolves as the bereaved person assimilates the finality
of the loss and this knowledge is integrated into attachment-related long-term
memory and mental schemas. This allows an effective attachment system to func-
tion again, and there is a reduction of overwhelming and intense sadness.
Although acute grief is usually followed by resilient adjustment (Bonanno
et al., 2002), Shear and Shair (2005) suggest that adjustment to the death may
become complicated by maladaptive attitudes and behaviors (and, this author
adds, perhaps physiological constraints of the neurobiological attachment
system). In Shear and Shair’s model, based on prior work by Myron Hofer and
Mardi Horowitz:

CG is viewed as a stress response syndrome that results from failure to

integrate information about death of an attachment figure into an effectively
functioning secure base schema and/or inability to effectively re-engage
the exploratory system in a world without the deceased. (Shear et al., 2007,
p. 453)

The difficulty with creating a neurobiological model of CG is that it is cur-

rently unknown whether, at a basic level, CG represents merely a person with
acute grief whose process of adaptation has been interrupted, or a wholly other
process than non-complicated adaptation. For example, CG may stem from a pre-
existing individual difference, which is already present at the time of the death
of the attachment figure. However, it may require the removal of the attachment
figure for this pre-existing condition to be revealed in behavior.
The two portions of Shear and Shair’s model of CG may be roughly mapped
onto the two types of physiological stress responses that have been discussed
above. The general stress response (e.g., the fight-or-flight response) may be seen
as similar to CG diagnostic criteria that include efforts to avoid thoughts of the
208  Mary-Frances O’Connor
deceased, feeling stunned, dazed, or shocked, and excessive irritability or anger
(Prigerson et al., 2009; Shear et al., 2011).1 The failure to integrate the reality of
the death may lead to a continuously realized acute grief, prolonging the acute
grief experience.
The attachment-specific stress response, driven by the loss of the rewarding
aspects of attachment, might be seen as analogous to CG diagnostic criteria that
include yearning for the deceased, feeling as if life is empty or meaningless with-
out the person who died, frequent preoccupying thoughts about the person who
died, and persistent difficulty trusting or caring about other people (Prigerson
et al., 2009; Shear et al., 2011). The ability to re-engage with the social world
following the death of an attachment figure may require the affiliative aspects
of attachment system to explore new relationships and turn to existing ones for
support.
In the next section, the empirical work supporting the role of physiology and
neurobiology in CG will be reviewed in detail.

Studies of the physiology and neurobiology of CG

Crying
Crying is a canonical sign of grief, and is clearly a physiological event. Crying is
also a very complex behavior, influenced by gender, culture, personality charac-
teristics, and the social environment (Romans & Clarkson, 2008). There is some
evidence that early in a crying episode there is high arousal in the cardiovascular
system, but that across the episode there is an increase in the regulation of system,
and heart rate slows again (Hendriks, Rottenberg, & Vingerhoets, 2007). Early
work that examined the difference in symptoms between CG and bereavement-
related depression found that crying was more strongly related to other CG symp-
toms than to other depression symptoms (Prigerson, Frank, et al., 1995).

Hypertension and heart rate

The most convincing early empirical evidence that there was a physiological
component to bereavement came from studies of morbidity and mortality asso-
ciated with bereavement (sometimes referred to as the “broken-heart phenom-
enon”; Stroebe, 1994). Work on that topic has been reviewed elsewhere (Buckley,
McKinley, Tofler, & Bartrop, 2009; Stroebe, Schut, & Stroebe, 2007), but newer
studies have developed this line of research by investigating the relationship
between morbidity and mortality and the subgroup of bereaved individuals who
have CG.
For instance, CG symptoms2 and blood pressure were measured in bereaved
participants at 6, 13, and 25 months (Prigerson et al., 1997). High blood pressure
was operationalized as systolic blood pressure over 140  mm Hg, and CG was
measured using the Grief Measurement Scale (Jacobs, Kasl, et al., 1986), modi-
fied to reflect only items from the 19-item Inventory of Complicated Grief (ICG;
 Physiological mechanisms of CG  209
Prigerson, Maciejewski, et al., 1995). CG symptoms at 6 months were associated
with increased risk of high blood pressure at 13 months. In addition, self-reported
incidence of cancer has been shown in two studies to be higher in those with CG
(Prigerson et al., 1997; Prigerson, Maciejewski, et al., 1995).
The mechanisms that account for these health changes have not yet been
robustly examined in CG,3 although there is a great deal of evidence regard-
ing changes in physiological mechanisms in general bereavement. A study by
Bonanno, Neria, Mancini, Coifman, Litz, and Insel (2007) examined the heart rate
changes associated with recalling events related to the loss (i.e., the participant’s
relationship to the deceased). This recall was done at approximately 4 months
post-death and again at approximately 18 months. CG symptoms were measured
using eight items (Horowitz et al., 1997). CG symptoms were only slightly and
non-significantly associated with baseline heart rate, but significantly associated
with decreased heart rate from 4 months to 18 months (Bonanno et al., 2007).
The authors interpret these data in light of both animal and human studies dem-
onstrating a withdrawal upon separation from attachment figures. In contrast to
the negative correlation of CG symptoms and heart rate, PTSD symptoms in this
study were seen to correlate positively with heart rate, such that higher PTSD
symptoms were associated with a higher baseline heart rate and an increased heart
rate across the two recall sessions.

Neurobiology
Several studies have now been conducted using functional magnetic resonance
imaging (fMRI) as a modality in bereavement research. However, most of them
have investigated general bereavement. If CG is a distinct phenomenon from non-
CG, there should be differences in the neurobiological correlates. To the author’s
knowledge, only one study has investigated group differences between bereaved
persons with CG and non-CG using fMRI (O’Connor et al., 2008).
Participants in this study were 23 women, each with a female family member
who had died of breast cancer in the prior 18 months. This group was chosen
because of the prevalence of CG among women with a family history of breast
cancer. These women have often lost multiple female family members, often
when the family member was still very young, and the survivors identify very
strongly with them, because of their own increased risk.
CG was diagnosed with a structured clinical interview (Prigerson & Jacobs,
2001), with a cut-off for CG extrapolated from an ICG of ≥  30 (Prigerson,
Maciejewski, et al., 1995). The participants included 11 women with CG and
12 women with non-CG. Exclusion criteria included Axis I psychiatric disorders
(including current depression) and medical disorders.
Participants provided a photograph of their deceased loved one, which was
matched with a photo of a stranger. Grief-related words were taken from an
interview about the death event (e.g., collapse, funeral, loss) and were matched
with neutral words (e.g., announce, ceiling, list). These words were embedded
into the photos to create composites. These picture–word composites resulted
210  Mary-Frances O’Connor
in a 2  ×  2 factorial design: (1) deceased + grief word, (2) stranger + grief word,
(3) deceased + neutral word, (4) stranger + neutral word. Sixty composites were
shown, in a manner comparable to a slide show. This task had previously been
validated in the scanner with participants’ skin conductance responses and subjec-
tive grief ratings in response to each slide (Gündel, O’Connor, Littrell, Fort, &
Lane, 2003).
Analyses revealed that, in response to reminders of the deceased, CG partici-
pants showed greater activation than those with non-CG in a subcortical area of
the brain called the nucleus accumbens. To be clear, this means not that this was
the only area of the brain that was active during the mental processing of the
picture–word composites, but that it was more active in those with CG than in
those with non-CG, above a statistical threshold of p  <  0.001 (uncorrected for
multiple comparisons).
Research on both animals and humans clearly demonstrates that the nucleus
accumbens is active during the processing of rewards. In this use of the term
reward, the reference is to the psychological construct of reward as a reinforcer
(i.e., as opposed to punishment), rather than a positive, experiential feeling of
reward. Reward can be decomposed into “wanting” and “liking,” and elegant
experimental designs have shown that the nucleus accumbens is activated when a
reward is “wanted” (Knutson, Adams, Fong, & Hommer, 2001). Quotation marks
are used around the terms here to distinguish the experiential aspect of want-
ing from the reinforcement value of “wanting” that is associated with nucleus
accumbens activation.
Additional analyses were conducted to explore the correlates of activation
in this region. Activation in the nucleus accumbens was not correlated with the
amount of time that had passed since the death event, the participant’s age, or
the self-reported positive/negative affect after the scan. The nucleus accumbens
activation was positively correlated with self-reported yearning at an interview
in the week prior to the scan (r = 0.42, p < 0.05). This result is understandable,
given that, when an object is reinforcing (i.e., it is “wanted”), there may also be a
yearning for that object.
It is also important to recognize that knowing that this region distinguishes
these two groups does not mean that the nucleus accumbens activation is causal
(i.e., we do not know that this region causes impaired adaptation during grief, or
whether it is a consequence of the symptoms of CG). It also does not tell us if the
region is related to individual differences, or if it is part of a network of activation
that changes across adaptation. In other words, one possibility is that those with
CG would show distinctive activation in this region as an individual difference –
perhaps even before the loss of a loved one. Alternatively, all individuals may
show greater activation in this region early in adaptation to a loved one’s death,
and decreasing activation in this region as they adapt psychologically. These two
alternatives require future research that scans each participant more than once
across time, in order to observe change during adaptation.
Finally, it is not possible to know from functional neuroimaging what neurons
in the nucleus accumbens region are the sources of this increased activation. For
 Physiological mechanisms of CG  211
example, this brain region is rich in oxytocin, opioid, and dopamine receptors, and
neurons that use one, two, or all three of these neurotransmitters may have been
more active in those with CG than in those with non-CG. Thus, future research
that investigates the levels of these neurotransmitters in the central nervous
system (or, in cases where there is a valid method, the behavioral output of these
central levels) might be a productive avenue of research in discriminating CG
from non-CG.
For those with CG, reminders of the deceased activated neural reward activity,
and this neural reinforcement may interfere with adapting to the loss in the present.
Alternatively, the nucleus accumbens activation may simply be a neurobiological
indicator of where the bereaved is in the adaptation process. Because activation
of this region is also seen in fMRI studies of adults viewing photos of their living
romantic partners and their children (Bartels & Zeki, 2004), it is reasonable to
hypothesize that those with CG are responding subcortically to the cue as though
the loved one were still alive.

Sleep
Sleep can be measured through polysomnography (PSG),4 a comprehensive
recording of the physiological changes that occur during sleep. These recordings
include brain waves, eye movements, and muscle, heart, and breathing activity.
PSG can identify when someone is sleeping, and also what type of sleep they are
experiencing (e.g., deep sleep, rapid eye movement [REM] sleep associated with
dreaming).
In the first study of sleep in CG, symptoms were measured with 13 of the
19 items on the ICG, because the study was begun prior to the publication of
the ICG and not all information was available for early participants (Prigerson,
Maciejewski, et al., 1995). Symptoms of CG were not associated with main
effects on PSG measures although mild subjective sleep impairment was reported.
In general, sleep continuity measures in subjects with CG alone were similar to
data from non-bereaved healthy control subjects previously published.
Because CG symptoms showed no main effects on PSG sleep measures,
it appears that CG symptoms do not entail the changes in sleep physiology
seen in depression (for a review of the relationship between REM sleep and
depression, see Berger & Riemann, 1993). The study did show that CG in
combination with comorbid depression was associated with a higher percentage
of REM sleep; however, two subjects primarily drove this effect. The authors
conclude in their discussion that, based on their physiological data, CG should
not be considered a form of depressive reaction to bereavement (McDermott
et al., 1997).
In a second study of the physiology of sleep and circadian rhythms of widowed
individuals (Monk et al., 2008), participants were assessed with the ICG, with
CG caseness defined as an ICG score of 25 or greater at 6 months (Prigerson,
Maciejewski, et al., 1995). Out of 18 individuals who were bereaved longer than
6 months, only four met criteria for CG. Sleep was measured through PSG in this
212  Mary-Frances O’Connor
study as well. The four CG participants were only approximately 0.5 standard
deviation worse in sleep duration (total minutes of PSG-measured sleep), sleep
efficiency (percentage of the night actually spent asleep), and sleep latency (the
number of minutes between lights out and sleep onset) than the whole sample
mean.
Another measure that was included in this study was core body temperature.
This is a physiological system with a very strong circadian rhythm. Early work by
Hofer and colleagues demonstrated that, when rat pups were separated from their
mother, they showed overactive or depressed behavior depending on whether they
were kept at the same temperature as when they were with her, or allowed to cool
without her body heat, respectively (Hofer, 1994). This was one piece of evidence
used to support the physiological co-regulation basis of grief.
Monk and colleagues recorded core body temperature continuously, around
the clock. Core body temperature normally is lowest at 1 a.m., with a steep rise
through the morning hours and a more gradual rise to mid-evening. In the pattern
of core body temperature, those with CG showed a shift of half an hour earlier in
circadian temperature rhythm phase. A temperature rhythm that is shifted forward
can lead to early awakening. Those with CG were also 1 standard deviation worse
in circadian temperature rhythm amplitude (i.e., their temperature did not have as
large a rise and fall across the day). No information was provided whether the four
CG participants differed in age from the other participants, which is important
information, as these same changes can be seen in normal aging.
In this second study, no differences were reported for those with CG related to
the percentage of sleep spent in REM, even though this variable was measured.
This suggests that this sample, albeit extremely small, is a replication of the first
study findings that those with CG do not show the same pattern of increased REM
sleep found in depression. Clearly more data need to be collected comparing CG
and depression to make a definitive statement on the role of REM sleep in these
disorders.

Genetics
How might genetics affect symptoms of CG? In the field of genetics, this para-
digm is referred to as a gene-by-environment interaction, abbreviated as G×E. In
this case, bereavement is the environmental portion of the interaction, and likely
genes are investigated to determine whether a genetic portion of the interaction
exists. In other words the question is posed: What genetic vulnerabilities more
often lead to CG after bereavement?
Only one study has currently been published on genetics and CG. The genetic
variation that has been investigated is the gene region that contains the code
for monoamine oxidase-A (MAO-A). This genetic variation functions to make
more or less of the enzyme MAO-A, depending on which variation the person
carries. MAO-A breaks down molecules that are familiar in psychopathology,
including serotonin, dopamine, epinephrine, and norepinephrine. Thus, some
people make more MAO-A and consequently they have less serotonin and
 Physiological mechanisms of CG  213
dopamine (because it is broken down more quickly and therefore less available).
Kersting and colleagues (2007) hypothesized that those who have the genetic
variation that causes elevated MAO-A activity could have an increased vulner-
ability to CG.
CG is a risk factor for major depression (Boelen & Prigerson, 2007), so the
study by Kersting and colleagues was done with psychiatric inpatients who were
diagnosed with major depression and had a history of bereavement. To measure
CG, the investigators used the ICG (Prigerson, Maciejewski, et al., 1995), cat-
egorizing those scoring 25 or above as having CG. The more active variant of the
MAO-A gene was significantly associated with CG in women, whereas there was
no such effect in male patients. This result means that, among depressed women,
those who had the more active genetic variation and who were bereaved (i.e.,
G×E) were more likely to have CG. The effects of this MAO-A variation have
been shown in women specifically in other psychiatric disorders, such as panic
disorder and obsessive–compulsive disorder (Camarena, Cruz, de la Fuente, &
Nicolini, 1998; Deckert et al., 1999).

Why study the physiology and neurobiology of CG?

As seen in this review, it is useful to consider including neurobiological and
physiological markers in study designs when investigating CG, as these mark-
ers can sometimes be used to discriminate between disorders in a useful way
(e.g., depression vs. CG, PTSD vs. CG), even when a clinical gestalt may be
murky. Because biomarkers show promise in distinguishing CG from non-CG, it
would also be useful to include them as outcome measures in treatment studies,
as a multi-method way of operationalizing outcome. As with biomarkers in most
affective disorders, none are yet ready to be used in a clinical setting to aid in
diagnosis.
In addition, a common misconception is that the only way that a physiological
understanding of CG would be used is through pharmacotherapy. However, an
understanding of the physiological mechanisms in this disorder can be used in
a host of ways. This includes understanding the ways in which production of
neurotransmitters and activation of the body’s systems are created endogenously
(e.g., through warm physical contact, meditation, cognitive–behavioral therapy,
rumination, suppression).
The stress of bereavement is created not just by the circumstance per se, but
also by the individual’s response to it. The individual’s response is a psychological
one and, concurrently, a physiological one. This stress response may specifically
include CG symptoms. To the degree that we discover whether these symptoms
are mediated by attachment (Langner & Maercker, 2005), then understanding the
physiology and neurobiology of attachment will no doubt assist in treating the CG
response to bereavement. Observing and documenting the physiological response
to bereavement, and how it shapes and is shaped by the psychological response,
may help us to improve adaptation even in the face of one of life’s most stressful
events.
214  Mary-Frances O’Connor

Summary and future directions

In summary, the study of the physiology and neurobiology of CG is only at the
earliest beginning, with studies nibbling around the edges of the disorder by
focusing on individual variables. Self-regulation, at the psychological as well as
physiological levels, may be important in coping with pangs of grief and assist
in acceptance of the death of an attachment figure. Human physiology and neu-
robiology embody the reactivity and recovery in oscillatory waves of grief. One
hypothesis is that increases in certain neurochemicals (e.g., catecholamines, oxy-
tocin, and cortisol) following the death of a loved one may make it more likely
that we reach out to other attachment figures, and make the strengthening of bonds
with the living more likely during the mourning period. Another hypothesis is that
a failure to mount this physiological response to the death of an attachment figure
could prevent the normal social exploratory behaviors, an avoidance symptom
often seen in CG.
In addition, the withdrawal of interpersonal regulators must be accommodated
following the death of an attachment figure. For some, this withdrawal leads to
changes in sleep, eating, concentration, and enjoyment of activities. Preliminary
evidence suggests that the physiology underlying these changes is dissimilar
between disorders (e.g., increased REM sleep in depression, but not in CG, and
increased heart rate across time in PTSD, but not in CG). Careful comparison of
these disorders is necessary to determine whether physiological markers can be
useful in diagnosis, and then in differential treatment for them.
The assimilation of the reality of the death must occur in the brain for the
working model of attachment to be revised. Yet another hypothesis is that if the
assimilation of the new information does not occur, for psychological reasons
(e.g., extreme guilt or avoidance) and/or biological ones (e.g., increased degrada-
tion of bioamines), then the adaptation to the death may be prolonged and lead
to CG.
Some physiological markers of CG will correlate with a separation distress
response and others will correlate with a general stress response. The physiologi-
cal markers that correlate with a general stress response may occur with other
stressful life events, but the physiological markers that correlate with the separa-
tion distress should be specific to the loss of an attachment figure. In addition, the
physiological markers correlated specifically to the loss of an attachment figure
may be pre-existing traits (endemic to the individual or to the relationship), or
these physiological markers may develop, or fail to recover, across time during
the adaptation process.
Most affective disorders (e.g., major depressive disorder) are better described
on a continuum than as discrete categories (Prisciandaro & Roberts, 2009) and
CG may well turn out to be similar in this regard. The relationship between the
clinical gestalt of CG and individual biomarkers is complex. It is highly unlikely
that there is a one-to-one correspondence between any particular physiological
or neurobiological marker and CG. For one thing, physiological systems are
multiply intertwined, and feed back information to each other, and therefore any
 Physiological mechanisms of CG  215
biomarker impacts a host of other biomarkers. However, by measuring these
markers, we may see what contributes to poor adaptation or what the physiologi-
cal predictors of CG are. Using physiological and neurobiological variables in
bereavement research as one part of a multi-method approach will only increase
our understanding of the phenomena.

Notes
1 The disorder is referred to as prolonged grief disorder in the Prigerson et al. (2009)
paper.
2 The disorder is referred to as traumatic grief in the Prigerson et al. (1997) paper.
3 Not reviewed here are self-reported physical health symptoms measured in some
studies and found to be associated with CG symptoms or caseness.
4 Additional studies examining self-report measures of sleep in persons with CG
also have been published, but the present chapter focuses solely on physiological
measurements.

References
Bartels, A., & Zeki, S. (2004). The neural correlates of maternal and romantic love.
Neuroimage, 21, 1155–1166.
Berger, M., & Riemann, D. (1993). Symposium: Normal and abnormal REM sleep regula-
tion: REM sleep in depression – an overview. Journal of Sleep Research, 2, 211–223.
Berridge, K.  C., Robinson, T.  E., & Aldridge, J.  W. (2009). Dissecting components of
reward: “Liking”, “wanting”, and learning. Current Opinion in Pharmacology, 9,
65–73.
Boelen, P. A., & Prigerson, H. G. (2007). The influence of symptoms of prolonged grief
disorder, depression, and anxiety on quality of life among bereaved adults: A prospec-
tive study. European Archives of Psychiatry in Clinical Neuroscience, 257, 444–452.
Bonanno, G. A., Neria, Y., Mancini, A., Coifman, K. G., Litz, B., & Insel, B. (2007). Is
there more to complicated grief than depression and posttraumatic stress disorder? A
test of incremental validity. Journal of Abnormal Psychology, 116, 342–351.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring, M., Sonnega, J.,
et al. (2002). Resilience to loss and chronic grief: A prospective study from preloss to
18-months postloss. Journal of Personality & Social Psychology, 83, 1150–1164.
Bowlby, J. (1980). Attachment and loss, vol. 3: Loss, sadness and depression. New York:
Basic Books.
Buckley, T., McKinley, S., Tofler, G., & Bartrop, R. (2009). Cardiovascular risk in early
bereavement: A literature review and proposed mechanisms. International Journal of
Nursing Studies, 47, 229–238.
Buckley, T., Mihailidou, A. S., Bartrop, R., McKinley, S., Ward, C., Morel-Kopp, M. C., et
al. (2011). Haemodynamic changes during early bereavement: potential contribution to
increased cardiovascular risk. Heart, Lung and Circulation, 20, 91–98.
Camarena, B., Cruz, C., de la Fuente, J. R., & Nicolini, H. (1998). A higher frequency of
a low activity-related allele of the MAO-A gene in females with obsessive–compulsive
disorder. Psychiatry and Genetics, 8, 255–257.
Deckert, J., Catalano, M., Syagailo, Y. V., Bosi, M., Okladnova, O., Di Bella, D., et al.
(1999). Excess of high activity monoamine oxidase A gene promoter alleles in female
patients with panic disorder. Human Molecular Genetics, 8, 621–624.
216  Mary-Frances O’Connor
Depue, R. A., & Morrone-Strupinsky, J. V. (2005). A neurobehavioral model of affiliative
bonding: Implications for conceptualizing a human trait of affiliation. Behavior and
Brain Science, 28, 313–350; discussion 350–395.
Gerra, G., Monti, D., Panerai, A. E., Sacerdote, P., Anderlini, R., Avanzini, P., et al. (2003).
Long-term immune-endocrine effects of bereavement: Relationships with anxiety
levels and mood. Psychiatry Research, 121, 145–158.
Gündel, H., O’Connor, M.-F., Littrell, L., Fort, C., & Lane, R. D. (2003). Functional neu-
roanatomy of grief: An fMRI study. American Journal of Psychiatry, 160, 1946–1953.
Hazan, C., Gur-Yaish, N., & Campa, M. (2004). What does it mean to be attached? In
Rholes, W. S., & Simpson, J. A. (Eds.), Adult attachment: New directions and emerging
issues (pp. 55–85). New York: Guilford Press.
Heinrichs, M., Baumgartner, T., Kirschbaum, C., & Ehlert, U. (2003). Social support and
oxytocin interact to suppress cortisol and subjective responses to psychosocial stress.
Biological Psychiatry, 54, 1389–1398.
Hendriks, M. C., Rottenberg, J., & Vingerhoets, A. J. (2007). Can the distress-signal and
arousal-reduction views of crying be reconciled? Evidence from the cardiovascular
system. Emotion, 7, 458–463.
Hofer, M. A. (1984). Relationships as regulators: A psychobiologic perspective on bereave-
ment. Psychosomatic Medicine, 46, 183–197.
Hofer, M. A. (1994). Early relationships as regulators of infant physiology and behavior.
Acta Paediatrica Supplement, 397, 9–18.
Hofer, M. A., Wolff, C. T., Friedman, S. B., & Mason, J. W. (1972). A psychoendocrine
study of bereavement. I. 17-Hydroxycorticosteroid excretion rates of parents following
death of their children from leukemia. Psychosomatic Medicine, 34, 481–491.
Horowitz, M.  J., Siegel, B., Holen, A., Bonanno, G. A., Milbrath, C., & Stinson, C.  H.
(1997). Diagnostic criteria for complicated grief disorder. American Journal of
Psychiatry, 154, 904–910.
Insel, T. R. (2003). Is social attachment an addictive disorder? Physiology & Behavior, 79,
351–357.
Jacobs, S. C., Kasl, S. V., Ostfeld, A. M., Berkman, L., Kosten, T. R., & Charpentier, P.
(1986). The measurement of grief: Bereaved versus non-bereaved. Hospice Journal, 2,
21–36.
Jacobs, S. C., Mason, J., Kosten, T. R., Kasl, S. V., Ostfeld, A. M., & Wahby, V. (1987).
Urinary free cortisol and separation anxiety early in the course of bereavement and
threatened loss. Biological Psychiatry, 22, 148–152.
Jacobs, S. C., Mason, J. W., Kosten, T. R., Wahby, V., Kasl, S. V., & Ostfeld, A. M. (1986).
Bereavement and catecholamines. Journal of Psychosomatic Research, 30, 489–496.
Kersting, A., Kroker, K., Horstmann, J., Baune, B. T., Hohoff, C., Mortensen, L. S., et al.
(2007). Association of MAO-A variant with complicated grief in major depression.
Neuropsychobiology, 56, 191–196.
Knutson, B., Adams, C. M., Fong, G. W., & Hommer, D. (2001). Anticipation of increasing
monetary reward selectively recruits nucleus accumbens. Journal of Neuroscience, 21,
RC159.
Kosfeld, M., Heinrichs, M., Zak, P.  J., Fischbacher, U., & Fehr, E. (2005). Oxytocin
increases trust in humans. Nature, 435, 673–676.
Kovacs, G.  L., Sarnyai, Z., & Szabo, G. (1998). Oxytocin and addiction: A review.
Psychoneuroendocrinology, 23, 945–962.
Langner, R., & Maercker, A. (2005). Complicated grief as a stress response disorder:
Evaluating diagnostic criteria in a German sample. Journal of Psychosomatic Research,
58, 235–242.
 Physiological mechanisms of CG  217
Lim, M. M., & Young, L. J. (2006). Neuropeptidergic regulation of affiliative behavior and
social bonding in animals. Hormones and Behavior, 50, 506–517.
McDermott, O. D., Prigerson, H. G., Reynolds, C. F. III, Houck, P. R., Dew, M. A., Hall,
M., et al. (1997). Sleep in the wake of complicated grief symptoms: An exploratory
study. Biological Psychiatry, 41, 710–716.
Monk, T. H., Begley, A. E., Billy, B. D., Fletcher, M. E., Germain, A., Mazumdar, S., et
al. (2008). Sleep and circadian rhythms in spousally bereaved seniors. Chronobiology
International, 25, 83–98.
Nelson, E.  E., & Panksepp, J. (1998). Brain substrates of infant–mother attach-
ment: Contributions of opioids, oxytocin, and norepinephrine. Neuroscience and
Biobehavioral Review, 22, 437–452.
O’Connor, M.  F., Wellisch, D.  K., Stanton, A.  L., Eisenberger, N.  I., Irwin, M.  R., &
Lieberman, M. D. (2008). Craving love? Enduring grief activates brain’s reward center.
NeuroImage, 42, 969–972.
Panksepp, J., Knutson, B., & Burgdorf, J. (2002). The role of brain emotional systems
in addictions: A neuro-evolutionary perspective and new ‘self-report’ animal model.
Addiction, 97, 459–469.
Panksepp, J., Nelson, E., & Bekkedal, M. (1997). Brain systems for the mediation of social
separation-distress and social-reward: Evolutionary antecedents and neuropeptide
intermediaries. Annals of the New York Academy of Sciences, 807, 78–100.
Prigerson, H. G., Bierhals, A. J., Kasl, S. V., Reynolds, C. F. III, Shear, M. K., Day, N., et
al. (1997). Traumatic grief as a risk factor for mental and physical morbidity. American
Journal of Psychiatry, 154, 616–623.
Prigerson, H. G., Frank, E., Kasl, S. V., Reynolds, C. F., III, Anderson, B., Zubenko, G. S.,
et al. (1995). Complicated grief and bereavement-related depression as distinct disor-
ders: Preliminary empirical validation in elderly bereaved spouses. American Journal
of Psychiatry, 152, 22–30.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine, 6, e1000121.
Prigerson, H. G., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale,
consensus criteria, and preliminary empirical test. In Stroebe M. S., Hansson, R. O.,
Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research: Consequences,
coping and care (pp. 613–645). Washington, DC: American Psychological Association.
Prigerson, H.  G., Maciejewski, P.  K., Reynolds, C.  F., Bierhals, A.  J., Newsom, J.  T.,
Fasiczka, A., et al. (1995). Inventory of Complicated Grief: A scale to measure mal-
adaptive symptoms of loss. Psychiatry Research, 59, 65–79.
Prisciandaro, J.  J., & Roberts, J.  E. (2009). A comparison of the predictive abilities of
dimensional and categorical models of unipolar depression in the National Comorbidity
Survey. Psychological Medicine, 39, 1087–1096.
Romans, S. E., & Clarkson, R. F. (2008). Crying as a gendered indicator of depression.
Journal of Nervous and Mental Disease, 196, 237–243.
Sbarra, D. A., & Hazan, C. (2008). Coregulation, dysregulation, self-regulation: An inte-
grative analysis and empirical agenda for understanding adult attachment, separation,
loss, and recovery. Personality and Social Psychology Review, 12, 141–167.
Shear, M. K., Monk, T., Houck, P., Melhem, N., Frank, E., Reynolds, C., et al. (2007). An
attachment-based model of complicated grief including the role of avoidance. European
Archives of Psychiatry and Clinical Neuroscience, 257, 453–461.
Shear, M. K., & Shair, H. (2005). Attachment, loss, and complicated grief. Developmental
Psychobiology, 47, 253–267.
218  Mary-Frances O’Connor
Shear, M.  K., Simon, N., Wall, M., Zisook, S., Neimeyer, R., Duan, N., et al. (2011).
Complicated grief and related bereavement issues for DSM-5. Depression and Anxiety,
28, 103–117.
Stroebe, M. (1994). The broken heart phenomenon: An examination of the mortality of
bereavement. Journal of Community & Applied Social Psychology, 4, 47–61.
Stroebe, M., Schut, H., & Stroebe, W. (2007). Health outcomes of bereavement. Lancet,
370, 1960–1973.
Young, L. J., & Wang, Z. (2004). The neurobiology of pair bonding. Nature Neuroscience,
7, 1048–1054.
Part V

Treatment of complicated
grief
Principles, paradigms, and
procedures
16 Prolonged grief disorder
Cognitive–behavioral theory and
therapy
Paul A. Boelen, Marcel van den Hout, and
Jan van den Bout

Introduction
For a minority of people, the death of a loved one precipitates the development
of prolonged grief disorder (PGD), or complicated grief (CG) as it has also been
named (e.g., Prigerson et al., 2009). PGD has been defined as a clinical con-
dition that encompasses specific grief reactions (including separation distress
and difficulties accepting the loss and moving on without the lost person) that
cause significant distress and disability at least 6 months after the death occurred.
Symptoms of PGD are distinct from normal grief, bereavement-related depres-
sion, posttraumatic stress disorder (PTSD), and other anxiety symptoms and syn-
dromes, and, if left untreated, associated with significant impairments in health
and quality of life (Prigerson et al., 2009; see also Chapter 7 in this volume).
Parallel to the growing recognition of PGD as a distinct disorder, increasing
attention is being given to its treatment, with the most well-tested and effective
recent therapies being based in cognitive–behavioral therapy (CBT). For instance,
Shear, Frank, Houck, and Reynolds (2005) found “complicated grief treatment” –
a 16-session treatment containing elements CBT such as imaginal exercises to
revisit the death (exposure) and working toward achievement of personal goals –
to be effective in reducing PGD. Wagner, Knaevelsrud, and Maercker (2006)
found Internet-based CBT, including elements of confronting the loss, cognitive
restructuring, and social sharing, to be effective in reducing PGD and concomitant
depressive and anxious symptoms. We examined the effectiveness of a 12-session
CBT for PGD, based on a cognitive–behavioral conceptualization of the condition
(Boelen, de Keijser, van den Hout, & van den Bout, 2007). Compared with sup-
portive counseling, CBT was considerably more effective in ameliorating PGD
symptoms.
The present chapter addresses the theory and treatment of PGD from the per-
spective of cognitive–behavioral theorizing, with a focus on our own theoretical
approach (Boelen, van den Hout, & van den Bout, 2006). Notably, this approach
bears resemblance to other recent theoretical approaches to PGD (Shear, Boelen,
& Neimeyer, 2011). Specifically, this chapter will (1) introduce a theoretical basis
of CBT for PGD, (2) describe key interventions included in this approach, and (3)
review research on its theoretical underpinnings and effectiveness.
222  Paul A. Boelen et al.

Theoretical basis of CBT for PGD

One puzzling aspect of PGD is that, although people with PGD are so bound
up with the loss that they have difficulty functioning, the loss continues to feel
unreal. That is, for people suffering PGD, the disbelief, pangs of pain, and separa-
tion distress that normally occur intensely early after the death exacerbate rather
than gradually fade. We formulated a cognitive–behavioral model that attempts
to explain why, in some individuals, acute grief reactions persist and exacerbate
(Boelen, van den Hout, & van den Bout, 2006). This model proposes that three
interrelated processes account for this: (1) insufficient elaboration and integra-
tion of the loss within autobiographical memory; (2) negative thinking; and (3)
anxious and depressive avoidance behaviors.
This model draws heavily on general cognitive–behavioral theorizing (Beck,
1976) and multirepresentational models of psychopathology, in which various
forms of mental representation (cognitions, schemas, image representations,
distributed networks) and behavioral and cognitive responses are combined to
explain psychopathology (Dalgleish, 2004; Teasdale, 1999). This model is par-
ticularly inspired by Ehlers and Clark’s (2000) cognitive model of PTSD. Next,
these three processes will be described in more detail.

Insufficient elaboration and integration of the loss

Prototypically, in uncomplicated grief, explicit (i.e., consciously accessible)
knowledge that the separation is irreversible gradually gets integrated with exist-
ing knowledge about the self and the lost person, which is part of the autobio-
graphical memory base. This process is fostered by actively elaborating on the
implications of the loss for the self in the past, present, and future, as well as
by confronting external changes caused by the loss. Thus, through a process of
(active) elaboration, explicit knowledge about the loss gradually gets connected
with implicit memory knowledge about the self and the relation with the lost
person. The effect of this process is that the loss becomes part of the life story of
the mourner and gradually becomes a less disruptive, more normalized (albeit still
painful) event (cf. Ehlers & Clark, 2000).
Our CBT model postulates that, in PGD, this process of elaboration and integra-
tion is stalled or incomplete. This has several effects. A first effect is that, because
memory knowledge about the separation is disconnected from other memory
knowledge, the death continues to be a very shocking, unbelievable event. At the
same time, because the loss is so emotional and consequential, all kinds of stimuli
easily elicit memories, thoughts, and feelings that are associated with the death
such that – eventually – everything is a reminder of it. Stated differently, the lack
of connectivity between memories of the loss and other knowledge is assumed to
cause these memories to continue to intrude into awareness, causing PTSD-like
symptoms of intrusions, and a continued sense of shock, as if the loss happened
very recently rather than months or years ago (cf. Conway & Pleydell-Pearce,
2000).
 CBT for PGD  223
A second effect is that too little adjustment of knowledge about the self takes
place, as a result of which the person is left with a reduced sense of clarity about
the self. Integration of the reality of the loss with knowledge about the self is a
prerequisite for mourners to be able to redefine who they are without the lost
person. To the extent that this integration fails, a mismatch exists between the
external reality (absence of loved one) and self-knowledge (partially defined in
terms of lost person). This, in turn, can lead to an impaired sense of self-clarity:
a situation in which roles, goals, personal attributes, and other aspects of the self
that the lost person co-defined are temporarily unclear, fragmented, and unstable
(Campbell et al., 1996). This reduced self-clarity probably contributes to the wish
to revert back to the pre-loss period, difficulties in accepting the loss, isolation,
and a sense that life lacks meaning – symptoms that are all hallmark features of
PGD (Boelen, Keijsers, & van den Hout, 2012).
The third effect is that no adjustment of the “relationship representation” takes
place, so separation distress persists. Individuals form mental representations of
relationship with close others as part of autobiographical memory (Bowlby, 1980;
Mikulincer, 2006). These contain information about emotional (anxiety, despair)
and behavioral responses (crying, searching) that are activated when the relation-
ship is threatened and that serve to maintain proximity and felt security. When
a loved one dies, a process normally unfolds in which the fact that the loss is
irreversible connects with the relationship representation. This coincides with a
gradual reduction of such reactions of separation distress. To the extent that there
is a lack of integration of the loss with the relationship representation, the absence
of the lost person will continue to generate symptoms of separation distress that
are central to PGD. This notion links up with Shear and Shair’s (2005) biobe-
havioral model of bereavement, which also proposes that “symptoms of acute
grief . . . usually resolve following revision of the internalized representation of
the deceased to incorporate the reality of the death. Failure to accomplish this
integration results in the syndrome of complicated grief’ (p. 253).
In sum, the notion that, in persons suffering from PGD, information about the
separation is insufficiently connected with memory knowledge about the self and
the lost person helps to explain how the loss can be experienced in different ways:
how it can be thought about constantly but still feel shocking and unreal, how
mourners can say that the loss has changed them, but still feel left with a reduced
self-clarity, and how they rationally know that their loss is permanent but, at a less
conscious level, continue to experience the separation as reversible.

Persistent negative thinking

As a second process, the CBT conceptualization proposes that, unlike people
who recover from loss, individuals with PGD have rigid negative cognitions and
assumptions that contribute to the maintenance and exacerbation of their acute
grief symptoms (Boelen, van den Hout, & van den Bout, 2006). Two categories of
cognitions are particularly important. The first includes negative global cognitions
about the self (“I am a worthless person without my husband”), life (“Life has
224  Paul A. Boelen et al.
no meaning any more”), and the future (“I will certainly never find joy again”).
These global negative views may develop when the loss shatters pre-existing
positive views. For instance, the loss of a child may lead to negative views of
self and life, when this event is strongly at odds with pre-existing positive views
and the person is unable to maintain positive views taking into account the loss
event. Such beliefs can also arise when the loss reactivates pre-existing negative
assumptions. The death of a loving partner may reactivate negative views that one
is a worthless person. The importance of global negative views of self and life in
emotional problems following loss accords with earlier theories of coping with
loss (Janoff-Bulman, 1992) and trauma (Foa & Rothbaum, 1998; see also Park,
2010). These theories have emphasized that recovery from such events hinges on
a person’s ability to maintain positive views of self and life following such events.
The second category includes catastrophic misinterpretations of one’s own
reactions to the loss. Bereaved people have to manage painful emotions, thoughts,
and memories. Acceptance of these responses fosters emotional processing.
Problems arise when people interpret these responses in a catastrophic fashion.
Mourners may label the intensity of their sadness as signaling loss of control,
view their numbness as announcing depression, and interpret vivid intrusions
as reflecting insanity. Such misinterpretations are assumed to fuel distress and
avoidance and to prevent the person from reviewing and adjusting to the loss’s
implications. Trauma research has shown that catastrophic interpretations of ini-
tial posttraumatic stress symptoms (e.g., “If I think back to this accident, I will go
out of my mind”) contribute to the development of chronic PTSD (Mayou, Ehlers,
& Bryant, 2002). After bereavement, similar misinterpretations are assumed to
contribute to acute grief reactions becoming chronic.

Anxious and depressive avoidance

Anxious avoidance refers to avoidance of confrontation with the reality, impli-
cations, and pain of the loss, driven by the fear that this confrontation will be
intolerable and unbearable. The importance of avoidance behaviors in maintain-
ing grief draws from early behaviorist accounts in which pathological grief was
essentially seen as a phobia for normal grief reactions (Ramsay, 1977). Anxious
avoidance resembles the concept of experiential avoidance, referring to attempts
to alter the frequency, duration, or form of negatively evaluated private events
such as thoughts, feelings, and memories (e.g., Hayes, Wilson, Gifford, Follette,
& Strohsahl, 1996). It is the opposite of experiential acceptance, reflecting the
willingness to endure unwanted private events without judgment and defense.
Anxious avoidance can manifest itself in situational avoidance of places, pictures,
and people associated with the loss. It may also take the form of cognitive avoid-
ance behaviors, including the suppression of unwanted thoughts and memories, or
rumination about events surrounding the death (e.g., “Why did the loss occur?,”
“How could it have been prevented?”) as a means to keep thoughts and memories
that are even more painful to think about out of awareness.
Anxious avoidance can be distinguished from depressive avoidance, which
refers to withdrawal from social, occupational, and recreational activities that
 CBT for PGD  225
could be rewarding and provide a continued sense of self. The concept draws from
early behavioral models of depression that implicated decreases in non-depres-
sive, rewarding behaviors as a key maintaining factor in clinical depression (e.g.,
Jacobson, Martell, & Dimidjian, 2001). After bereavement, depressive avoidance
can occur when the loss interrupts access to reinforcers for healthy behavior. As
Ramsay (1977) put it, “A widow whose reinforcers consisted of doing everything
with and for her husband suddenly finds herself left with no positive reinforcers
when he dies” (p. 133). In addition, it can occur when mourners lack the skills
needed to achieve valued goals in the absence of the lost person, or when they
think that engaging in activities without the lost person is disrespectful to him or
her. Negative cognitions are assumed to be important in depressive avoidance as
well, especially those concerning the effects of engaging in potentially helpful
behaviors (e.g., “Meeting friends will not make me feel better”) and one’s abilities
to do so (e.g., “I am unable to take up new responsibilities”). The reduction of
interest and competence in autonomous functioning, implicated in the concept of
depressive avoidance, also results from inhibition of the biobehavioral explora-
tory system circuitry that occurs with the activation of attachment responses fol-
lowing loss (Elliot & Reis, 2003; Shear et al., 2011).
Both forms of avoidance are detrimental. For instance, anxious avoidance is
detrimental because it causes distress and interferes with the elaboration and inte-
gration of the irreversibility of the loss and the implications thereof. Depressive
avoidance is detrimental because it interferes with the experience of positive emo-
tions and maintains negative views of the self, life, and future.

Interaction between, and the mediating role of, the three processes
The three processes are assumed not only to directly contribute to symptoms of
PGD, but also to influence each other. For instance, elaboration of the loss, and
subsequent integration of the reality of the loss with memory knowledge about
the self and the lost relationship, is likely to be blocked when reviewing the con-
sequences of the loss brings to mind negative thoughts about the self, life, and the
future. Likewise, a tendency to engage in anxious avoidance is likely to prevent
such integration. Negative cognitions and avoidance behaviors also have a mutual
impact. For instance, negative cognitions about the self and life are likely to main-
tain a depressive cycle of withdrawal and inactivity. Catastrophic misinterpreta-
tions of grief reactions can contribute to anxious avoidance behaviors, which, in
turn, prevent correction of such misinterpretations.
Important also is that the three processes are assumed to mediate the impact
of various established risk factors for poor bereavement outcome. These include
personality characteristics of the bereaved individual, such as neuroticism and
insecure attachment style; features characterizing the loss, such as who died
and the mode of death; and events and circumstances occurring in the aftermath
(e.g., perceived social responses). Put another way, it is proposed that the three
processes are intermediate mechanisms that explain why, for instance, people
who are insecurely attached have an elevated chance of developing PGD (cf.
Wijngaards-de Meij et al., 2007). This is so because these people are likely to
226  Paul A. Boelen et al.
have more difficulties in accepting and integrating the reality of the loss (Process
1), maintaining a positive view of self (Process 2), and engaging in helpful coping
behaviors (Process 3), as a result of which they have a greater chance of develop-
ing PGD. The notion of mediation is important because it sheds light on change-
able mechanisms (e.g., negative cognitions) that can be targeted in treatment, to
curb the effect of more static, less easily changeable risk factors (e.g., personality
features) on the development and maintenance of PGD.

Cognitive–behavioral treatment
The aim of CBT for PGD is to alleviate persisting acute grief symptoms and to
help the person to achieve valued goals. To accomplish this, (1) the loss needs to
be integrated with existing knowledge, (2) unhelpful thinking patterns need to
be identified and altered, and (3) unhelpful avoidance strategies need to replaced
by more helpful ones. Different conventional CBT interventions can be used to
achieve these aims. Examples of interventions are described below.

Promoting elaboration and integration of the loss

A key intervention to directly target the lack of integration is exposure. During
exposure, the person is encouraged to gradually confront the painful reality of
the loss and to elaborate upon its implications. Several means can be used to
achieve this aim. The person suffering from PGD can be asked to tell or write a
detailed story about the events surrounding the death. This can be followed up
by repeated reliving of the most painful aspects (or so-called hot spots) of the
story. Such exposure is similar to revisiting the death, an intervention central to
Shear’s “complicated grief treatment” (Shear et al., 2005, 2011). This exercise
uses a procedure similar to prolonged exposure in the treatment of PTSD (Foa &
Rothbaum, 1998) and includes a reliving and subsequent further discussion of the
most troubling moments surrounding the death.
Imaginal exposure or reliving can be complemented with writing assignments.
People suffering from PGD can be instructed to write a detailed account of the
moments surrounding the loss: a procedure that is central to Internet-based CBT
for PGD (Wagner et al., 2006; Chapter 17 in this volume). In our own experience,
it is particularly useful to encourage them to write a letter to the lost person, care-
fully reviewing what is missed most now that he or she is dead. Within-session
exposure can also be complemented with exposure to situations or stimuli outside
therapy. Visiting the hospital where the loved one died or visiting places the lost
person always used to visit may help people with PGD to accept that the loss
occurred and to put it in the past.
As applied in this manner, exposure is not used to promote emotional habitu-
ation to painful memories and emotions. Instead, it is used to identify the most
painful memories that need to be confronted and worked through in treatment and
the most important guilty, angry, shameful, and frightening beliefs that need to be
re-evaluated. It is also used to encourage the person with PGD to fully connect
 CBT for PGD  227
with the reality of the loss and to lessen the disbelief. In addition, it is meant to
help the person to experience that experiential avoidance of the loss is fruitless,
that he or she has the strength to confront the loss, and that doing so lessens the
pain and fosters adjustment.

Changing maladaptive cognitions and assumptions

Cognitive restructuring aims to alleviate emotional suffering by (1) identifying
(maladaptive) cognitions that underlie a person’s suffering in particular situations
and circumstances; (2) examining the validity and utility of these cognitions; and
(3) reformulating these cognitions incorporating information gathered in step 2
into cognitions that are associated with less suffering and facilitate constructive
action. Cognitive restructuring focuses on the idiosyncratic meanings of the loss
and its sequelae, with a particular focus on global negative views of the self, life,
and the future and catastrophic misinterpretation of grief reactions. In identify-
ing maladaptive cognitions, it is important to search for relevant cognitions. Not
every negative cognition can be changed or should be changed. Instead, the thera-
pist should look for those cognitions that are central to the problems of the person,
that interfere with the achievement of valued goals, and that are falsifiable. For
example, it is hard to dispute the cognition “Life is meaningless” but easier to
discuss the validity of the cognition “Life has no meaning now, and I probably
won’t find meaning in life in the future.”
When relevant cognitions are identified, both verbal and behavioral techniques
can be used to change these cognitions. An example of the former one is using
Socratic questioning to investigate the validity (“How do I know that what I think
is true?,” “What evidence is there in favor of and against this thought?”) and util-
ity (“What will happen if I continue thinking this way?,” “What is the worst thing
that can happen if what I think it true?”) of a particular cognition.
Behavioral techniques include behavioral experiments. These are specified
actions/assignments that patients undertake in order to test specific maladaptive
cognitions and catastrophic misinterpretations as well as the validity of alternative,
more positive predictions (cf. Bennett-Levy et al., 2004). They are particularly
suitable for testing specific negative predictions (with an If . . . , then . . . format)
that lead to negative feelings and block constructive action. For instance, a cogni-
tion such as “If I think over the implications of this loss, I will get so sad that I
would go crazy” can be tested by encouraging the person to gradually review the
consequences of the loss within the safe context of the therapy. A negative cogni-
tion such as “If I share my feelings about this loss, then nobody would probably
respond in a supportive manner” could be tested by encouraging the person to set
up a meeting with a friend to talk about how he or she has felt recently.
Writing assignments can be used to complement within-session cognitive
restructuring. For instance, people suffering from PGD can be instructed to write
a supporting letter to an imaginary friend who has the exact same problems, with
an emphasis on trying to help this friend to re-evaluate maladaptive cognitions
about the loss (see Wagner et al., 2006; Chapter 17 in this volume).
228  Paul A. Boelen et al.

Reducing anxious and depressive avoidance

In targeting anxious avoidance, cognitive restructuring is used to identify and
discuss the prediction that underlies the avoidance of particular loss-related
stimuli (e.g., “If I looked at a photo of my deceased wife, the pain would be so
intense that I’d turn into an emotional wreck”). Then, behavioral experiments, as
described in the previous section, can be used to test the validity of the prediction
further (Bennett-Levy et al., 2004). This often means that people are encouraged
to confront the avoided stimuli, in order to experience that doing so alleviates
rather than attenuates the suffering. Behavioral experiments resemble exposure
interventions. Notably, though, behavioral experiments are explicitly meant to
change specific assumptions, whereas exposure interventions can have various
aims (including identification of “hot spot” memories and maladaptive cogni-
tions, and encouraging elaboration of the loss).
Anxious avoidance may coincide with particular strategies developed to
minimize distress  – strategies that are reminiscent of “safety behaviors” in
anxiety disorders. For instance, people suffering PGD may engage in compulsive
proximity-seeking behavior (e.g., visiting the graveyard twice a day), or rumina-
tive thinking about why the loss occurred, in order to minimize confrontation with
the pain associated with the irreversibility of the death. Response prevention can
be used to gradually eliminate such behavior. The procedure resembles response
prevention as applied in the treatment of obsessive–compulsive disorder (OCD).
Yet, unlike in the treatment of OCD, in which response prevention is used to alter
predicted external threat, in PGD treatment it is used to confront valid thoughts
and feelings that have to be dealt with in treatment.
Behavioral activation is an important intervention in targeting depressive
avoidance. In behavioral activation, people are instructed to register activities
and mood for some days or weeks, in order to experience how activity improves
mood. Then, behavioral interventions are applied to help the person to re-engage
in rewarding activities (Jacobson et al., 2001). In addition, treatment could focus
on identification of social, occupational, or recreational goals, and planning
actions necessary to achieve goals. Dependent on the nature of the goals, this
can be accompanied by social skill training, problem-solving skill training, and
time management training (Jacobson et al., 2001). Personal goal work is also
central to “complicated grief treatment” (Shear et al., 2011); it is based on the
idea that working toward the achievement of valued goals fosters the experi-
ence of positive emotions, the ability to solve problems, and the motivation to
confront painful information. In behavioral activation, there is a clear focus on
action, irrespective of certain aversive thoughts and mood states. A key idea
behind this approach is that it is not necessary to change mood before behavior
can be changed but, on the contrary, behavior change can precede improvement
of mood. A further idea is that activation and working toward the achievement
of valued goals can facilitate self-clarity and a continued sense of self (Ehlers,
2006).
 CBT for PGD  229

Research on CBT theory and treatment for PGD

In this section we will review evidence that supports the CBT theory and treat-
ment of PGD.

Theory
Because the underpinnings of CBT for PGD draw heavily on existing cogni-
tive–behavioral theorizing (Beck, 1976; Dalgleish, 2004; Ehlers & Clark, 2000),
the extensive body of evidence supporting the view that unhelpful thoughts,
behaviors, and memory processes indeed work in concert, maintaining all kinds
of emotional suffering, provides indirect support for the CBT approach to PGD.
However, an increasing number of studies have directly investigated causes and
correlates of PGD from the perspective of CBT.
For instance, earlier research findings of a linkage between grief severity
and negative views about meaningfulness of the world and the worthiness of
the self (Schwartzberg & Janoff-Bulman, 1991) and self-blame and other-blame
(Field & Bonanno, 2001) accord with the importance of negative cognitions in
grief. In several of our own studies, we found evidence for a significant linkage
between negative cognitions and assumptions and PGD severity. For instance,
in a prospective study, stronger endorsement of negative assumptions about the
self, life, and the future in the early stages of grief predicted more severe PGD
and depression across 2 years post loss (Boelen, van den Bout, & van den Hout,
2006). Furthermore, in cross-sectional and prospective studies (Boelen, van den
Bout, & van den Hout, 2003, 2010), we found evidence that catastrophic misin-
terpretations of grief reactions were associated with more severe PGD as well as
with tendencies to experientially avoid the pain of the loss. The importance of
these misinterpretations in predicting PGD was also supported in a large-scale
longitudinal study by Van der Houwen, Stroebe, Schut, Stroebe, and van den Bout
(2010).
Research has also provided evidence for the importance of avoidance behaviors
in PGD. For instance, generic measures of deliberate avoidance of loss-related
stimuli have been found to be significantly associated with loss-related distress in
several studies (e.g., Bonanno, Papa, Lalande, Nanping, & Noll, 2005; Shear et al.,
2007). In a cross-sectional study, we found evidence that indices of situational and
cognitive avoidance were associated with PGD (Boelen & van den Hout, 2008).
Importantly, the strength of the association was stronger in mourners who strongly
endorsed catastrophic misinterpretations, attesting to the notion that avoidance
strategies interact with catastrophic misinterpretations in maintaining PGD. A
further cross-sectional study showed that anxious and depressive avoidance are
separate constructs having distinct associations with PGD severity (Boelen & van
den Bout, 2010). Several studies have pointed at a linkage between PGD and
difficulties to retrieve specific memories (e.g., Chapter 13 in this volume). Given
that such difficulties are assumed to reflect avoidant tendencies, these studies also
support the importance of avoidance in maintaining PGD.
230  Paul A. Boelen et al.
Relatively few studies have as yet addressed the hypothesis that PGD is asso-
ciated with a lack of connectivity between explicit knowledge of the loss and
implicit memory knowledge. Although this is mainly an implicit process occur-
ring at the level of the autobiographical memory base, it is postulated to manifest
itself in at least two introspectively accessible phenomena. The first is a sense of
unrealness that can be defined as a subjective sense of uncertainty or ambivalence
about the irreversibility of the separation. The second is a reduced sense of clarity
about the self. In a series of studies we found evidence that increased levels of
PGD are indeed associated with a greater sense of “unrealness” about the loss
(even when controlling for negative cognitions, avoidance, and concomitant
depression) and with an impaired sense of self-clarity (Boelen, 2010; Boelen et
al., 2012). Two studies by Maccallum and Bryant provided further support for a
linkage between PGD and insufficient emotional processing at the level of auto-
biographical memory. In the first of these, people with PGD were found to report
more self-defining memories (i.e., vivid memories comprising enduring concerns
about the self) related with the lost person than people without PGD (Maccallum
& Bryant, 2008). This can also be taken as evidence that PGD is associated with
an impairment in the adjustment of self-representational knowledge. In a second
study, PGD patients were found to have an attentional bias for information related
with death and separation (Maccallum & Bryant, 2010) – a finding that accords
with the notion that PGD is characterized by a reduced integration of such knowl-
edge with other autobiographical knowledge.
Two studies provided evidence for the hypothesis that a lack of integration,
negative thinking, and avoidance behaviors indeed mediate the impact of estab-
lished personality-related and situational risk factors on PGD severity. In a cross-
sectional study, these three processes were found to mediate the associations of
neuroticism and attachment insecurity with PGD severity (Boelen & Klugkist,
2011). In a longitudinal study by Van der Houwen et al. (2010), catastrophic mis-
interpretations were found to mediate the impact of several risk factors on PGD,
including gender, neuroticism, and expectedness of the death.

Treatment
CBT has consistently been found to be an effective treatment for a wide range
of disorders (Beck, 2005). Again, this indirectly supports the relevance of this
approach for treating PGD. More direct evidence for the effectiveness of CBT
interventions comes from early studies by Mawson, Marks, Ramm, and Stern
(1981) and Sireling, Cohen, and Marks (1988). In these studies, exposure to
loss-related stimuli was found to lead to reduction in symptoms of problematic
forms of grief. As noted, some of the more recently conducted treatment studies
have also provided evidence for the effectiveness of CBT for PGD. In a large
trial, Shear et al. (2005) compared the effectiveness of “complicated grief treat-
ment” with the effects of interpersonal psychotherapy. The former treatment was
significantly more effective in terms of effect sizes and the time it took before
significant reductions in CG symptoms occurred than the latter approach. Wagner
 CBT for PGD  231
et al. (2006) subjected patients with PGD to an Internet-based CBT treatment. In
comparison with patients in a waiting-list control group, those who underwent the
treatment experienced a greater reduction in PGD and related symptoms.
In our own treatment trial we randomly allocated 54 people with PGD to one
of three treatments: one of two CBT treatments or non-directive supportive treat-
ment (Boelen et al., 2007). The two CBT conditions consisted of six 45-minute
manual-based sessions of cognitive restructuring (CR) and six sessions of expo-
sure therapy (ET) applied in two orders (CR + ET and ET + CR). The six sessions
of CR focused on explanation of the rationale of CR and learning to identify,
dispute, and alter negative cognitions. ET sessions included narrating the story of
the loss in detail, identification of internal and external reminders of the loss that
were avoided, and gradual confrontation with these reminders. Different forms of
exposure were used (e.g., exposure in vitro when patients avoided particular mem-
ories, response prevention when they engaged in compulsive proximity-seeking
behaviors). Results showed that symptoms of PGD and general psychopathology
declined significantly more in people allocated to the CBT treatments than in
those allocated to supportive counseling. For instance, in the intention to treat
analysis, pre-treatment to post-treatment effect sizes (Cohen’s d) on the Inventory
of Complicated Grief, a well-validated measure of PGD (Prigerson et al., 1995),
were 0.87 for the CR + ET condition and 1.29 for the ET + CR condition, com-
pared with 0.42 for the counseling condition. Effect sizes for CBT conditions are
large according to conventional guidelines and resemble the effect size of 1.35 in
the intention-to-treat analyses found by Shear et al. (2005) for their PGD treat-
ment. In follow-up analyses, we found that stronger reduction in PGD severity
was significantly associated with stronger reductions in negative cognitions and
avoidance. Although this was not a formal test of mediation, outcomes support the
relevance of targeting negative thinking and avoidance in the treatment of PGD
(Boelen, de Keijser, van den Hout, & van den Bout, 2011).

Closing comments
In CBT perspectives on PGD (Boelen, van den Hout, & van den Bout, 2006; Shear
et al., 2005, 2011; Wagner et al., 2006) it is hypothesized that memory processes,
negative cognitions, and avoidance behaviors play a key role in the development
and maintenance of PGD. As outlined in this chapter, there is increasing evidence
supporting this hypothesis. Nonetheless, more work needs to be done to test basic
premises of CBT perspectives on PGD. For instance, studies conducted to date
have mostly relied on self-report measures. It would be relevant for future stud-
ies to use other methods to study the role of memory processes, cognitions, and
behaviors in PGD, including methods based on social cognition and diary-keeping
methods to map out avoidance behaviors. In addition, experimental research is
needed to test the proposed directions of causality between variables.
More work obviously also needs to be done in the area of CBT treatment for
PGD. Although the findings summarized in this chapter support the potential
strength of CBT as a treatment of PGD, there is still ample scope for improving
232  Paul A. Boelen et al.
this treatment. For instance, in our own study (Boelen et al., 2007), only 32.6%
of patients randomly assigned to the CBT conditions showed clinically significant
reductions in PGD severity. This being the case, it seems important to further
refine the recent promising treatments. It is important for future studies to try to
disentangle the effects of specific components of treatment for specific subgroups
of people suffering PGD. It is also important to enhance clarity on mechanisms
of change of CBT, that is, whether or not CBT indeed produces alleviation of dis-
tress because it lessens negative cognitions and avoidance. Notwithstanding these
considerations, there are reasons to be optimistic about the explanatory value and
clinical usefulness of applying cognitive–behavioral theorizing to the study and
treatment of persons with PGD.

References
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International
Universities Press.
Beck, A. T. (2005). The current state of cognitive therapy: A 40-year retrospective. Archives
of General Psychiatry, 62, 953–959.
Bennett-Levy, J., Butler, G., Fennell, M. J. V., Hackmann, A., Mueller, M. & Westbrook,
D. (Eds.) (2004). The Oxford guide to behavioural experiments in cognitive therapy.
Oxford, Oxford University Press.
Boelen, P. A. (2010). A sense of “unrealness” about the death of a loved-one: An explora-
tory study of its role in emotional complications among bereaved individuals. Applied
Cognitive Psychology, 24, 238–251.
Boelen, P. A., & van den Bout, J. (2010). Anxious and depressive avoidance and symp-
toms of prolonged grief, depression, and posttraumatic stress-disorder. Psychologica
Belgica, 50, 49–67.
Boelen, P.  A., van den Bout, J., & van den Hout, M. A. (2003). The role of negative
interpretations of grief reactions in emotional problems after bereavement. Journal of
Behavior Therapy and Experimental Psychiatry, 34, 225–238.
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2006). Negative cognitions and
avoidance in emotional problems after bereavement: A prospective study. Behaviour
Research and Therapy, 44, 1657–1672.
Boelen, P. A., van den Bout, J., & van den Hout, M. A. (2010). A prospective examination
of catastrophic misinterpretations and experiential avoidance in emotional distress fol-
lowing loss. Journal of Nervous and Mental Disease, 198, 252–257.
Boelen, P. A., & van den Hout, M. A. (2008). The role of threatening misinterpretations and
avoidance in emotional problems after loss. Behavioural and Cognitive Psychotherapy,
36, 71–88.
Boelen, P. A., van den Hout, M. A., & van den Bout, J. (2006). A cognitive–behavioral
conceptualization of complicated grief. Clinical Psychology: Science and Practice, 13,
109–128.
Boelen, P. A., de Keijser, J., van den Hout, M. A., & van den Bout, J. (2007). Treatment of
complicated grief: A comparison between cognitive behavioral therapy and supportive
counseling. Journal of Consulting and Clinical Psychology, 75, 277–284.
Boelen, P. A., de Keijser, J., van den Hout, M. A., & van den Bout, J. (2011). Factors
associated with outcome of cognitive behavioral therapy for complicated grief: A pre-
liminary study. Clinical Psychology & Psychotherapy, 18, 284–291.
 CBT for PGD  233
Boelen, P. A., Keijsers, L., & van den Hout, M. A. (2012). The role of self-concept clarity
in prolonged grief disorder. Journal of Nervous and Mental Disease, 200, 56–62.
Boelen, P. A., & Klugkist, I. (2011). Cognitive behavioural variables mediate the associa-
tions of neuroticism and attachment insecurity with prolonged grief disorder severity.
Anxiety, Stress, & Coping, 24, 291–307.
Bonanno, G. A., Papa, A., Lalande, K., Nanping, Z., & Noll, J. G. (2005). Grief process-
ing and deliberate grief avoidance: A prospective comparison of bereaved spouses and
parents in the United States and China. Journal of Consulting and Clinical Psychology,
73, 86–98.
Bowlby, J. (1980). Attachment and loss, vol. 3: Loss: Sadness and depression. New York:
Basic Books.
Campbell J. D., Trapnell, P. D., Heine, S. J., Katz, I. M., Lavallee, LF, & Lehmann D. R.
(1996). Self-concept clarity: Measurement, personality correlates and cultural bounda-
ries. Journal of Personality and Social Psychology, 70, 141–156.
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical
memories in the self-memory system. Psychological Review, 107, 261–288.
Dalgleish, T. (2004). Cognitive approaches to posttraumatic stress disorder: The evolution
of multirepresentational theorizing. Psychological Bulletin, 130, 228–260.
Ehlers, A. (2006). Understanding and treating complicated grief: What can we learn from
posttraumatic stress disorder? Clinical Psychology: Science and Practice, 13, 135–140.
Ehlers, A., & Clark, D.  M. (2000). A cognitive model of posttraumatic stress disorder.
Behaviour Research and Therapy, 38, 319–345.
Elliot, A. J., & Reis, H. T. (2003). Attachment and exploration in adulthood. Journal of
Personality and Social Psychology, 85, 317–331.
Field, N. P., & Bonanno, G A. (2001). The role of blame in adaptation in the first 5 years
following the death of a spouse. American Behavioral Scientist, 44, 764–781.
Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitive–behavioral
therapy for PTSD. New York: Guilford.
Hayes, S.  C., Wilson, K.  G., Gifford, E.  V., Follette, V.  M., & Strohsahl, K. (1996).
Experiential avoidance and behavioral disorders: A functional dimensional approach to
diagnosis and treatment. Journal of Consulting and Clinical Psychology, 64, 1152–1168.
van der Houwen, K., Stroebe, M., Schut, H., Stroebe, W., & van den Bout, J. (2010).
Mediating processes in bereavement: The role of rumination, threatening misinterpreta-
tions, and deliberate grief avoidance. Social Science and Medicine, 71, 1669–1676.
Jacobson, N. S., Martell, C. R., & Dimidjian, S. (2001). Behavioral activation treatment for
depression: Returning to contextual roots. Clinical Psychology: Science and Practice,
8, 255–270.
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma.
New York: Free Press.
Maccallum, F., & Bryant, R.  A. (2008). Self-defining memories in complicated grief.
Behaviour Research and Therapy, 46, 1311–1315.
Maccallum, F., & Bryant, R. A. (2010). Attentional bias in complicated grief. Journal of
Affective Disorders, 125, 316–322.
Mayou, R. A., Ehlers, A., & Bryant, B. (2002). Posttraumatic stress disorder after motor
vehicle accidents: 3-year follow-up of a prospective longitudinal study. Behaviour
Research and Therapy, 40, 665–675.
Mawson, D., Marks, I., Ramm, E., & Stern, R. S. (1981). Guided mourning for morbid
grief: A controlled study. British Journal of Psychiatry, 138, 185–193.
234  Paul A. Boelen et al.
Mikulincer, M. (2006). Attachment, caregiving, and sex within romantic relationships: A
behavioral systems perspective. In Mikulincer, M., & Goodman, G. S. (Eds.), Dynamics
of romantic love: Attachment, caregiving, and sex (pp. 23–44). New York: Guilford.
Park, C. L. (2010). Making sense of the meaning literature: An integrative review of mean-
ing making and its effects on adjustment to stressful life events. Psychological Bulletin,
136, 257–301.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged Grief Disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Medicine 6(8), e1000121.
Prigerson, H.  G., Maciejewski, P.  K., Reynolds, C.  F., Bierhals, A.  J., Newsom, J.  T.,
Fasiczka, A., et al. (1995). Inventory of Complicated Grief: A scale to measure mal-
adaptive symptoms of loss. Psychiatry Research, 59, 65–79.
Ramsay, R. W. (1977). Behavioural approaches to bereavement. Behaviour Research and
Therapy, 15, 131–135.
Schwartzberg, S.  S., & Janoff-Bulman, R. (1991). Grief and the search for meaning:
exploring the assumptive worlds of bereaved college students. Journal of Social and
Clinical Psychology, 10, 270–288.
Shear, M.  K., Boelen, P.  A., & Neimeyer, R.  A. (2011). Treating Complicated Grief:
Converging approaches. In Neimeyer, R. A., Harris, D. L., Winokuer, H. R., & Thornton,
G. F. (Eds.), Grief and bereavement in contemporary society: Bridging research and
practice (pp. 139–163). New York: Routledge.
Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. III. (2005). Treatment of complicated
grief: A randomized controlled trial. JAMA, 293, 2601–2608.
Shear, K., Monk, T., Houck, P., Melhem, N., Frank, E., Reynolds, C. III, & Sillowash, R.
(2007). An attachment-based model of complicated grief including the role of avoid-
ance. European Archives of Psychiatry and Clinical Neuroscience, 257, 453–461.
Shear, K., & Shair, H. (2005). Attachment, loss, and complicated grief. Developmental
Psychobiology, 47, 253–267.
Sireling, L., Cohen, D., & Marks, I. (1988). Guided mourning for morbid grief: A con-
trolled replication. Behavior Therapy, 19, 121–132.
Teasdale, J. D. (1999). Multi-level theories of cognition–emotion relations. In Dalgleish, T.,
& Power, M. J. (Eds.), Handbook of cognition and emotion (pp. 665–682). Chichester,
UK: Wiley.
Wagner, B., Knaevelsrud, C., & Maercker, A. (2006). Internet-based cognitive–behavio-
ral therapy for complicated grief: A randomized controlled trial. Death Studies, 30,
429–453.
Wijngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W., van den Bout, J., van der
Heijden, P., & Dijkstra, I. C. (2007). Neuroticism and attachment insecurity as predic-
tors of bereavement outcome. Journal of Research in Personality, 41, 498–505.
17 Internet-based bereavement
interventions and support
An overview
Birgit Wagner

Introduction
Interpersonal communication and relationships have changed dramatically with
the growing influence of the Internet. The new platforms offered by the Internet
have not only transformed social and professional life, but also opened up new
channels of communication for those experiencing bereavement. Specifically,
the Internet allows bereaved individuals to seek social support without physical
interaction. In recent years, numerous Internet-based discussion forums have been
established for bereaved populations (e.g., forums for bereaved parents, widowers
and widows, and suicide survivors). These Internet discussion groups are usually
self-help groups; many are developed and moderated by bereaved individuals.
Online memorial sites on which people who have lost family members or friends
grieve and mourn publicly are another expression of grief. In this often very
personal form of public grieving, bereaved individuals or communities create
memorial websites including photographs to describe the deceased person’s life
and dying.
Parallel to the development of the social media of the Internet, new forms of
psychotherapeutic interventions  – therapist-supported or self-help  – have been
developed for a variety of patient groups. Accumulating research has shown
that Internet-based interventions  – particularly cognitive–behavioral interven-
tions – can be beneficial for psychological health, with treatment effects compa-
rable to those of face-to-face treatments (Barak, Hen, & Boniel-Nissim, 2008).
Specifically, Internet-based interventions for depression have been delivered in
different forms, from self-help treatments delivered without therapist guidance
to mainly text-based interventions with high therapist involvement (Andersson,
2006; Ruwaard et al., 2009; Spek et al., 2006). Computerized interventions for
depression with therapist support showed a mean effect size comparable to face-
to-face treatment for depression, whereas interventions with little or no thera-
pist contact had a significantly smaller treatment effect size (Spek et al., 2007).
Further, interventions aimed at patients with posttraumatic stress disorder (Lange
et al., 2003) and anxiety have proved most effective, whereas interventions target-
ing patients with somatic problems (e.g., weight loss) have turned out to be less
effective (Barak et al., 2008).
236  Birgit Wagner
A number of Internet-based bereavement interventions have been developed
in recent years. The interventions are delivered in various forms, from text-based
approaches with therapist feedback (Kersting, Kroker, Schlicht, & Wagner, 2010;
Wagner, Knaevelsrud, & Maercker, 2006) to self-help treatments delivered with-
out therapist feedback (van der Houwen, Schut, van den Bout, Stroebe, & Stroebe,
2010). This chapter describes these different treatment approaches, beginning
with low-threshold online bereavement support groups, efficacy of bereavement
interventions, continuing with Internet-based interventions that include therapist
support, and finally discussing self-help bereavement interventions. The chapter
presents the procedures and key components of therapist-supported Internet-based
grief interventions, and discusses important indications and contraindications for
Internet-based therapies.

Online bereavement support groups

The advent of Internet social networking sites has led to a dramatic increase in
the numbers of online support groups in recent years. There are Internet support
groups for practically every interest group (e.g., breast cancer support groups,
survivors of suicide), operating through various Internet applications (e.g., chat
rooms, forums, bulletin boards). These groups can be especially beneficial for
individuals who feel socially stigmatized or have difficulty linking up with
others facing the same challenges in the offline world. The opportunity for social
exchange with others in the same situation is one of the main advantages of online
support groups (Davison, Pennebaker, & Dickerson, 2000). These groups can
provide a sense of belonging (McKenna & Bargh, 2000) and help users to learn
to cope with their bereavement. Moreover, the lack of geographic boundaries
enables even individuals living in remote areas to participate in highly specific
groups. In face-to-face groups, aspects of physical appearance such as attractive-
ness and ethnicity play a key role in determining participants’ interactions and
the development of friendships. Whereas these first impressions are difficult to
overcome in face-to-face settings (McKenna & Seidman, 2005), in online set-
tings, first impressions are based on the kind of opinions expressed and the kind
of information users reveal about themselves, rather than on physical features.
Mourning and loss support groups account for some 10% of the electronic
support groups in the health and wellness section of Yahoo! Groups (van der
Houwen, Stroebe, Schut, Stroebe, & van den Bout, 2010), making bereavement
the third most popular topic. However, there has been little research into these
groups and their effects on the mental health of users. One of the first empirical
studies in the field compared parent suicide survivors using online support groups
with those participating in face-to-face support groups (Feigelman, Gorman, Beal,
& Jordan, 2008). The study revealed that the Internet group experienced greater
social stigmatization and less social support and showed higher symptom levels of
depression, grief, and suicide ideation. These results are in line with the findings
of the cross-sectional Swiss Bereaved Parents Study, which reported that 36% of
 Internet-based interventions and support  237
respondents were active in Internet discussion groups for bereaved parents. This
group showed significantly higher incidence of symptoms of depression, posttrau-
matic stress, and complicated grief (Kelly, 2008). However it is unclear whether
these symptoms were caused by the frequent use of online groups, in which grief-
related feelings were regularly actualized and re-experienced, or if bereaved indi-
viduals with higher symptom levels were more likely to seek out and participate in
online support groups. Furthermore, the two groups were not randomly allocated
to the Internet versus the non-discussion group, and therefore conclusions regard-
ing the effect should be drawn carefully. Of the bereaved parents participating in
Internet discussion groups, 15% spent more than 2 hours per day in these forums,
22% used them for less than 1 hour per day, and 33% logged on at least once
per week. In addition, 45% of participants in online discussion groups felt better
accepted in their online relationships than in their face-to-face ones.
Van der Houwen and colleagues (van der Houwen, Stroebe, et al., 2010)
recruited participants for their study exclusively over the Internet. Of their
respondents, 62% used online mutual bereavement support groups, spending an
average of 7.4 hours per week in these forums. This group was compared with
bereaved individuals who had never used online bereavement support groups.
The online support group members were significantly younger, more likely to
have lost a child, and less likely to belong to a religious community. They showed
poorer mental health, higher emotional loneliness, and lower levels of social sup-
port. Vicary and Fraley (2010) evaluated students’ reactions to the shootings at
Virginia Tech in 2007 and their use of social media networks on the Internet. Only
a few hours after the shooting, students had created online support groups on
Facebook. Students were interviewed 2 weeks and 2 months after the shootings,
and their depression and posttraumatic stress were assessed. The results showed
that depression and symptoms of posttraumatic stress disorder (PTSD) 2 months
after the shooting were not related to the number of Facebook groups the student
had joined, the number of messages the student had posted, or the frequency of
using the Facebook wall to discuss the shootings. Overall, Internet use was not
related to depression or PTSD-related symptoms 2 months after the shootings.
However, most students indicated that they perceived the use of the Internet as
beneficial.

Effectiveness of bereavement interventions

A number of meta-analyses and reviews of bereavement treatments (Allumbaugh
& Hoyt, 1999; Currier, Neimeyer, & Berman, 2008; Kato & Mann, 1999; Rowa-
Dewar, 2002; Schut, Stroebe, van den Bout, & Terheggen, 2001; Wittouck, Van
Autreve, De Jaegere, Portzky, & van Heeringen, 2011) describe the results of
preventive interventions, interventions for complicated grievers and for high-risk
groups, individual and group therapies, and cognitive–behavioral related treat-
ments. However, studies of the effectiveness of bereavement interventions yield
inconclusive results, and some reviewers claim that there is no strong empirical
238  Birgit Wagner
evidence that these interventions are effective (Currier et al., 2008; Schut &
Stroebe, 2005). A number of methodological limitations have been addressed:
Interventions are largely methodologically flawed, and inclusion criteria are inco-
herent because of a lack of commonly agreed DSM-IV criteria (Kato & Mann,
1999). Only a small number of studies fulfill the inclusion criteria of a randomized
controlled trial. The results of these various reviews and meta-analyses show that
the overall benefits of bereavement interventions are small (Schut et al., 2001).
Interventions appeared most effective if they were aimed at high-risk groups or if
the grief process was already complicated (Allumbaugh & Hoyt, 1999; Currier et
al., 2008; Schut et al., 2001).
Schut and colleagues (2001) critically query meta-analyses of bereavement
interventions because of the enormous theoretical and conceptional differences
of these studies. In their review of efficacy studies they showed the difference of
effects of primary preventive interventions, preventive interventions of high-risk
groups, and the treatment of individuals who suffer from symptoms of compli-
cated grieving. The secondary preventive interventions for high-risk groups (e.g.,
high levels of distress, traumatic loss, loss of a child) showed modest effects,
and improvements were either temporary or non-existent. The tertiary preventive
intervention for individuals suffering higher levels of distorted and prolonged
grieving symptoms seemed to show more beneficial effects. Wittouck and col-
leagues (2011) have published the newest meta-analysis of prevention and treat-
ment of complicated grief. In this meta-analysis, a stringent inclusion criterion
was used, and therefore only randomized controlled trials in which the outcome
variable was complicated grief measures were included. In this meta-analysis 14
studies met the inclusion criteria. Of these studies, nine were studies on preven-
tive grief interventions and five examined the treatment of grief interventions. The
preventive grief interventions in this meta-analysis showed inconsistent support.
Altogether four studies in this meta-analysis reported positive results with respect
to complicated grief measures. All four of these studies were based on cognitive–
behavioral techniques. The results of this meta-analysis revealed that treatment
interventions aimed at patients suffering from complicated grief seemed to be
efficacious in the reduction of complicated grief symptoms.

Internet-based intervention for complicated grief

First studies of treatments involving cognitive–behavioral components in
bereavement interventions have shown treatment efficacy (Boelen, de Keijser,
van den Hout, & van den Bout, 2007; Shear, Frank, Houck, & Reynolds, 2005;
Wagner et al., 2006; Wittouck et al., 2011). Parallel with this, the effectiveness
of cognitive–behavioral therapy for PTSD has been well documented in a series
of meta-analyses (Bradley, Greene, Russ, Dutra, & Westen, 2005). Confrontation
with intrusive memories of the traumatic event has proven to be a key element
of effective psychotherapy for PTSD. Exposure to traumatic stimuli (e.g., memo-
ries, pictures, or fear-provoking situations) has been found to significantly reduce
avoidance behavior. Given the similarity of some symptoms of complicated grief
 Internet-based interventions and support  239
disorder to those of PTSD, it is conceivable that cognitive–behavioral treatments
designed for PTSD (Foa & Jaycox, 1999) might trigger similar healing mecha-
nisms among individuals with complicated grief.
An Internet-based therapy for complicated grief (Wagner et al., 2006) has been
developed on the basis of the Interapy treatment approach for PTSD (Lange et al.,
2003). Lange, van de Ven, Schrieken, and Emmelkamp (2001) developed an online
intervention for PTSD, in which communication between therapist and patient is
exclusively text-based and asynchronous. The intervention consists of structured
writing assignments based on the cognitive–behavioral therapy approach and the
written disclosure procedure developed by Pennebaker and colleagues (Berry &
Pennebaker, 1993). Lange and colleagues’ first studies of their online intervention
(Lange et al., 2001) showed a large reduction in PTSD symptoms and high treat-
ment efficacy, comparable to that reported for face-to-face cognitive–behavioral
interventions.
The Internet-based intervention for complicated grief consists of three mod-
ules: (1) self-confrontation, (2) cognitive reappraisal, and (3) social sharing. As
the therapy is conducted exclusively over the Internet, a detailed online diag-
nostic procedure is needed to ensure that this form of therapy is appropriate for
the patient. Because there is no possibility for therapists to respond immediately
in crisis situations, some patient groups are not ideally suited to Internet-based
psychotherapy. The current consensus is that the Internet offers insufficient
support for people who are severely depressed or suicidal, or have dissociative
and/or psychotic symptoms. The same applies to people who have a history of
alcohol or substance abuse. All measures were self-reports administered through
an online diagnostic assessment. Newer studies conducted in the field of online
interventions use telephone interviews for structured clinical interviews. Patients
are set ten 45-minute writing assignments over a 5-week period. The instructions
are based on a treatment protocol, but individually tailored to patients’ needs.
The treatment is delivered by psychologists trained in cognitive–behavioral psy-
chotherapy and with special training in the use of writing assignments to treat
complicated grief. At the beginning of each phase, the patient receives detailed
psychoeducation on the principles of the treatment module and sets a date and
time to complete the writing assignment. After every two assignments, the patient
receives feedback and instructions for the next two assignments (for a case study,
see Wagner, Knaevelsrud, & Maercker, 2005). The treatment begins with a self-
confrontation phase. Imaginative confrontation is a technique that encourages
the patient to revisit feelings, emotions, and images surrounding the death of a
significant person that are repressed or trigger fear. This first phase of the therapy
focuses on confronting painful memories, thoughts, and feelings concerning the
circumstances of the death. The patient is asked to describe the most painful
moment in as much detail as possible, focusing on the most difficult aspects. In
the second phase, cognitive reappraisal, the patients are instructed to write a sup-
portive and encouraging letter to a hypothetical friend. They are asked to imagine
that this friend has also experienced the loss of a significant other and is now
facing the same difficulties. The letter should reflect on guilt feelings, challenge
240  Birgit Wagner
dysfunctional automatic thinking and behavior patterns, and correct unrealistic
assumptions. The patients are encouraged to think about rituals to remember the
deceased by, to re-access positive memories of the deceased, and to identify ways
of activating resources such as social contacts, positive competences, and experi-
ences. In the final phase, patients take symbolic leave of the traumatic event by
writing a letter to a significant person, to someone who witnessed the traumatic
event, or to themselves. The letter can be sent after finishing the therapy, but this
is not obligatory.
Results of a randomized controlled trial showed that, relative to a waiting list
group, the treatment group experienced statistically and clinically significant
reductions in the main symptoms of complicated grief and in general psychopa-
thology at posttreatment and at 3-month follow-up (Wagner et al., 2006). These
beneficial treatment effects were maintained at 1.5-year follow-up (Wagner &
Maercker, 2007). However, there was no control group against which the outcomes
of the treated sample could be compared. Of the participants who began the treat-
ment, only 8% did not complete the intervention. The impact of the intervention
on posttraumatic growth and optimism was also evaluated (Wagner, Knaevelsrud,
& Maercker, 2007). In the cognitive restructuring module, participants were to
address questions such as the following in their letter to the hypothetical friend:
“Is it possible that your friend learnt something from the death of her daughter or
from what happened after this loss? Has he or she discovered something about life
that she would otherwise not have known – or found out only much later?” The
results indicated that posttraumatic growth increased significantly in the treatment
group, while optimism remained unchanged (Wagner et al., 2007). An explanation
for this finding could be that optimism is a relatively stable aspect of personality
(Scheier & Carver, 1985) and that the treatment was too short to affect it.
Based on the outcomes of the intervention for patients with complicated grief,
an Internet-based prevention program was developed for high-risk complicated
grief groups (Wagner & Maercker, 2008). This intervention is shorter than the
intervention for complicated grief, consisting of two 45-minute sessions per week
over 3 weeks. The intervention consists of the following modules, each involving
one writing assignment: (1) describing the circumstances of the death, (2) using
the life-imprint method to explore the biography and life imprint of the deceased,
(3) keeping a daily diary of social activities and sleep hygiene, (4) cognitive
restructuring of dysfunctional thoughts, such as responsibility for the death and
feelings of guilt, (5) communication within the family, (6) gender-specific coping
with bereavement, and (7) bond with the deceased. Findings from a pilot study
showed that this intervention resulted in significantly reduced symptoms of grief
and depression (Wagner & Maercker, 2008).

Internet-based treatment for parents after prenatal loss

Pregnancy loss occurs in up to 20% of recognized pregnancies (Savitz, Hertz-
Picciotto, Poole, & Olshan, 2002). The loss of a child during pregnancy causes
significant psychological distress for many women, and may lead to long-lasting
 Internet-based interventions and support  241
bereavement difficulties and psychological illnesses (Kersting, Dorsch, Kreulich,
& Baez, 2004; Kersting et al., 2005, 2009; Mann, McKeown, Bacon, Vesselinov,
& Bush, 2008). Women with a history of pregnancy loss are at particular risk
of disturbances in their psychological adaptation in a new pregnancy (Bergner,
Beyer, Klapp, & Rauchfuss, 2008). A number of interventions have been devel-
oped for women in the aftermath of pregnancy loss; however, few patients are
actually offered this specific treatment after prenatal loss. Currently, the efficacy
of an Internet-based therapy for complicated grief (Wagner et al., 2006) in this
specific patient group is being evaluated. To date, the outcomes of the 5-week
intervention replicate previous findings (Kersting et al., 2011; Kersting, Kroker,
Schlicht, & Wagner, 2011), with the intervention group showing significantly
reduced symptoms of grief, PTSD, and depression at posttreatment relative to the
waiting list group. This symptom reduction was maintained at 3-month follow-
up. An Internet-based therapy has a number of specific advantages for parents
who have lost a child during pregnancy. Members of this subgroup of bereaved
patients are proficient users of online technologies; young women, in particu-
lar, regularly use the Internet to obtain information about health-related topics.
The temporal independence of this form of therapy is especially useful for these
patients, who are often unable to participate in face-to-face treatment as many
who have children would need to organize childcare.

Self-help interventions for bereaved individuals

Internet-based interventions can be delivered in different forms, with various
levels of therapist involvement or as self-help treatments without therapist con-
tact. Based on the research of Lange et al. (2003), Wagner et al. (2006), and the
principles of the Pennebaker paradigm, van der Houwen, Schut, and colleagues
(2010) have developed and evaluated a self-help intervention for bereaved indi-
viduals. The treatment involves no personalized therapist feedback and consists
of just five structured writing assignments. These five assignments cover the three
treatment modules outlined by Lange et al. (2003): (1) exposure; (2) cognitive
restructuring; (3) integration and restoration. This self-help approach has been
found to lead to decreased feelings of emotional loneliness and increased positive
mood. However, no significant change was found in grief-related symptoms or
depression. The attrition rate in this study was 46% (59% in the intervention and
27% in the control condition) over the course of the study.
These results are similar to findings about self-help interventions reported in a
meta-analysis of Internet-based depression interventions (Andersson, Carlbring,
Berger, Almlov, & Cuijpers, 2009), which found that therapist support had a
strong influence on treatment outcomes. Computerized interventions with thera-
pist support showed an average between-group effect size of d = 0.61, which is
comparable with the effects reported for face-to-face treatments for depression.
However, interventions without or with very little therapist contact had a sig-
nificantly smaller treatment effect of d = 0.25. Further, studies on self-guided pro-
grams have shown not only reduced treatment effects, but also substantial attrition
242  Birgit Wagner
rates (Christensen, Griffiths, Mackinnon, & Brittliffe, 2006; Clarke et al., 2002,
2005; Kaltenthaler et al., 2008). Analyses of the relationship between the amount
of therapist time in minutes and treatment outcomes have revealed a significant
correlation between therapist time per participant and the between-group effect
sizes of Internet-based interventions (Palmqvist, Carlbring, & Andersson, 2007).
Based on the findings of their Swedish studies, Andersson and colleagues (2009)
have suggested that about 100 minutes per patient spent by the therapist giving
comments on homework and providing feedback can be sufficient for a 10-week
program. Overall, these studies suggest that a minimal level of therapist contact is
needed to decrease attrition rates and to reduce symptoms.
However, the influence of therapist contact on treatment efficacy in bereave-
ment interventions is still unclear and further research is needed to show if the
findings for Internet-based interventions for depression can be generalized to
bereaved people. The differences across Internet-based bereavement interven-
tions might be also attributable to symptom severity. As Schut et al. (2001) noted,
tertiary interventions for individuals suffering higher levels of distorted and pro-
longed grieving symptoms show better alleviation of complicated grief symptoms
than interventions aimed at all bereaved people or high-risk groups. The self-help
intervention for the bereaved (van der Houwen, Schut, et al., 2010) was aimed
as primary intervention, whereas the Internet-based intervention with therapist
support (Wagner et al., 2006) was aimed at patients who were highly distressed.
Therefore, future research is needed to evaluate the relationship of therapist sup-
port in Internet-based interventions.

Advantages and disadvantages of Internet-based therapeutic

interventions
Internet-based interventions for bereaved individuals are still a relatively new
research topic, and future studies need to focus on who benefits and who does
not. Research on Internet-based therapy for related mental disorders, such as
PTSD or depression, has identified a number of advantages and disadvantages
of this form of treatment (Andersson et al., 2009; Wagner & Maercker, 2010).
One major advantage of Internet-based therapeutic interventions is their anonym-
ity and the lack of geographic boundaries. The anonymity of the Internet helps
patients overcome their initial shame and encourages them to confront themes
such as guilt and social difficulties and to disclose painful feelings. The lack of the
physical presence of the therapist during the online intervention facilitates inti-
macy and increased self-disclosure (Cook & Doyle, 2002). Therefore, structured
writing assignments provide a promising alternative to imaginary confrontations
during face-to-face therapy. Another important advantage is the transparency of
the therapy process. The texts produced can be archived, which not only gives
patients the opportunity to follow up on their therapy at a later time, but also
provides new options for the therapist. The asynchronous, time-delayed nature of
the patient–therapist communication gives the therapist time to reflect on his or
her responses rather than being pressured to give the patient a quick reply.
However, online therapies also have limitations that require further research
 Internet-based interventions and support  243
and call for caution to be exercised in their implementation. First, Internet-based
treatment may not suit all patients; it is important to screen patients carefully and
to work only with those who are likely to benefit from this form of therapy (Suler,
2004). One potential challenge of Internet-based treatment is the potential for
misunderstandings in the absence of spontaneous clarification. On the one hand,
important information might be withheld from the therapist, potentially leading to
incorrect conclusions being drawn. On the other hand, the patient might misinter-
pret the therapist’s feedback. For this reason, it is especially important to establish
maximum transparency with the patient (e.g., providing the name and telephone
contact details of the organization or clinic providing the therapy).
Another concern relates to how the therapist can respond if a patient becomes
suicidal or homicidal and expresses such thoughts in an assignment (Wagner,
Schulz, & Knaevelsrud, 2012). It is practically impossible to respond to the
patient’s crisis in a reliable and timely way within the Internet-based approach.
It therefore seems advisable to obtain the contact numbers of a personal medical
doctor for cases of emergency. Moreover, a careful online or telephone-based
diagnosis procedure is essential before the treatment, and clearly defined exclu-
sion criteria must be established to avoid treating patients with suicidal, dissocia-
tive, or psychotic tendencies over the Internet.

General conclusions
The Internet has become an integral part of everyday work, family, and local com-
munity life. It is thus no surprise that bereaved individuals also turn to the Internet
as a coping mechanism. Internet-based communication offers new possibilities for
social networking and support beyond those available in a face-to-face context.
It facilitates relatively anonymous, asynchronous, and text-based communication
beyond geographical and temporal boundaries. However, although the body of
research on online bereavement behavior is growing, little is yet known about
the characteristics of those who seek mutual or psychotherapeutic bereavement
support online. Specifically, research is needed into the widely used Internet
bereavement support groups and their effects on users’ well-being. Although users
describe these online support groups as beneficial (Vicary & Fraley, 2010) and
participation in these groups does not seem to influence mental health negatively
over a 2- or 3-month period, long-term follow-ups are needed to provide clearer
insights into the consequences of online bereavement support for users’ mental
and physical health.
Other forms of Internet-based bereavement support are interventions involv-
ing various levels of therapist guidance and self-help interventions. Cognitive–
behavioral interventions aimed at patients suffering from complicated grief
have shown high treatment efficacy (Wagner, Knaevelsrud, & Maercker, 2006,
2007; Wagner & Maercker, 2008), with symptom reduction being maintained at
long-term follow-up (Wagner & Maercker, 2007). In fact, the effect sizes were
comparable to those reported for traditional psychological treatment. Self-help
interventions aimed at all bereaved individuals (primary and secondary interven-
tions) have shown lower or no effects on grief-related symptoms. However, these
244  Birgit Wagner
interventions resulted in significantly decreased feelings of emotional loneliness
and increased positive mood (van der Houwen, Schut, et al., 2010). All interven-
tions reviewed in this paper are based on a cognitive–behavioral framework and
share components (e.g., psychoeducation, cognitive restructuring, exposure) that
have proved effective in face-to-face settings. However, little is yet known about
the mechanisms through which structured writing or written disclosure leads to
change in bereaved individuals. Whereas some studies have reported positive
treatment effects for grief-related symptoms, others have not. The crucial dif-
ference may lie in the level of grief complications of the participants and the
therapist feedback.
To conclude, the development of Internet-based bereavement interventions is
still at an early stage and research in this field is scarce. Further empirical stud-
ies on Internet-based interventions and online bereavement support groups are
essential.

References
Allumbaugh, D. L., & Hoyt, W. T. (1999). Effectiveness of grief therapy: A meta-analysis.
Journal of Counseling Psychology, 46, 370–380.
Andersson, G. (2006). Internet-based cognitive–behavioral self help for depression. Expert
Review of Neurotherapeutics, 6, 1637–1642.
Andersson, G., Carlbring, P., Berger, T., Almlov, J., & Cuijpers, P. (2009). What makes
Internet therapy work? Cognitive Behaviour Therapy, 38, 55–60.
Barak, A., Hen, L., & Boniel-Nissim, M. (2008). A comprehensive review and a meta-
analysis of the effectiveness of Internet-based psychotherapeutic interventions. Journal
of Technology in Human Services, 26, 109–159.
Bergner, A., Beyer, R., Klapp, B. F., & Rauchfuss, M. (2008). Pregnancy after early preg-
nancy loss: A prospective study of anxiety, depressive symptomatology and coping.
Journal of Psychosomatic Obstetric Gynaecology, 29, 105–113.
Berry, D. S., & Pennebaker, J. W. (1993). Nonverbal and verbal emotional expression and
health. Psychotherapy and Psychosomatics, 59, 11–19.
Boelen, P. A., de Keijser, J., van den Hout, M. A., & van den Bout, J. (2007). Treatment of
complicated grief: A comparison between cognitive–behavioral therapy and supportive
counseling. Journal of Consulting and Clinical Psychology, 75, 277–284.
Bradley, R., Greene, J., Russ, E., Dutra, L., & Westen, D. (2005). A multidimensional meta-
analysis of psychotherapy for PTSD. American Journal of Psychiatry, 162, 214–227.
Christensen, H., Griffiths, K. M., Mackinnon, A. J., & Brittliffe, K. (2006). Online ran-
domized controlled trial of brief and full cognitive behaviour therapy for depression.
Psychological Medicine, 36, 1737–1746.
Clarke, G., Eubanks, D., Reid, E., Kelleher, C., O’Connor, E., DeBar, L. L., et al. (2005).
Overcoming depression on the Internet (ODIN) (2): A randomized trial of a self-help
depression skills program with reminders. Journal of Medical Internet Research, 7,
e16.
Clarke, G., Reid, E., Eubanks, D., O’Connor, E., DeBar, L. L., Kelleher, C., et al. (2002).
Overcoming depression on the Internet (ODIN): A randomized controlled trial of an
Internet depression skills intervention program. Journal of Medical Internet Research,
4, e14.
 Internet-based interventions and support  245
Cook, J. E., & Doyle, C. (2002). Working alliance in online therapy as compared to face-to-
face therapy: Preliminary results. CyberPsychology & Behavior, 5, 95–105.
Currier, J.  M., Neimeyer, R. A., & Berman, J.  S. (2008). The effectiveness of psycho-
therapeutic interventions for bereaved persons: A comprehensive quantitative review.
Psychological Bulletin, 134, 648–661.
Davison, K. P., Pennebaker, J. W., & Dickerson, S. S. (2000). Who talks? The social psy-
chology of illness support groups. American Psychologist, 55, 205–217.
Feigelman, W., Gorman, B.  S., Beal, K.  C., & Jordan, J.  R. (2008). Internet support
groups for suicide survivors: A new mode for gaining bereavement assistance. Omega
(Westport), 57, 217–243.
Foa, E. B., & Jaycox, L. H. (1999). Cognitive–behavioral theory and treatment of posttrau-
matic stress disorder. In Spiegel, D. (Ed.), Psychotherapeutic frontiers: New principles
and practices (pp. 23–61). Washington, DC: American Psychiatric Press.
van der Houwen, K., Schut, H., van den Bout, J., Stroebe, M., & Stroebe, W. (2010a). The
efficacy of a brief internet-based self-help intervention for the bereaved. Behaviour
Research and Therapy, 48, 359–367.
van der Houwen, K., Stroebe, M., Schut, H., Stroebe, W., & van den Bout, J. (2010b).
Online mutual support in bereavement: An empirical examination. Computers in
Human Behavior, 26, 1519–1525.
Kaltenthaler, E., Sutcliffe, P., Parry, G., Beverley, C., Rees, A., & Ferriter, M. (2008). The
acceptability to patients of computerized cognitive behaviour therapy for depression: A
systematic review. Psychological Medicine, 38, 1521–1530.
Kato, P. M., & Mann, T. (1999). A synthesis of psychological interventions for the bereaved.
Clinical Psychology Review, 19, 275–296.
Kelly, L. (2008). Psychische Gesundheit und Internetdiskussionsforen für Eltern nach dem
Tod ihres Kindes. Unpublished dissertation, Zürich University, Switzerland.
Kersting, A., Dorsch, M., Kreulich, C., & Baez, E. (2004). Psychological stress response
after miscarriage and induced abortion. Psychosomatic Medicine, 66, 795–796; author
reply 796.
Kersting, A., Dorsch, M., Kreulich, C., Reutemann, M., Ohrmann, P., Baez, E., et al.
(2005). Trauma and grief 2–7 years after termination of pregnancy because of fetal
anomalies: A pilot study. Journal of Psychosomatic Obstetric Gynaecology, 26, 9–14.
Kersting, A., Kroker, K., Schlicht, S., Baust, K., & Wagner, B. (2011). Efficacy of cognitive
behavioral internet-based therapy in parents after the loss of a child during pregnancy:
Pilot data from a randomized controlled trial. Archives of Women’s Mental Health, 14,
465–477.
Kersting, A., Kroker, K., Schlicht, S., & Wagner, B. (2011). Internet-based treatment
after pregnancy loss: Concept and case study. Journal of Psychosomatic Obstetric
Gynaecology, 32, 72–78.
Kersting, A., Kroker, K., Steinhard, J., Hoernig-Franz, I., Wesselmann, U., Luedorff, K., et
al. (2009). Psychological impact on women after second and third trimester termination
of pregnancy due to fetal anomalies versus women after birth: A 14-month follow up
study. Archives of Women’s Mental Health, 12, 193–201.
Lange, A., Rietdijk, D., Hudcovicova, M., van de Ven, J. P., Schrieken, B., & Emmelkamp,
P.  M.  G. (2003). Interapy: A controlled randomized trial of the standardized treat-
ment of posttraumatic stress through the internet. Journal of Consulting and Clinical
Psychology, 71, 901–909.
Lange, A., van de Ven, J.  P., Schrieken, B., & Emmelkamp, P.  M.  G. (2001). Interapy.
Treatment of posttraumatic stress through the Internet: A controlled trial. Journal of
Behavior Therapy and Experimental Psychiatry, 32, 73–90.
246  Birgit Wagner
McKenna, K., & Bargh, J. (2000). Plan 9 from cyberspace: The implications of the Internet
for personality and social psychology. Personality and Social Psychology Review, 4,
57–75.
McKenna, K., & Seidman, G. (2005). You, me, and we: Interpersonal processes in elec-
tronic groups. In Amichai-Hamburger, Y. (Ed.), The social net: Understanding human
behavior in cyberspace (pp. 191–217). New York: Oxford University Press.
Mann, J.  R., McKeown, R.  E., Bacon, J., Vesselinov, R., & Bush, F. (2008). Predicting
depressive symptoms and grief after pregnancy loss. Journal of Psychosomatic
Obstetric Gynaecology, 29, 274–279.
Palmqvist, B., Carlbring, P., & Andersson, G. (2007). Internet-delivered treatments with
or without therapist input: Does the therapist factor have implications for efficacy and
cost? Expert Review of Pharmacoeconomics & Outcomes Research, 7, 291–297.
Rowa-Dewar, N. (2002). Do interventions make a difference to bereaved parents? A sys-
tematic review of controlled studies. International Journal of Palliative Nursing, 8,
452–457.
Ruwaard, J., Schrieken, B., Schrijver, M., Broeksteeg, J., Dekker, J., Vermeulen, H., et
al. (2009). Standardized web-based cognitive behavioural therapy of mild to moder-
ate depression: A randomized controlled trial with a long-term follow-up. Cognitive
Behaviour Therapy, 38, 206–221.
Savitz, D. A., Hertz-Picciotto, I., Poole, C., & Olshan, A. F. (2002). Epidemiologic meas-
ures of the course and outcome of pregnancy. Epidemiological Review, 24, 91–101.
Scheier, M.  F., & Carver, C.  S. (1985). Optimism, coping, and health: Assessment and
implications of generalized outcome expectancies. Health Psychology, 4, 219–247.
Schut, H., & Stroebe, M. S. (2005). Interventions to enhance adaptation to bereavement.
Journal of Palliative Medicine, 8, 140–147.
Schut, H., Stroebe, M.  S., van den Bout, J., & Terheggen, M. (2001). The efficacy of
bereavement interventions: Determining who benefits. In Stroebe M. S., Hansson, R. O.,
Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research: Consequences,
coping, and care (pp. 705–738). Washington, DC: American Psychological Association.
Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. (2005). Treatment of complicated
grief: A randomized controlled trial. JAMA, 293, 2601–2608.
Spek, V., Cuijpers, P., Nyklicek, I., Riper, H., Keyzer, J., & Pop, V. (2007). Internet-based
cognitive behaviour therapy for symptoms of depression and anxiety: A meta-analysis.
Psychological Medicine, 37, 319–328.
Suler, J. (2004). The online disinhibition effect. CyberPsychology & Behavior, 7, 321–326.
Vicary, A. M., & Fraley, R. C. (2010). Student reactions to the shootings at Virginia Tech
and Northern Illinois University: Does sharing grief and support over the Internet affect
recovery? Personal and Social Psychology Bulletin, 36, 1555–1563.
Wagner, B., Knaevelsrud, C., & Maercker, A. (2005). Internet-based treatment for compli-
cated grief: Concepts and case study. Journal of Loss and Trauma, 10, 409–432.
Wagner, B., Knaevelsrud, C., & Maercker, A. (2006). Internet-based cognitive–behavio-
ral therapy for complicated grief: A randomized controlled trial. Death Studies, 30,
429–453.
Wagner, B., Knaevelsrud, C., & Maercker, A. (2007). Post-traumatic growth and opti-
mism as outcomes of an internet-based intervention for complicated grief. Cognitive
Behaviour Therapy, 36, 156–161.
Wagner, B., & Maercker, A. (2007). A 1.5-year follow-up of an Internet-based intervention
for complicated grief. Journal of Traumatic Stress, 20, 625–629.
 Internet-based interventions and support  247
Wagner, B., & Maercker, A. (2008). An Internet-based cognitive–behavioral preventive
intervention for complicated grief: A pilot study. Giornale Italiano di Medicina del
Lavoro ed Ergonomia, 30, 47–53.
Wagner, B., & Maercker, A. (2010). Internet-based intervention for posttraumatic stress
disorder. In Brunet, A., Ashbaugh, R. A., & Herbert, F. C. (Eds.), Internet use in the
aftermath of trauma (pp. 255–267). Amsterdam: IOS Press.
Wagner, B., Schulz, W., & Knaevelsrud, C. (2012). Efficacy of an Internet-based interven-
tion for posttraumatic stress disorder in Iraq: A pilot study. Psychiatry Research, 195,
85–88.
Wittouck, C., Van Autreve, S., De Jaegere, E., Portzky, G., & van Heeringen, K. (2011). The
prevention and treatment of complicated grief: A meta-analysis. Clinical Psychology
Review, 31, 69–78.
18 Family therapy for complicated
grief
David W. Kissane, Talia I. Zaider, Yuelin Li, and
Francesca Del Gaudio

Introduction
Loss never occurs in a vacuum but rather is shared by variously interconnected
people, whose most common constellation is the family (Kissane & Bloch, 1994).
The quality of relationships involved therein proves determinative of the pattern of
adaptation to loss (Kissane, Bloch, Dowe, et al., 1996; Kissane, Bloch, Onghena,
et al., 1996), whether through mutual support and shared grief that steadily heals,
or through distortion and perpetuation of that relational functioning, which, in
turn, exacerbates the intensity and length of mourning. Family therapy has usefully
complemented individual and group approaches to bereavement care (Kissane,
McKenzie, Bloch, Moskowitz, McKenzie, & O’Neill, 2006). Might we entertain
boldly the hope that family therapy initiated during palliative care could prevent
the development of complicated grief? Such a prophylactic approach through a
model of family-centered care is exactly what we have been studying, and we
explore it in this chapter located within the section on therapeutic approaches to
complicated grief.
Let us set the stage for this exploration by locating our work within the lit-
erature of family interventions in bereavement care, review our conceptualiza-
tion of which families cope well and which do more poorly when bereaved, and
provide an overview of our model of therapy. We share preliminary results from
our current National Institutes of Health (NIH)-funded randomized trial of differ-
ent doses of family therapy delivered to at-risk families, whose therapy is com-
menced during anti-cancer treatment and continued into bereavement, once the ill
patient has been lost.

Why family therapy?

There is a long history of family therapists writing about their work with the
bereaved, but few trials conducted to show empirically the benefits of this
approach (Goldstein, Alter, & Axelrod, 1996; Kissane et al., 2006). Lieberman
first reported cases in which family work was necessary to overcome chronic and
complicated grief, for which treatment had been first attempted individually and
failed (Lieberman, 1978). Subsequently working with children and adolescents,
Rosenthal reported beneficial results from family involvement (Rosenthal, 1980).
 Family therapy for CG  249
In contrast to these two early successes, Williams and Polak intruded prematurely
in having the therapist visit relatives with the coroner’s assistant following motor
vehicle deaths (Williams & Polak, 1979), whereas pediatric family therapy
produced improved child behaviors at 1 year following a parent’s death, though
differences waned by the 2-year follow-up (Black & Urbanowicz, 1987). The
latter study did not select at-risk families and struggled with both engagement
and compliance. Goldstein, Alter, and Axelrod (1996) conducted an eight-session
open trial of a psychoeducational family intervention in an outpatient cancer
center, which helped participants adjust.
By the turn of the century, the promise hoped for by Paul and Grosser (1965)
had unfortunately not materialized into a strong body of evidence in support of
family interventions with the bereaved. Our Melbourne-based Family-Focused
Grief Therapy (FFGT) randomized controlled trial showed promise by amelio-
rating distress and depression post-death through its preventive family interven-
tion (Kissane, Lichtenthal, & Zaider, 2007–2008; Kissane, Bloch, McKenzie,
McDowall, & Nitzan, 1998). Since then, a randomized intervention with adoles-
cents and parents with HIV failed to find differences between adolescents who
lost a parent and the non-bereaved at 2-year follow-up (Rotheram-Borus, Stein, &
Lin, 2001; Rotheram-Borus, Weiss, Alber, & Lester, 2005). Other sterling efforts
have been family bereavement programs for parentally bereaved children and
adolescents (Rauch & Muriel, 2004; Sandler et al., 2003). Some promise does
emerge from research across the past decade suggesting that family intervention
has much to offer as a model of bereavement care.

Which theoretical models guide this family therapy?

Let us turn now to theoretical models that inform family processes in mourning
to consider further why family intervention for complicated grief has intuitive
appeal. Four theories form the backbone of such conceptualization: attachment
theory (Bowlby, 1969), cognitive processing theory in adaptation to trauma
(Creamer, Burgess, & Pattison, 1992), group adaptation (Whitaker & Lieberman,
1964), and pre-existing resilience (Bonanno, Wortman, & Nesse, 2004). We will
briefly consider the relevance of each.

Attachment theory
When considering loss from a relational perspective, attachment theory comes
quickly to the fore. The most important relationships are generally found in fami-
lies, whether nuclear, family-of-origin, or extended family (Shaver & Tancredy,
2001). The nature of the bonds of attachment strongly influence the experience
of grief (Ainsworth & Eichberg, 1991). When the emotional impact of a loss is
shared among family members, restorative coping responses can be activated as
relatives comfort and support one another. Family therapy facilitates both ele-
ments of the dual-process model (Stroebe & Schut, 2001) through inviting shar-
ing of grief alongside improved family functioning, in which communication,
250  David W. Kissane et al.
cooperation, and mutual support are enhanced. Patterns of relationship transmit-
ted across generations through the family exhibit styles of attachment that either
facilitate or hinder adaptive mourning.

Cognitive processing and meaning construction

Information about the death and integration of an understanding of what this
means follows successful cognitive processing (Janoff-Bulman, 1989), including
making sense of events in accordance with one’s previously established belief
system, as well as modification of the person’s assumptive world (Parkes, 1972,
1998), a schema of ideas, values, attitudes, and beliefs that each person organ-
izes about his or her life in the world, to accommodate the new events. Like any
trauma, illness and death disrupt the assumptive world schema. At a family level,
disclosure of thoughts and feelings helps shape the assumptive worldviews of the
family as a whole (Janoff-Bulman, 1989; Janoff-Bulman & Berg, 1998), lead-
ing to cognitive reappraisal as either confrontation or avoidance strategies unfold
within the family.
How each family communicates and resolves differences of opinion impacts
dynamically on each individual’s cognitions. Families can challenge negative con-
clusions and guide the integration of positive meaning (Folkman & Moskowitz,
2000). Family traditions influence the regulation of grieving, recognizing when
some degree of avoidance is healthy, but too much is detrimental. Families pursue
cognitive reframing iteratively, using the diverse views of members to test out
options, solve problems, and mutually support each other in finding new meaning
and coping adaptively. In this sense, the meaning attributed to role change within
families (Neimeyer, 2011) is a key determinant of adaptive outcome.

Group adaptation
Group discussions shift dynamically between enabling and restrictive solutions,
as some members offer a constructive suggestion to resolve an issue, while others
urge caution derived from a more fearful viewpoint (Whitaker & Lieberman,
1964). The group, in this case the family, grapples with these options. As debate
unfolds, the family seeks consensus, with adaptive choices generally resulting
from constructive views. Alas, sometimes a dominant person may impose a
deleterious point of view; alternatively, more indecisive individuals can be con-
vinced by the majority. Any difference of opinion can generate ongoing conflict,
these disagreements splitting the family and reducing its teamwork. Just as the
cohesiveness of a psychotherapy group is the hallmark of its effectiveness in
promoting the development and maturity of its membership (McKenzie, 1995),
so too with families.

Harnessing family resilience

Robustness and character strength are key ingredients when families are con-
fronted with the loss of a loved one (Bonanno et al., 2004; Boss, 2006; Shapiro,
 Family therapy for CG  251
2008). Use of evident strengths is both cogent and strategic, for it avoids criti-
cism of the family and, indeed, affirms one potential pathway through which it
can heal its pain (Zaider & Kissane, 2007). The family provides its own natural
support network, through which its constituents can be sustained by its inherent
resiliency (Boss, 2006). We broaden the conception of family to that of the fictive
kin, those available persons who prove to be supportive, concerned, and willing to
help (Landau, 2007). Close friends, neighbors, or members of the extended family
can all bring their generosity, compassion, love, insight, and support to enrich the
circle of care.
Thus, affirming any apparent strengths of a family becomes a key dimension
of therapy. No matter how dysfunctional they might appear, some strength can
be identified in every family! This is endorsed as a dimension of their resilience.

Can we recognize those families in greater need of help

during palliative care?
In the palliative care setting, before a patient with life-threatening illness dies,
family members can be invited to complete a Family Relationships Index (FRI)
(Moos & Moos, 1981), a 12-item true–false scale, which gives information about
family cohesiveness, expressiveness of thoughts and feelings, and conflict resolu-
tion, and has good sensitivity to identify families at risk of psychosocial morbidity
during bereavement (Edwards & Lavery, 2005; Kissane et al., 2003). Those fami-
lies considered at risk are invited to attend a family meeting, where their issues
and concerns about caring for the dying family member can be appraised and
continuing family therapy contracted with those present (Kissane, 2000). Therapy
readily continues into bereavement once a therapeutic alliance is established.
Repeated studies have confirmed a typology of families based on their rela-
tional functioning as defined by the FRI (Kissane et al., 1996, 2003). Two types
of families are well functioning. The first is termed supportive, and these families
are highly cohesive, communicate readily, comfort and support one another, are
free of conflict, and have a resilience that protects their membership from psychi-
atric disorders and complicated grief. The second well-functioning type is termed
conflict resolving, because their open communication and high cohesion protect
their members from difficulties, despite prominent differences of opinion. They
are also free of psychiatric disorder during bereavement. Neither supportive nor
conflict-resolving families are in need of preventive family therapy.
Three types of families have reduced cohesiveness and communication, and
are troubled by conflict. Intermediate families have mild reductions in communi-
cation and teamwork, carry some members in need of psychological help, and are
readily amenable to being engaged in therapy. Generally six to eight sessions of
therapy over as many months will help intermediate families share their grief and
protect members from morbid psychiatric outcomes (Kissane et al., 2006).
Of the more dysfunctional family types, sullen families have poor communi-
cation and cohesion, but muted anger; they carry the highest rates of depression
(anger turned in) and accept help, generally needing 8–12 sessions of therapy
over 12–18 months to prevent complicated grief (Kissane et al., 2007–2008). In
252  David W. Kissane et al.
contrast, hostile families are fractured by conflict, use distance as their means
of survival, and may not be willing to come together for therapy. If they can be
engaged, therapists are wise to set achievable goals, but can expect 10–16 sessions
of therapy over 18 months to be needed to prevent morbid outcomes for those fami-
lies engaged in treatment (Kissane et al., 2006). An acceptable compromise is to
work with an accessible part of such a family and help those open to this assistance.
This empirical typology helps conceptually to recognize families where attach-
ment processes are more disturbed and accurately predicts rates of psychiatric
disorder (Kissane & Bloch, 2002). However, we avoid classifying families by
these names clinically, lest harm be done by pejorative labeling. It is sufficient to
say to families that experience teaches us the value of meeting with the family as a
whole to help patients and their caregivers manage the illness and care provision.

Engaging bereaved families

When a bereaved person presents for individual therapy, it is possible to ask them
to bring along a relative to help extend the family story and deepen understanding
about what has happened (Kissane & Hooghe, 2011). This approach is consistent
with family therapy theory, in which one member serves as the “symptom bearer”
or apparent focus of concern and becomes the rationale for inviting other relatives
to assemble. Therapists may need to reach out actively to others to ask for their
help. A wise therapist works with whoever is available and later invites outliers to
attend, telling them something of the dynamics discovered and emphasizing the
role they could take up to the benefit of the family or ill family member.
Similarly, family therapy can be conducted as a mode of bereavement care
after unexpected deaths, including traumatic or violent deaths, in pediatric loss,
and whether the death was that of a child, sibling, or parent. Although our research
program has been focused on the cancer setting, its extension into other clinical
settings becomes an important feature of its future dissemination.

Ethical aspects of family-centered care

Whereas cancer patients are the primary focus of care in oncology, their family
members become second-order patients in the palliative and terminal care set-
ting. Indeed, the philosophy of palliative care includes caring for the bereaved.
Practical issues present ethical dilemmas when family care is pursued, including
the competing needs of individuals or subsystems within the family, maintaining
confidences, family secrets, and the boundaries of therapy. A fuller discussion of
these many issues is found in our book on FFGT (Kissane & Bloch, 2002).

Conducting therapy in the home

Arranging therapy in the home of the primary caregiver is common in the hos-
pice setting. This practice increases the ability of the ill patient to take part in
early sessions despite his or her frailty and thus become known to the therapist
 Family therapy for CG  253
before death intervenes. Families are very appreciative of this and the therapist is
later able to bring the deceased back figuratively into the therapy, given personal
knowledge of the person. The wishes and hopes of the lost relative can be power-
fully used to motivate the family to sustain efforts at mutual support. When high
conflict is recognized in families, therapists are wise to conduct the initial assess-
ment sessions on neutral territory rather than the home, until confidence develops
about its safety in the home.

Goals of family therapy

Therapists work to facilitate a constructive dialogue that tells the family’s story of
illness and coping with loss. The following specific goals are pursued:

1 to understand the impact that the illness or loss has on family life and coping,
not only for individuals but also for the family as a whole;
2 to examine relationship patterns between family members, contrasting close
alliances with distant relationships, and considering their contributions to
mutual support or the development of conflict;
3 to clarify any transgenerational patterns in family lifestyle, recognizing those
that continue to be adopted and those left behind, in the process naming
family strengths and reaching agreement about any perceived vulnerabilities;
and
4 to foster their mutual support of each other as they mourn and invite their
overt choice about adaptive relational styles that will enhance coping.

Approach to family assessment and therapy

Therapists join initially with each individual through a linear dialogue that learns
who each person is, their place and role in the family, their contribution to caregiv-
ing, and their hopes, expectations, and concerns about the family. Engaging the
family as a whole, however, is dependent on circular rather than linear questions;
the former style of questioning invites each individual, in turn, to step into the
shoes of others and share perceptions about their coping, strengths, vulnerabili-
ties, and understanding of the illness and prognosis, including their acceptance of
what the future holds (Dumont & Kissane, 2009). As understanding is developed,
this initial focus on the status of the illness is shifted to a focus on how the family
communicates about this, works as a team, and resolves differences in opinions
and values. Through the use of integrative summary comments and reflexive
questions, the therapist searches to understand how each family is emotionally
supportive of one another, what challenges it grapples with, what hopes and aspi-
rations they have, and what fears, issues, or concerns are worthy of continued
focus. This agenda setting builds consensus about what might be accomplished
through continued family meetings.
Patterns of family relationship and styles of coping can be made explicit
through examination of these across the generations, thus recognizing traditions,
254  David W. Kissane et al.
values, and ways of coping with stress, change, and death (Kissane & Bloch,
2002). Families will sometimes draw confidence from this; at other times, they
will elect to choose alternative styles of relating and coping. Therapist skill is
needed to recognize and name these patterns, which facilitate the family’s devel-
opment of insight and resultant opportunity to make choices. This technique
proves non-critical as families recognize how easily the ways of prior generations
can be repeated.
Hypothetical consideration of what is sought in the future (in 1, 6, or 12 months)
invites consideration of the family’s priorities during the patient’s remaining life,
discussion of death, and how the family will cope. Patients can poignantly express
their wishes and concerns for their loved ones, expressing gratitude, aspirations,
and bequests alongside special requests for family cohesion, mutual support, and
care in the years ahead. These comments in the final months or weeks of a dying
person’s life can be powerfully recalled during bereavement work.
Grief work then continues in bereavement with sustained focus on the family’s
relational goals, mutual support, and coping with the loss. Progress is affirmed,
with the emphasis on active sharing of grief countering any tendencies toward
avoidance and social withdrawal. The time between sessions is gradually extended
from monthly, to sessions every 2 and then 3 months, until termination of therapy
becomes appropriate.

Family therapy and complicated grief

In this chapter, we have adopted the definition of complicated grief as prolonged
grief whose intensity persists rather than beginning to wane by 6 months of
bereavement (Prigerson et al., 2009). Consensus seems to be emerging that some
form of separation distress is expressed as continued longing for or yearning for
the deceased, which disrupts the functionality of daily life and persists beyond 6
months of bereavement. Using a DSM-style format, accompanying symptoms
have varied between counts of four present out of eight, or five out of nine addi-
tional symptoms. As discussed further below, we used the 11-item Consensus
Criteria for Complicated Grief scale provided by Prigerson in 2005, which
required four out of eight additional symptoms reflecting chronic grief.
We closed recruitment at 170 families in our current trial of FFGT, comprising
some 616 consenting subjects at baseline. Table 18.1 provides an illustration of
their sociodemographic features, including mean age, gender, race and ethnicity,
and marital status. Our recruitment rate was 27% of eligible families.

Improvement in family communication

As family therapy proceeds, we have studied each member’s perception of their
family’s communication as a process measure of therapy to see how this changes
session by session. The dominant patterns are improvement in communication or
maintenance of a high level of communication, where this was present from the
beginning. Linear mixed-effects modeling, accounting for clustered family data,
was used to determine whether family members’ perceptions of communication
 Family therapy for CG  255
Table 18.1 Demographics of participants from 170 families in our current randomized
controlled trial of family-focused grief therapy (FFGT)

Sample characteristics (n = 616) Patients (n = 130) Other family members (n = 486)

Mean age (range) in years 55.6 (21–92) 42.9 (12–85)

Gender
Male 54 (42%) 192 (40%)
Female 76 (58%) 294 (60%)

Racea
White (non-Hispanic) 96 (73%) 360 (74%)
Hispanic 20 (15%) 49 (10%)
Black (African American) 11 (8%) 50 (10%)
Asian 4 (3%) 16 (3%)
Other/unknown 2 (1%) 11 (2%)

Marital status
Married/living with partner 99 (76%) 267 (55%)
Single 17 (13%) 184 (38%)
Divorced/separated 10 (8%) 22 (4%)
Widowed 4 (3%) 13 (3%)

Notes
Not all patients participated in therapy.
a Three subjects identified as both Hispanic and Black.

changed across sessions. Results suggested that, overall, family members per-
ceived a significant increase in communication across FFGT sessions (β = 1.26,
standard error = 0.18, t = 7.07, p < 0.001).

Promise in ameliorating and preventing depression

Table 18.2 shows the comparison for individual subjects of mean scores at base-
line and after 6 months of bereavement on the Beck Depression Inventory-II
(BDI) questionnaire (Steer, Ball, Ranieri, & Beck, 1999; Steer, Brown, Beck, &
Sanderson, 2001), with these scores further broken out by family subtype. Clearly
scores for depression reduce with time, but mean scores for therapy arms reduce
to a greater degree. Selecting the 15% of subjects most distressed at baseline on
the Brief Symptom Inventory-global scale as a selection criterion for distress,
significant reductions in their BDI depression occurred, with reduction of 5.3
points on the BDI with six sessions of family therapy (p = 0.024 compared with
usual care) and 8.6 points with 10 sessions (p = 0.010) in a linear mixed-effects
model comparing the change scores across intervention conditions, adjusting for
baseline BDI scores.
256  David W. Kissane et al.
Table 18.2 Mean and standard deviation scores on the Beck Depression Inventory (BDI)
for 220 subjects after 6 months’ bereavement displayed by both intervention
arm and family type

Number completed, Number completed, mean

Randomization mean (SD) BDI (SD) BDI scores at 6
Family type arm scores at baseline months of bereavementa
All family Standard care n = 158, 13.5 (8.9) n = 54, 11.2 (10.2)
members
FFGT 6 sessions n = 184, 14.6 (9.1) n = 89, 9.4 (7.6)
combined
FFGT 10 sessions n = 194, 12.2 (9.2) n = 77, 9.8 (8.7)
Intermediate Standard care 11.6 (8.1) 10.9 (10.7)
families
FFGT 6 sessions 12.4 (8.9) 8.1 (6.8)
FFGT 10 sessions 15.3 (9.4) 12.4 (10.1)
Sullen families Standard care 14.1 (9.0) 13.0 (11.6)
FFGT 6 sessions 14.4 (8.1) 9.8 (8.1)
FFGT 10 sessions 9.4 (8.0) 7.8 (6.7)
Hostile families Standard care 16.8 (9.5) 13.7 (9.5)
FFGT 6 sessions 18.4 (11.7) 10.4 (6.1)
FFGT 10 sessions 11.8 (8.8) 7.9 (8.4)

a Data available for only those subjects completing this phase of study by May 2011.

Potential to prevent complicated grief disorder

Using the 11-item measure Complicated Grief Consensus Criteria (copyright
Prigerson & Maciejewski, 2005, personal communication), complicated grief
was diagnosed when (A) beyond 6 months of bereavement, (B) a distressing or
disrupting sense of longing or yearning for the deceased persisted, (C) with an
additional four out of eight symptoms of (1) difficulty accepting the loss, (2) dif-
ficulty moving on, (3) numbness or difficulty connecting with others, (4) loss of
trust, (5) bitterness, (6) meaninglessness/emptiness, (7) future without purpose,
(8) feeling on edge and easily startled, and (D) this state caused functional impair-
ment in social, occupational, or domestic life.
For the first 174 subjects reaching 6 months of bereavement, we found a
rate of complicated grief of 25% among individuals receiving standard care
compared with 16% for those receiving family therapy. These data raise the
possibility that family therapy will reduce rates of development of complicated
grief preventively.

Therapy processes
How do therapists achieve these results? Close examination of family therapy
sessions, along with feedback elicited from the therapists, has confirmed the
 Family therapy for CG  257
strategies that therapists follow. Fidelity coding of the first three sessions (n = 144)
of recorded therapy delivered to 74 families (299 individuals) by 32 therapists
made use of the FFGT fidelity coding measure (Chan, O’Neill, McKenzie, Love,
& Kissane, 2004). Inter-rater reliability was satisfactory at 88%.
The therapist behaviors that were rated could be broadly classified as:

1 engaging/joining with the family (e.g., “The therapist elicited expectations

for the therapy from family members”);
2 conducting a relational assessment of family functioning (e.g., “The therapist
asked about family communication”);
3 reviewing multigenerational influences (e.g., “The therapist attempted to link
family patterns across the generations”);
4 exploring family identity and values (e.g., “The therapist attempted to elicit
information on an overall family motto or tradition”); and
5 maintaining a realistic focus of therapy (e.g., “The family and therapist iden-
tified shared goals and concerns to become the focus of family meetings”).

The proportion of family sessions featuring these five processes ranged from
78% (engaging the family successfully) to 59% (creating goals for therapy). Some
98% of assessments elicited the story of the illness, 97% elicited family concerns,
91% identified patterns of relating, 74% a comprehensive discussion about family
communication, 75% the family’s capacity for teamwork, 72% reinforced family
strengths, 66% clarified family roles and values, and 60% beliefs. Less use was
made of summaries (42%), family mottos (32%), exploration of family conflict
in the assessment phase (32%), and the formalization of a comprehensive family
treatment plan (20%). The last may be partly understood as therapists waiting
until supervision to fully formulate their treatment plan.

Challenges identified by the therapists

As part of our peer supervision process after every session, therapists com-
pleted structured process notes in which they were asked to identify challenging
moments during the session. The three most common types of challenges identi-
fied by therapists were:

1 establishing and maintaining boundaries in the home setting (e.g., “Family

members come and go here, adding to the chaos and stress”);
2 attending to grief and open discussion of death and dying (e.g., “Discussing
spiritual beliefs and death and dying was challenging”); and
3 determining a useful focus of therapy when there are longstanding familial
issues (e.g., “There was too much to talk about and not enough time. We had
a marathon session”).

The following selected statements from therapists poignantly capture some of the
challenging aspects of this work:
258  David W. Kissane et al.
From a personal perspective, realizing the degree of Robert’s illness had a
significant impact. At different moments during the session, Robert felt sick,
rested his head between his hands, down to his knees, and closed his eyes.

I felt uncomfortable asking them to pierce their veil of optimism by con-

cretely contemplating a future hypothetical where she is dead if this isn’t
something her doctors are talking to her about (though I did it anyway!).

Working with the reality that this is such a nice, loving family, where the
children are going to lose their father to a terrible illness, is very hard.

When Ellen asked, “How will I know when it is time? I don’t feel ready yet,”
my initial reaction was panic. What could I possibly say? As I looked into her
eyes, it became clear why we were all here. It was hard to sit through their
pain.

Despite these challenges, the therapist accompanies the family through the multi-
ple losses associated with advanced disease and ultimately death, often getting to
know their ill relative rather intimately at the end of his or her life. Almost without
exception, therapists were able to attend the funeral, which furthers their involve-
ment with the family. This continuity of care from palliative care into bereave-
ment is invaluable to the family and empowers the therapist to join the family’s
experience of loss and “put in circulation” (White, 1988, p. 24) the memory of the
deceased family member.

Discussion
The data from our trial of FFGT delivered preventively to at-risk families enrolled
during palliative care and continued into bereavement show promise for the
capacity of family care to prevent complicated grief and depression arising in
bereavement. Most families show profiles of increased or sustained family com-
munication over therapy sessions. Our data are in keeping with the literature
that suggests that family therapy is an important adjunct to individual and group
therapy models of bereavement care (Lieberman, 1978).
FFGT is distinguishable from other approaches to the treatment of complicated
grief (e.g., Kavanagh, 1990; Shear, Frank, Houck, & Reynolds, 2005) in two ways:
(1) FFGT focuses on prevention, and therefore targets at-risk families prior to the
death of a loved one, and (2) FFGT privileges the family, rather than the sup-
portive relationship with an individual therapist, as the preferred context in which
grief is processed. The cohesive and well-attuned family will mobilize effectively
around its vulnerable members in the setting of bereavement. This might include
recognizing when one party requires individual professional attention and facili-
tating such support-seeking behavior (e.g., when a pre-existing psychiatric condi-
tion is exacerbated in the setting of bereavement). The fractured, conflict-ridden,
or non-communicative family deprives its individual members of a key resource
for processing shared grief, and increases the risk of avoidance, isolation, and
 Family therapy for CG  259
prolonged distress. The hypothesized mechanism of change in FFGT is therefore
the strengthening of family bonds. The therapist specifically looks to improve
three areas: constructive communication, capacity to tolerate and negotiate differ-
ences, and collaborative problem solving. Both arms of the dual process model of
mourning are effectively attended to by the family as a whole.
Because of the therapist’s involvement with the family through a time made
difficult by one of their number dying, regular peer supervision proves helpful
to discuss clinical challenges and process personal grief reactions (Zaider &
Kissane, 2009). Contact and exchange of information with the medical team is
also encouraged to keep the therapist aware of disease-related developments, the
prognosis, hospital admissions, or test results. Our model of family therapy has
been taught to many social workers, psychologists, and psychiatrists. We believe
that competent clinicians find little difficulty in understanding and applying the
model to bereaved families. By improving communication about illness and death
together with the resultant grieving and coping, mutual support and teamwork
are fostered. Families are helped to tolerate differences of opinion, optimize their
functioning as a whole, and share their grief together.
Our model has drawn considerable interest from countries where family tradi-
tions are strong and decision making is family centered. It has a natural fit with
Japanese, Chinese, South Asian, and Mediterranean families. It is sensitive to
the cultural needs of families, yet helps blended families to make sense of their
past and respect the strengths of both families of origin. Exploration of religious
traditions and the family’s use of ritual are grist to the mill. Families are especially
appreciative of therapy in the home, beginning during palliative care, so that the
therapist gets to know the dying patient and can later recall their comments to
the bereaved. This continuity of care has merit, but does not prohibit the recruit-
ment of family members when therapy begins after death. Thus FFGT has util-
ity as a model of family-centered care not only during hospice care, but also in
bereavement.

Conclusion
In this chapter, we have shown evidence of the promise for family therapy to
both ameliorate and prevent complicated grief. We recognize that individual or
group therapy for the bereaved may be all that is possible in some geographic
settings. Nevertheless, family therapy can be extraordinarily complementary and,
for many, it could be the primary mode of therapeutic intervention.

Acknowledgements
This research has been supported by the National Research Council of Australia,
the Bethlehem Griffiths Research Foundation, and the National Institutes of
Health (R01 CA 115329 DW Kissane, Principal Investigator; and R03 CA138131
TI Zaider, Principal Investigator). We thank the many clinicians, therapists,
research collaborators, and colleagues who have supported this work across two
decades.
260  David W. Kissane et al.

References
Ainsworth, M. D. S., & Eichberg, C. G. (1991). Effects on infant–mother attachment of
mother’s experience related to loss of attachment figure. In Stevenson-Hinde, J., &
Marris, P. (Eds.), Attachment across the life cycle (pp. 160–183). New York: Routledge.
Black, D., & Urbanowicz, M.  A. (1987). Family intervention with bereaved children.
Journal of Child Psychology and Psychiatry, 28(3), 467–476.
Bonanno, G. A., Wortman, C. B., & Nesse, R. M. (2004). Prospective patterns of resilience
and maladjustment during widowhood. Psychology and Aging, 19(2), 260–271.
Boss, P. (2006). Loss, trauma and resilience: Therapeutic work with ambiguous loss. New
York: Norton.
Bowlby, J. (1969). Attachment and loss, vol. 1: Attachment. New York: Basic Books.
Chan, E.  K., O’Neill, I., McKenzie, M., Love, A. & Kissane, D. (2004). What works
for therapists conducting family meetings: Treatment integrity in family focused
grief therapy during palliative care and bereavement. Journal of Pain and Symptom
Management, 27(6), 502–512.
Creamer, M., Burgess, P., & Pattison, P. (1992). Reaction to trauma: A cognitive processing
model. Journal of Abnormal Psychology, 101(3), 452–459.
Dumont, I., & Kissane, D.  W. (2009). Techniques for framing questions in conducting
family meetings in palliative care. Palliative Supportive Care, 7(2), 163–170.
Edwards, B., & Lavery, V. (2005). Validity of the Family Relationships Index as a screen-
ing tool. Psychooncology, 14, 546–554.
Folkman, S., & Moskowitz, J.  T. (2000). Positive affect and the other side of coping.
American Psychologist, 55(6), 647–654.
Goldstein, J., Alter, C.  L., & Axelrod, R. (1996). A psychoeducational bereavement-
support group for families provided in an outpatient cancer center. Journal of Cancer
Education, 11(4), 233–237.
Janoff-Bulman, R. (1989). Assumptive worlds and the stress of traumatic events:
Applications of the scheme construct. Social Cognition, 7, 113–136.
Janoff-Bulman, R., & Berg, M. (1998). Disillusionment and the creation of value: From
traumatic losses to existential gains. In Harvey, J. (Ed.), Perspectives on loss: A source-
book (pp. 35–47). Philadelphia, PA: Brunner Mazel.
Kavanagh, D. J. (1990). Towards a cognitive–behavioural intervention for adult grief reac-
tions. British Journal of Psychiatry, 157, 373–383.
Kissane, D. (2000). Family grief therapy: A model for working with families during pallia-
tive care and bereavement. In Baider, L., Cooper, C., & De-Nour, A. (Eds.), Cancer and
the family (pp. 175–197). Chichester: Wiley.
Kissane, D., & Bloch, S. (1994). Family grief. British Journal of Psychiatry, 164, 728–740.
Kissane, D., & Bloch, S. (2002). Family focused grief therapy: A model of family-centred
care during palliative care and bereavement. Buckingham: Open University Press.
Kissane, D., Bloch, S., Dowe, D., Snyder, R., Onghena, P., McKenzie, D., & Wallace,
C. (1996). The Melbourne family grief study I: Perceptions of family functioning in
bereavement. American Journal of Psychiatry, 153, 650–658.
Kissane, D. W., Bloch, S., McKenzie, M., McDowall, A. C., & Nitzan, R. (1998). Family
grief therapy: A preliminary account of a new model to promote healthy family func-
tioning during palliative care and bereavement. Psychooncology, 7(1), 14–25.
Kissane, D., Bloch, S., Onghena, P., McKenzie, D., Snyder, R., & Dowe, D. (1996). The
Melbourne family grief study II: Psychosocial morbidity and grief in bereaved families.
American Journal of Psychiatry, 153, 659–666.
 Family therapy for CG  261
Kissane, D.  W., & Hooghe, A. (2011) Family therapy for the bereaved. In Neimeyer,
R. A., Harris, D. L., Winokuer, H. R., & Thornton, G. F. (Eds.), Grief and bereavement
in contemporary society: Bridging research and practice (pp.  287–302). New York:
Routledge.
Kissane, D., Lichtenthal, W., & Zaider, T. (2007–2008). Family care before and after
bereavement. Omega, 56, 21–32.
Kissane, D. W., McKenzie, M., Bloch, S., Moskowitz, C., McKenzie, D. P., & O’Neill, I.
(2006). Family focused grief therapy: A randomized, controlled trial in palliative care
and bereavement. American Journal of Psychiatry, 163(7), 1208–1218.
Kissane, D.  W., McKenzie, M., McKenzie, D.  P., Forbes, A., O’Neill, I., & Bloch, S.
(2003). Psychosocial morbidity associated with patterns of family functioning in pal-
liative care: Baseline data from the Family Focused Grief Therapy controlled trial.
Palliative Medicine, 17(6), 527–537.
Landau, J. (2007). Enhancing resilience: Families and communities as agents for change.
Family Process, 46(3), 351–365.
Lieberman, S. (1978). Nineteen cases of morbid grief. British Journal of Psychiatry, 132,
159–163.
McKenzie, K. R. (1995). Rationale for group psychotherapy in managed care. In McKenzie,
K. R. (Ed.), Effective use of group therapy in managed care (pp. 1–25). Washington,
DC: American Psychiatric Press.
Moos, R.  H., & Moos, B.  S. (1981). Family environment scale manual. Stanford, CA:
Consulting Psychologists Press.
Neimeyer, R. (2011). Reconstructing meaning in bereavement. In Watson, M., &
Kissane, D. (Eds.), Handbook of psychotherapy in cancer (pp. 247–257). Chichester:
Wiley-Blackwell.
Parkes, C. (1972). Bereavement: Studies of grief in adult life. London: Tavistock.
Parkes, C. (1998). Bereavement studies of grief in adult life (3rd edn.). Madison, CT:
International University Press.
Paul, N., & Grosser, G. (1965). Operational mourning and its role in conjoint family
therapy. Community Health Journal, 1, 339–345.
Prigerson, H. G., Horowitz, M. J., Jacobs, S. C., Parkes, C. M., Aslan, M., Goodkin, K., et
al. (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for
DSM-V and ICD-11. PLoS Med, 6(8), e1000121.
Rauch, P. K., & Muriel, A. C. (2004). The importance of parenting concerns among patients
with cancer. Critical Reviews in Oncology/Hematology, 49(1), 37–42.
Rosenthal, P. A. (1980). Short-term family therapy and pathological grief resolution with
children and adolescents. Family Process, 19(2), 151–159.
Rotheram-Borus, M. J., Stein, J. A., & Lin, Y. Y. (2001). Impact of parent death and an
intervention on the adjustment of adolescents whose parents have HIV/AIDS. Journal
of Consulting and Clinical Psychology, 69(5), 763–773.
Rotheram-Borus, M.  J., Weiss, R., Alber, S., & Lester, P. (2005). Adolescent adjust-
ment before and after HIV-related parental death. Journal of Consulting and Clinical
Psychology, 73(2), 221–228.
Sandler, I. N., Ayers, T. S., Wolchik, S. A., Tein, J. Y., Kwok, O. M., Haine, R. A., et al.
(2003). The family bereavement program: Efficacy evaluation of a theory-based preven-
tion program for parentally bereaved children and adolescents. Journal of Consulting
and Clinical Psychology, 71(3), 587–600.
Shapiro, E. R. (2008). Whose recovery, of what? Relationships and environments promot-
ing grief and growth. Death Studies, 32(1), 40–58.
262  David W. Kissane et al.
Shaver, P., & Tancredy, C. (2001). Emotion, attachment and bereavement: A conceptual
commentary. In Stroebe, M., Hansson, R., Stroebe, W., & Schut, H. (Eds.), Handbook
of bereavement research: Consequences, coping and care (pp. 63–68). Washington,
DC: American Psychological Association.
Shear, K., Frank, E., Houck, P. R., & Reynolds, C. F. III. (2005). Treatment of complicated
grief: A randomized controlled trial. JAMA, 293(21), 2601–2608.
Steer, R., Ball, R., Ranieri, W., & Beck, A. (1999). Dimensions of the Beck Depression
Inventory-II in clinically depressed outpatients. Journal of Clinical Psychology, 55(1),
117–128.
Steer, R., Brown, G., Beck, A., & Sanderson, W. (2001). Mean Beck Depression
Inventory-II scores by severity of major depressive episode. Psychological Reports,
88(3), 1075–1076.
Stroebe, M., & Schut, H. (2001). Models of coping with bereavement: A review. In Stroebe,
M., Hansson, R., Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research:
Consequences, coping and care (pp. 375–403). Washington, DC: APA Books.
Whitaker, D. S., & Lieberman, M. A. (1964). Psychotherapy through the group process.
Chicago: Adline.
White, M. (1988). Saying hello again. The incorporation of the lost relationship in the
resolution of grief. Dulwich Centre Newsletter, Spring (p. 24). South Australia.
Williams, W. V., & Polak, P. R. (1979). Follow-up research in primary prevention: A model
of adjustment in acute grief. Journal of Clinical Psychology, 35(1), 35–45.
Zaider, T., & Kissane, D. (2007). Resilient families. In Monroe, B., & Oliviere, D. (Eds.),
Resilience in palliative care. Oxford: Oxford University Press.
Zaider, T., & Kissane, D. (2009). The assessment and management of family distress
during palliative care. Current Opinion in Supportive and Palliative Care, 3(1), 67–71.
19 Brief group therapies for
complicated grief
Interpretive and supportive approaches
William E. Piper and John Ogrodniczuk

In 1986, several members of our team of researchers and clinicians (W. Piper,
M. McCallum, and H. Azim) worked together in the outpatient Walk-in Clinic of
the Department of Psychiatry, at the University of Alberta Hospital in Edmonton,
Alberta, Canada. Because of the high volume of patients seen in our Walk-in
Clinic, approximately 2,000 new referrals each year, most patients were treated
with group therapy rather than individual therapy. To obtain a more accurate
account of the utilization of group therapy, we decided to conduct an informal
in-house review of the objectives and apparent achievements of several therapy
groups provided in the clinic, particularly the time-limited short-term therapy
groups, given their cost-effective potential.
Our review revealed that the short-term therapy groups appeared to be experi-
encing task overload. That is, the groups seemed to be trying to achieve too many
things: crisis intervention; support for day-to-day problem solving; assessment
of suitability for long-term group therapy; training for new therapists who were
inexperienced with group therapy; and treatment for symptom reduction, insight,
and development of interpersonal skills and personality change. Thus, in our feed-
back to the clinic, we concluded that the short-term therapy groups would prob-
ably benefit from having a narrower focus with a more limited but more realistic
set of objectives. Because the topic of death loss was very prevalent among the
patients who were participating in the short-term therapy groups, we decided to
conduct a pilot therapy group to determine how well a short-term therapy group
that focused on death losses and that had more limited objectives would work. As
it turned out, the group worked very well. After discussions with the therapists
and clinic administrators, we began providing loss groups for patients who met
criteria for complicated grief (CG).
In general, the current chapter focuses on the treatment of CG by means of
short-term group therapies. In particular, the chapter focuses on our program of
providing one of two models of brief group therapies for CG (Piper, Ogrodniczuk,
Joyce, & Weideman, 2011). Unfortunately, although CG is a familiar condition,
it lacks a standard definition or a standard set of diagnostic criteria. Typical
symptoms of CG include shock, denial, sadness, irritability, preoccupation with
the lost person, yearning for the lost person, and searching for the lost person.
In addition, dysfunctional behavior with family, friends, and work associates, as
264  William E. Piper and John Ogrodniczuk
well as health-compromising behaviors such as excessive drinking and exces-
sive smoking, is common. Also common is comorbidity with disorders such as
depression. The presence of symptoms and dysfunctional behavior as described
above, particularly if expressed in high intensity and long duration, is what most
clinicians regard as CG; and that is how we regarded it in our initial work with
the concept.
During the 20 years in which we conducted the three clinical trials, definitions
and criteria for some concepts changed. Fortunately, however, other key concepts
remained virtually constant. Such was the case with Horowitz’s Impact of Event
Scale. The two subscales of this measure (intrusion and avoidance) provided us
with two indicators of CG. Intrusion refers to the degree to which thoughts, feel-
ings, and memories about the lost person intrude upon the day-to-day function-
ing of the patient. Avoidance refers to the patient’s active resistance to thoughts,
feelings, and memories about the lost person. Thus, CG, as defined by Horowitz,
served as both a cut-off criterion for CG (operational definition) and an outcome
variable. For example, in our prevalence study a patient had to attain a score of 10
or above on either the intrusion or avoidance subscale and a score of 2 or above
in one of the six areas of the Social Adjustment Scale – Self-report. In addition,
the death loss had to have occurred at least 3 months prior to the assessment of
the patient.
It has been our impression that losses, the effects of losses on survivors, and
the presence of CG criteria usually do not receive adequate attention in intake
interviews. In part, this is because assessors are required to obtain a large and
diverse amount of information in a limited amount of time. This involves conduct-
ing a thorough mental status examination, an inquiry about relevant diagnostic
criteria, and a review of possible causes and precipitants of the patient’s current
problems. Not much time remains available for assessing loss information. To
circumvent this problem, we attempted to identify a few items from various
loss questionnaires that could successfully detect the presence or absence of CG
(Piper, Ogrodniczuk, & Weideman, 2005). Two such items/questions were found.
They were: “During the past 7 days, pictures about the loss popped into my mind”
and “During the past 7 days, I have tried not to think about the loss.” If the patient
tested positive on either item, the probability that the patient had CG was 0.90.
The assessor could then follow up with a more detailed set of questions to confirm
or disconfirm the initial impression. The fact that CG is not in the Diagnostic and
statistical manual of mental disorders (DSM) probably contributes to the neglect
of the recognition of CG.

Prevalence of CG
Large-scale prevalence studies for CG, that is, those with several hundred sub-
jects, are quite rare. Nevertheless, from the few large-scale studies that have
been conducted it has been estimated that approximately 20% of all bereaved
individuals meet criteria for CG (Zisook & Lyons, 1989–1990). The percent-
age of psychiatric outpatients who meet the criteria for CG may be even higher.
 Brief group therapies for CG  265
Studies have reported estimates ranging from 15% to 33%. The findings from our
own prevalence study (Piper, Ogrodniczuk, Azim, & Weideman, 2001), which
involved 729 patients, were at the high end of this range, 33%. This suggested that
approximately one third of the outpatients who walked through the doors of our
clinics met criteria for CG.

Differences between our two grief therapies and other

therapies in the field
Taken together, the characteristics of our approach to treatment differ from other
approaches to treatment in the literature. First, it was a psychosocial form of treat-
ment, in contrast to a biological form such as medication. Second, it was psycho-
dynamic in orientation with a conceptual emphasis on conflictual components
such as wishes, anxiety, and defenses, and a technical emphasis on clarification,
confrontation, and interpretation. Third, it was a group form of treatment, thereby
involving an entire set of patients in interaction with a therapist and one another,
as opposed to a more private, one-to-one relationship between patient and thera-
pist. Fourth, it was a time-limited form of treatment with a predefined beginning
and ending, as opposed to a form of therapy with an open-ended contract. Fifth,
patients were expected to decide what to talk about at the beginning of each ses-
sion and throughout treatment, rather than being directed by the therapist as in
the case of some cognitive–behavioral approaches. Given these characteristics,
we could have named the treatment Time-Limited, Short-Term, Dynamically-
Oriented Group Therapy. However, for the sake of parsimony, we settled on the
simple labels supportive therapy and interpretive therapy.

Differences between our interpretive and supportive

therapies
The overall objective for interpretive therapy is to enhance patient insight about
repetitive conflicts and trauma associated with their loss. The conflict may be
intrapsychic or external in nature. The overall objective for supportive therapy
is to improve the patient’s immediate adaptation to his or her life situation. In
interpretive therapy, among other techniques, the therapist maintains pressure on
the patients to talk, encourages the patients to explore uncomfortable emotions,
and provides the patients with interpretations. (Interpretations are statements that
make reference to dynamic components such as wishes, fears, and defenses.) In
supportive therapy, among other techniques, the therapist gratifies the patients,
provides guidance, and provides non-interpretive interventions. (Non-interpretive
interventions include brief questions, clarifications, and empathic comments.)

Technical manuals and feedback

It is important to keep in mind that what we have designated as interpretive and
supportive forms of dynamically oriented group therapy represent part of only a
266  William E. Piper and John Ogrodniczuk
few reports in the literature about using group therapy with CG patients. To help
therapists adhere to the intended forms of therapy, we constructed a technical
manual for each of the two forms of therapy. We also provided feedback to each
therapist regarding how successful he or she had been in carrying out the two
forms of treatment as planned. This feedback was presented to the therapists after
every other session in the form of ratings completed by independent raters.
An example of a group therapy being used to treat patients with CG is inter-
personal psychotherapy (IPT; Wilfley, MacKenzie, Welch, Ayers, & Weissman,
2000). Although this approach is well described by Wilfley et al. in their book,
they report no efficacy studies for this form of therapy. Another example of a group
therapy being used to treat patients with CG is described in a study conducted by
Horowitz, Marmar, Weiss, DeWitt, and Rosenbaum (1984). Unfortunately, the
study did not make clear how much time had elapsed between the deaths and
treatment. Most of the variables that were analyzed were process variables and
not outcome variables.
The following material from typical therapy sessions illustrates how a therapist
can respond to the group either interpretively or supportively.

Illustration of therapist’s interpretive statements

(Lengthy silence followed by therapist’s interpretive statements.)

Therapist: I think the group is resisting putting what you are feeling into words.
I wonder if you can put it into words.
Ed: Yes, I’m trying but I can’t find the words.
Therapist: But this is everybody’s difficulty, not just Ed’s. What about others?
Brenda: Well, I feel that our problems are hopeless.
Therapist: That is definitely the feeling, along with feeling sad. You seem to be
on the verge of tears. All of you are in the same boat – feeling hope-
less and sad as it sinks in a sea of tears.
Brenda: Well, our therapist is trying his best but we are just hopeless.
Therapist: You don’t know what to do with liking me and hating me at the same
time. Liking me because I helped you recognize your feelings and
hating me because our group is almost finished and you are losing me.

This interaction illustrated how the therapist can use interpretive technical fea-
tures to maintain pressure on the patients to talk, encourage the patients to explore
uncomfortable emotions, and provide the patients with interpretations of conflict.

Illustration of therapist’s supportive statements

(Lengthy silence followed by therapist’s supportive statements.)

Therapist: Despite the silence, I think that members of the group are working
hard.
Ed: Yes, I’m trying hard and I’m almost able to put things into words.
 Brief group therapies for CG  267
Therapist: Why don’t you take a stab at it? Sometimes if one person is able to
start others are able to start too.
Brenda: Well, I’ve been feeling hopeless and sad but, in a funny way, I’ve also
felt good about what I’ve learned in the group.
Therapist: Yes, this has been happening in spite of the fact that our group will
end next week. I suspect that in some ways we will all miss parts of
each other.

This interaction illustrated how the therapist can use supportive technical
features to gratify the patients, provide guidance, and provide non-interpretive
interventions.

Clinical trials
Since 1986, we have conducted over 90 short-term loss groups. Over one half of
these groups participated in one of three clinical trials conducted by our research
team. We refer to the three as the control, comparative, and composition studies.
Two forms of dynamically oriented short-term therapy (interpretive and support-
ive models) were studied. They were similar to each other in structure (e.g., one
90-minute therapy session per week for 12 weeks), but differed from each other
in style of therapy (e.g., the therapist’s focus on transference in the interpretive
model).
The overall objective of treatment was to enhance patient insight about repeti-
tive conflicts and trauma associated with the losses that are assumed to serve as
impediments to a normal mourning process. A related objective was to help the
patients develop tolerance for ambivalence toward the people whom they have
lost. In the sessions, the therapist attempted to create a climate of tolerable tension
and deprivation. In the interpretive therapy, the therapist attempted to (a) maintain
pressure on the patients to talk, (b) encourage them to explore uncomfortable
emotions, (c) make interpretations about conflicts, (d) direct attention to subjec-
tive impressions of the therapist, (e) make links between the patients’ relation-
ships with the therapist or each other and the patients’ relationships with others in
their lives, (f) focus on the patients and therapist in the here-and-now treatment
situation, and (g) direct attention to the patients’ subjective impressions of others
outside the treatment situation.
In contrast, in supportive therapy the therapist attempted to (a) gratify the
patients, (b) make non-interpretive interventions such as reflections, questions,
and provision of information, (c) provide guidance similar to the role of a kindly
family doctor, (d) engage in problem-solving strategies with the patients, (e) offer
explanations that locate the responsibility for the patients’ difficulties outside the
patient, (f) praise the patients, and (g) display personal information, opinions, and
core values.
We believed that a useful way to conceptualize each of the main features of
interpretive group therapy was to regard each of its features as a dimension. For
some dimensions (e.g., focusing on transference), the higher their level, the more
the therapy was regarded as interpretive. For other dimensions (e.g., gratifying
268  William E. Piper and John Ogrodniczuk
the patients), the lower the level, the more the therapy is regarded as interpre-
tive. Supportive therapy can be conceptualized in a similar way, that is, for some
dimensions (e.g., making clarifications) the higher their level, the more the therapy
is regarded as supportive and for other dimensions (e.g., making interpretations)
the lower the level, the more the therapy is regarded as supportive.

Control trial
The control clinical trial involved only interpretive therapy (Piper, McCallum, &
Azim, 1992). We had not yet begun to offer a supportive form of group therapy for
loss patients. An immediate treatment condition was compared with a waiting-list
control condition.
On the basis of 94 patients from 16 therapy groups, who had been randomly
assigned to the conditions, the findings clearly indicated superior outcomes for
patients in the immediate treatment (control) condition. Examination of the tech-
nical features of treatment confirmed its interpretive nature.

Comparative trial
Although the results for interpretive group therapy were clearly favorable, not
all patients benefited. Stemming from work carried out by Rockland (1989) and
Werman (1984) in individual therapy and following work that we had been pursu-
ing in the case of short-term individual therapy for a wide range of outpatients
in our clinic (Piper, Joyce, McCallum, & Azim, 1998), we suspected that patient
personality variables may influence the outcome of therapy. We developed and
labeled one such personality variable, the patient’s Quality of Object Relations
(QOR). We defined QOR as a person’s internal enduring tendency to establish
certain kinds of relationships that range along an overall dimension from 1
(primitive) to 9 (mature). An assessor conducts a 1-hour interview that focuses
on the nature of the patient’s relationships during three stages of life (childhood,
adolescence, and adulthood). The assessor uses a scoring manual to determine
the overall score. In the individual therapy studies we found a matching effect.
Patients with higher QOR scores tended to improve more in interpretive indi-
vidual therapy, and patients with lower QOR scores tended to improve more in
supportive individual therapy. We decided to check for the presence of this type of
effect in the comparative trial. On the basis of 139 patients in 16 therapy groups,
strong evidence for the matching effect was found (Piper, Debbane, Bienvenu,
& Garant, 1984). High-QOR patients benefited more from interpretive group
therapy, and low-QOR patients benefited more from supportive group therapy.

Composition trial
In the comparative trial, the composition of each therapy group was mixed (het-
erogeneous) in terms of the patients’ QOR scores. We wondered if the matching
effect would be even more pronounced in groups composed of all high-QOR
patients who received interpretive therapy and groups composed of all low-QOR
 Brief group therapies for CG  269
patients who received supportive therapy. This led to our decision to conduct the
composition trial (Piper, Ogrodniczuk, Joyce, Weideman, & Rosie, 2007). Like
other terms in the literature such as complicated grief, the literature is replete with
references to the term composition and yet seems to lack a common definition. For
the purposes of this study, composition was defined as the proportion of patients
in a group with high QOR scores.
Four conditions were created experimentally in the composition trial: (1)
homogeneous, high-QOR interpretive groups; (2) homogeneous, low-QOR
supportive groups; (3) heterogeneous, mixed-QOR interpretive groups; and (4)
heterogeneous, mixed-QOR supportive groups. Our primary hypothesis was that
patients in the two homogeneous conditions would experience greater benefit than
the patients in the two heterogeneous conditions.
On the basis of 135 patients in 18 therapy groups, we found that the best
outcomes were achieved by the homogeneous high-QOR patients who received
interpretive group therapy and the poorest outcomes were achieved by the homo-
geneous low-QOR patients who received supportive therapy. Outcome for the
mixed-QOR groups fell in between. Even though our primary hypothesis did not
receive support, we nevertheless found evidence for an important composition
effect. Using a QOR cut-off score of 4.2, the score that we found to be a useful
differentiator in previous studies, we found that the greater was the proportion
of high-QOR patients in the group, the better was the outcome. This was true
regardless of the patient’s own QOR score or the form of therapy (interpretive or
supportive) that the patient had received. It is possible that the high-QOR patients
beneficially provided peer support and served as models of useful problem-
solving behavior. In their absence, the primitive behaviors of low-QOR patients
may have led to a group culture in which the provision of support and engagement
in problem solving was regarded as intrusive and was met with suspicion and
resistance. These, of course, are speculative ideas, which require future research
testing and confirmation.

Follow-up
Once the efficacy of a form of therapy has been demonstrated, usually through
randomized controlled trials, the question that follows naturally is: Will the
effects last? To address the question concerning lasting effects, follow-up data
are typically examined. However, for many disorders, follow-up data either are
not available or contain significant flaws. Typically, the nature of the follow-up
sample is problematic. Inevitably, some patients do not return for their follow-up
assessments. Thus, the sample will have decreased in size, which will compro-
mise the statistical power of the analyses. The representativeness of the follow-up
sample can also be questioned. Although missing data always create ambiguities,
if the number of missing data is relatively small and there is evidence that the
sample of missing data is similar to the sample of non-missing data, the findings
are usually considered worthy of consideration. We believe that this is the case
concerning the follow-up data from our composition trial.
Achieving lasting benefits from brief psychotherapies for certain disorders has
270  William E. Piper and John Ogrodniczuk
proven to be difficult. Such has been the case for the treatment of depression. This
is relevant to the treatment of patients with CG because of the overlap of symp-
toms with depression (e.g., sadness). Overall, the findings concerning the long-
term benefits of brief therapies for depression have been disappointing. In the
National Institute of Mental Health treatment of depression collaborative study
(Elkin et al., 1989), patients received an average of 16 sessions of treatment. Only
33% of the patients who began therapy met recovery criteria, and nearly 40% of
those relapsed within 18 months. The investigators concluded that “16 weeks of
these specific forms of treatment are insufficient for most patients to achieve full
recovery and lasting remission” (p. 782). Similar negative conclusions have been
made regarding remission rates in the STAR*D clinical trial (Trivedi et al., 2006)
by Fava and colleagues (2004) and in a recent meta-analytic review conducted by
de Maat, Dekker, Schoevers, and de Jonghe (2006). In their review, the remission
rate for cognitive therapy was 38%, and the relapse rate was 27%. Although these
findings have been viewed as promising for advocates of psychosocial therapies,
the relapse rates unfortunately have remained high at approximately 30%.
In the follow-up data from our composition trial, of the 110 completers in the
study, 84 (77%) provided follow-up data. In addition, we compared patients who
provided follow-up data with those who did not on 22 initial status variables.
Only 1 of 22 variables was significant. Patients who did not provide follow-up
data had higher posttherapy Beck Depression Inventory scores (Beck & Steer,
1987) (mean = 33, SD = 12) than patients who provided follow-up data (mean = 25,
SD = 12), t(106) = 2.99, p = 0.003. However, given the number of analyses con-
ducted, this could easily be a result of error. Thus, there was little evidence of
differences between patients who did and patients who did not provide follow-up
data. Before examining the findings from our composition trial, several distinc-
tions among types of follow-up and patient outcome need to be clear:

• Maintenance refers to recovery during treatment followed by maintenance of

recovery during follow-up.
• Relapse refers to recovery during treatment followed by relapse during
follow-up.
• Delayed refers to non-recovery during treatment followed by recovery during
follow-up.
• Non-recovery refers to non-recovery during treatment followed by non-
recovery during follow-up.

These distinctions were used with the data from the 18 therapy groups of our
composition trial. Outcome was represented by the achievement of clinical sig-
nificance for two well-known grief outcome variables in the research literature:
intrusion and avoidance. They were the two subscales from the Impact of Event
Scale (Horowitz, Wilner, & Alvarez, 1979). Clinical significance was determined
by the procedure of Jacobson and colleagues (Jacobson, Follette, & Revenstorf,
1984; Jacobson & Revenstorf, 1988). Recovery was defined as dropping below
the clinical significance cut-off score and relapse as rising above the cut-off
 Brief group therapies for CG  271
score. Next, we checked whether the overall proportion of patients who achieved
clinical significance at posttherapy was different from the proportion of patients
who achieved clinical significance at follow-up. Using the test for a difference
between two dependent proportions, there was a significant increase from post
therapy (30/59 = 0.51) to follow-up (42/59 = 0.71) in the proportion of patients
who achieved clinical significance for intrusion, z(N = 59) = 2.83, p = 0.004. There
was also a significant increase from posttherapy (34/64 = 0.53) to follow-up
(46/64 = 0.72) for avoidance, z(N = 64) = 2.88, p = 0.004. Following this, we deter-
mined the percentages of patients for the four types of outcome described previ-
ously. For intrusion, there were 46% (27/59) maintenance patients, 5% (3/59)
relapse patients, 25% (15/59) delayed-recovery patients, and 24% (14/59) non-
recovery patients. The results for avoidance were very similar with 48% (31/64)
maintenance patients, 5% (3/64) relapse patients, 23% (15/64) delayed-recovery
patients, and 23% (15/64) non-recovery patients. Thus, a substantial percentage
of the sample, nearly 75%, achieved clinically significant improvement on grief
outcomes (intrusion, avoidance) by the end of follow-up, and a smaller proportion
failed to maintain their recovered status or failed to recover at all.
If we focus on intrusion, a total of 71% of the patients achieved maintenance
or delayed recovery, which is almost 75% of the sample. If one considered only
the outcome from pre- to posttherapy, only 46%, which is a little below half of the
sample, would have been regarded as improved. Use of the follow-up data creates
quite a different impression regarding the degree of improvement.

Meta-analytical reviews and their critiques

In the field of psychotherapy, effectiveness is often expressed in terms of effect
size. It is a statistic that expresses the amount of change in a single sample (e.g.,
from pretreatment to posttreatment) or the difference in the amount of change
between two samples (e.g., treated and control group). Because it is expressed
in terms of standard deviation units, comparisons can be made across different
outcomes, treatments, and studies. In the studies cited in this chapter, effect size
is generally defined as the mean of the pretreatment outcome score minus the
mean of the posttreatment outcome score divided by the standard deviation of the
pretreatment score. It is commonly symbolized as d. This approach to reviewing
the literature is commonly referred to as meta-analytic.
During the past decade only a small number of meta-analytic reviews of the
efficacy of treatments for CG have been conducted. This reflects the fact that
only a small number of clinical trials and other outcome studies have been con-
ducted. In the literature, there are many descriptions of group therapy treatments
of CG including some books on cognitive–behavioral therapy (CBT; Rose, 1989)
and IPT (Wilfley et al., 2000). It would have been desirable if the reviews had
provided a consensus concerning the efficacy of loss interventions. However,
this has not been the case. Even less desirable is the fact that reviewers more
often than not have been negatively critical about the methodology and conclu-
sions of other reviewers. At times, the criticism has been sharp and accusatory.
272  William E. Piper and John Ogrodniczuk
For example, Larson and Hoyt (2007) have suggested that Fortner (1999) and
Neimeyer (2000) damaged the reputation of grief counseling by misrepresenting
empirical findings.
Allumbaugh and Hoyt (1999) presented an overall effect size of 0.43 in their
review, which is somewhat smaller than effect sizes commonly reported in psy-
chotherapy outcome studies. Kato and Mann (1999) provided an overview of four
major theories of bereavement with their quantitative analysis. Methodological
flaws in all of the studies made the results hard to interpret. They reported an over-
all effect size of 0.11. They were left with the question whether the interventions
were powerful enough or the sample was too small to detect differences. Larson
and Hoyt (2007) criticized Kato and Mann regarding their method of selecting
studies for their review. In addition, they criticized them for constructing an effect
size for each measure in their studies.
Neimeyer (2000) produced a meta-analytical review that included 23 stud-
ies. The overall effect size was 0.15. Neimeyer’s review generated controversy
because it relied heavily on an unpublished dissertation completed by Fortner in
1999. He used an idiosyncratic statistic called Treatment Induced Deterioration
Effect (TIDE), which had not been subjective to a peer review. Schut, Stroebe,
van den Bout, and Terheggen (2001) provided a review of bereavement interven-
tion research. They concluded that the length of time between the loss and the
onset of treatment was directly related to better outcome. Similarly to Kato and
Mann (1999) they also concluded that most of the studies in their review had
methodological problems.
Currier, Neimeyer, and Berman (2008) provided one of the most comprehen-
sive reviews of the field to date. They included 61 studies in their review. They
concluded that the interventions reviewed had small effect sizes from pretherapy
to posttherapy. In the case of our three clinical trials, Table 19.1 presents the effect
sizes from pretherapy to posttherapy for our two primary outcome variables,
intrusion and avoidance. Thus, our effect sizes tended to be larger than most of
those reported in literature.
Setting aside the rather unpleasant and at times acrimonious tone of many of
the reviews, the critiques have been useful in identifying variables to explore in
future studies involving the treatment of CG. Such moderating variables include
time since the loss, level of CG distress, individual versus group therapy, random
assignment to conditions, and therapist training differences.

Table 19.1  Effect sizes for the two primary outcome variables

Controlled Comparative Composition

Outcome variable (and measure) trial trial trial
Intrusion (intrusion subscale of the Impact 0.56 1.09 0.56
of Event Scale; Horowitz et al., 1979)
Avoidance (avoidance subscale of the 0.70 1.05 0.85
Impact of Event Scale; Horowitz et al.,
1979)
 Brief group therapies for CG  273
Most of these variables seemed to be cited in the literature as afterthoughts
rather than being chosen a priori as potential moderator variables. Given the find-
ing that the effect sizes associated with our two models of group therapy were
larger than most effect sizes reported in the reviews, one reason for this may be
the fact that many of the events in the therapies provided here-and-now oppor-
tunities to be addressed immediately rather than impediments to the process of
therapy. For example, patient lateness or absenteeism may raise concerns about
whether the patient will return. Dropping out may trigger feelings associated with
witnessing a death. These events are usually regarded as troublesome events in a
therapy group. However, in loss groups these are naturally occurring phenomena
that are reminiscent of events related to patients’ losses. Discussion of such events
is often avoided, and such reticence only contributes to similar behaviors of other
patients in the group. Finally, the patients came to the Walk-in Clinic because they
were suffering from symptoms and dysfunction related to CG. They did not come
in response to advertisements about a study in which they could participate and
at the same time receive treatment. No advertisements were used in our studies.
Instead, patients were usually referred by their family physician.

Evidence-based treatment
From time to time, researchers have afforded special importance and status to
aspects of their work if carried out in a particular manner. Currently, it appears
to be very important, if not essential, to be able to refer to one’s work or findings
as evidence based. Two recently published books, Evidence-Based Treatment
for Personality Dysfunction edited by Magnavita (2010) and Evidence-Based
Psychotherapy edited by Goodheart, Kazdin, and Sternberg (2006), thoroughly
reviewed the terminology associated with evidence-based treatments. Early ter-
minology advocated by a task force of the American Psychological Association
initially favored the term validated as in empirically validated treatment. The
criteria required for this designation were stringent (e.g., randomized control
trial design, manualized treatment, and replication). However, because the word
validated conveyed to many people the impression that further research with the
treatment would probably be unnecessary and the stringent criteria created arti-
ficial conditions, it was replaced with the term supported. Later an even broader
category of evidence-based practice was recommended. This again emphasizes
the policy of offering only treatments that are evidence based. The questions that
obviously followed were: What criteria define evidence based and what criteria
do not? A debate has ensued over this issue. Some therapists restrict evidence to
mean the findings of randomized clinical trials. Other therapists accept correla-
tional (naturalistic) and case studies as well as randomized clinical trials.
Over 55 years ago, Cronbach and Meehl (1955) published their classic paper
on the topic of construct validity. They argued that the criteria for a construct
consisted of the entire set of relationships with other constructs. They referred
to this pattern of relationships as a nomological network. In a similar way, the
criteria for evidence-based treatment consist of the entire set of relationships with
274  William E. Piper and John Ogrodniczuk
other relevant concepts and outcomes. In defining the criteria for evidence-based
treatment, however, we would argue that one should similarly avoid a narrow con-
ception and allow criteria to evolve from the main findings concerning treatment
and outcome. In regard to our research, this would mean including the entirety of
information about prevalence, control findings, comparative findings, composi-
tional findings, and processes.
Anyone who has ever carried out a psychotherapy clinical trial can understand
and appreciate the difficulties and challenges that one must confront. One may
experience a sense of satisfaction even when just one significant finding results.
However, one should not stop at this point. For example in our research, clarifying
aspects of prevalence, control, comparison, and compositional findings goes far
beyond the value and usefulness of only the control trial findings.
The achievement of evidence-based treatment in and of itself should not be
regarded as a kind of academic trophy that signals the place to stop. In the United
States there is a National Registry for evidence-based treatments. There are three
levels of programs (or treatments). In ascending order of strength, they are labeled
promising, effective, and model. Although they are helpful in providing criteria
that can be applied to programs, they should not encourage investigators to stop
at the lowest level of the Registry. If they do, there is a clear danger that the term
evidence based will become just another catch-word to overcrowd rather than
clarify our language. In regard to our evidence-based group treatments for CG, the
evidence should be regarded as a beginning, not an end.
Publications concerning the effectiveness of short-term group therapy inevi-
tably address the topics of cost and cost-effectiveness. In a recent review of 36
studies that compared individual and group forms of CBT, Tucker and Oei (2007)
concluded that the evidence was not strong enough to favor one form of treatment
over the other. Another worthwhile study would be to investigate the main effects
of length of therapy and form of therapy (individual vs. group), as well as the
interaction of the two variables. Our research team used this design in a study
that investigated the main effects of length of treatment (6 months vs. 24 months)
and form of treatment (individual vs. group) in Montreal in the early 1980s (Piper
et al., 1984). The patients who received group therapy did better in long-term
therapy than in short-term therapy. Also, patients who received individual therapy
did better in short-term therapy than in long-term therapy. Thus, in the case of
complicated grief and form of therapy, the more interesting findings involve inter-
action effects rather than main effects.

Future activities
In regard to our future activities, although we believe that we have identified some
interesting and clinically useful findings concerning the matching of forms of
therapy and patients’ personality characteristics, as well as the entire composition
of the group, we know very little about the specific mechanisms that follow from
these features to bring about favorable outcome. Consequently, we are currently
 Brief group therapies for CG  275
embarking on an exploration and, we hope, an explanation of the therapy process
of the 18 groups from our composition trial as revealed by audiotapes and tran-
scripts of therapy sessions. We would like to discover what mediated the composi-
tion–outcome relationship. On theoretical grounds we hypothesize that the greater
the percentage of high-QOR patients in a group: (1) the more the content of the
group will reflect constructive, mutually productive, and hostility-free interac-
tions; (2) the greater the focus will be on other patients rather on oneself when
a patient speaks; and (3) the more the group will engage in dynamic work. The
identification of mediating mechanisms may suggest how they can be activated
by means other than restrictive group composition (e.g., by excluding low-QOR
patients). Instead, patient preparation or therapist’s technique could be used. That
would facilitate including greater numbers of psychiatric patients with low-QOR
in short-term therapy groups for CG.

References
Allumbaugh, D., & Hoyt, W. (1999). Effectiveness of grief counseling: A meta-analysis.
Journal of Counseling Psychology, 46, 370–380.
Beck, A. T., & Steer, R. A. (1987). Beck Depression Inventory manual. New York: Harcourt
Brace Jovanovich.
Cronbach, L., J., & Meehl, P., E. (1955). Construct validity in psychological tests.
Psychological Bulletin, 52, 281–302.
Currier, J.  M., Neimeyer, R. A., & Berman, J.  S. (2008). The effectiveness of psycho-
therapeutic interventions for bereaved persons: A comprehensive quantitative review.
Psychological Bulletin, 134, 648–661.
Elkin, I., Shea, T, Watkins, J. T, Imber, S. D., Sotsky, S. M., Collins, J. E., et al. (1989).
National Institute of Mental Health treatment of depression collaborative research pro-
gram. Archives of General Psychiatry, 46, 971–982.
Fava, G. A., Ruini, C., Rafanelli, C., Finos, L., Conti, S., & Grandi, S. (2004). Six-year out-
come of cognitive behaviour therapy for prevention of recurrent depression. American
Journal of Psychiatry, 161, 1872–1876.
Fortner, B.  V. (1999). The effectiveness of grief counselling and theory: A quantitative
review. Unpublished manuscript.
Goodheart, C. D., Kazdin, A. E., & Strenberg, R. J. (Eds.). (2006). Evidence-based psycho-
therapy: Where practice and research meet. Washington, DC: American Psychological
Association.
Horowitz, M.  J., Marmar, C.  R., Weiss, D., DeWitt, K.  N., & Rosenbaum, R. (1984).
Brief psychotherapy of bereavement reactions: The relationship of process to outcome.
Archives of General Psychiatry, 41, 438–448.
Horowitz, M. J., Wilner, N., & Alvarez, W. (1979). Impact of Event Scale: A measure of
subjective stress. Psychosomatic Medicine, 41, 209–218.
Jacobson, N. S., Follette, W. C., & Revenstorf, D. (1984). Psychotherapy outcome research:
Methods for reporting variability and evaluating clinical significance. Behavior
Therapy, 15, 336–352.
Jacobson, N. S., & Revenstorf, D. (1988). Statistics for assessing the clinical significance
of psychotherapy techniques: Issues, problems, and new developments. Behavior
Assessment, 10, 133–145.
276  William E. Piper and John Ogrodniczuk
Kato, P. M., & Mann, T. (1999). A synthesis of psychological interventions for the bereaved.
Clinical Psychology Review, 19, 275–296.
Larson, D. G., & Hoyt, W. T. (2007). What has become of grief therapy? An evaluation of
the empirical foundations of the new pessimism. Professional Psychology: Research
and Practice, 38, 347–355.
de Maat, S., Dekker, J., Schoevers, R., & de Jonghe, F. (2006). Relative efficacy of psy-
chotherapy and pharmacotherapy in the treatment of depression: A meta-analysis.
Psychotherapy Research, 16, 562–572.
Magnavita, J.  J. (Ed.). (2010). Evidence-based treatment for personality dysfunction:
Principles, methods, and processes. Washington, DC: American Psychological
Association.
Neimeyer, R. (2000). Searching for the meaning: Grief therapy and the process of recon-
struction. Death Studies, 24, 541–558.
Piper, W. E., Debbane, E. G., Bienvenu, J. P., & Garant, J. (1984). A comparative study
of four forms of psychotherapy. Journal of Consulting and Clinical Psychology, 52,
268–279.
Piper, W. E., Joyce, A. S., McCallum, M., & Azim, H., F. (1998). Interpretive and sup-
portive forms of psychotherapy and patient personality variables. Journal of Consulting
and Clinical Psychology, 66, 558–567,
Piper, W. E., McCallum, M., & Azim, H. F. A. (1992). Adaptation to loss through short-
term group psychotherapy. New York: Guilford Press.
Piper, W.  E., Ogrodniczuk, J.  S., Azim, H.  F., & Weideman, R. (2001). Prevalence of
loss and complicated grief among psychiatric outpatients. Psychiatric Services, 53,
1069–1074.
Piper, W.  E., Ogrodniczuk, J.  S., Joyce, A.  S., & Weideman, R. (2011). Short-term
group therapies for complicated grief: Two research-based models. Washington, DC:
American Psychological Association.
Piper, W. E., Ogrodniczuk, J. S., & Weideman, R. (2005). Screening for complicated grief:
When less may provide more. Canadian Journal of Psychiatry, 50, 680–683.
Piper, W. E., Ogrodniczuk, J. S., Joyce, A. S., Weideman, R., & Rosie, J. S. (2007). Group
composition and group therapy for complicated grief. Journal of Consulting and
Clinical Psychology, 75, 116–125.
Rockland, L.  H. (1989). Supportive, therapy: A psychodynamic approach. New York:
Basic Books.
Rose, S. D. (1989). Working with adults in groups: Integrating cognitive–behavioral and
small group strategies. San Francisco: Jossey-Bass.
Schut, H., Stroebe, M.  S., van den Bout, J., & Terheggen, M. (2001). The efficacy of
bereavement interventions: Determining who benefits. In Stroebe, M. S., Hansson, R. O.,
Stroebe, W., & Schut, H. (Eds.), Handbook of bereavement research: Consequences,
coping, and care (pp. 705–738). Washington, DC: American Psychological Association.
Trivedi, M. H., Rush, A. J., Wisniewski, S. R., Nierenberg, A. A., Warden, D., Ritz, L.,
et al. (2006). Evaluation of outcomes with citalopram for depression using measure-
ment-based care in STAR*D: Implications for clinical practice. American Journal of
Psychiatry, 163, 28–40.
Tucker, M., & Oei, T.  P.  S. (2007). Is group more cost effective than individual cogni-
tive behaviour therapy? The evidence is not solid yet. Behavioural and Cognitive
Psychotherapy, 35, 77–91.
 Brief group therapies for CG  277
Werman, D.  S. (1984). The practice of supportive psychotherapy. New York: Brunner/
Mazel.
Wilfley, D. E., MacKenzie, K. R., Welch, R. R., Ayres, V. E., & Weissman, M. M. (2000).
Interpersonal psychotherapy for group. New York: Basic Books.
Zisook, S., & Lyons, L. (1989–1990). Bereavement and unresolved grief in psychiatric
outpatients. Omega, 20, 307–322.
20 Complicated grief after violent
death
Identification and intervention
E. K. Rynearson, Henk Schut, and
Margaret Stroebe

Introduction
In periods of peace especially, violent deaths account for only a small percent-
age of total deaths, with, for example, 7% of annual deaths in the United States
falling within the category of murder, suicide, and accidents (National Centers
for Disease Control, 2009). However, there is considerable clinical evidence
to support the premise that violent dying has specific and enduring effects on
bereavement and grief (Rando, 1993; Rynearson, 2001). The violent death of a
loved one is a traumatizing experience. In research on family members of murder
victims, researchers have drawn attention to the likelihood of strongly intrusive
and avoidant thoughts combined with hyperarousal, suggesting the presence of
posttraumatic stress reactions (Parkes, 1993; Rynearson, 1994). Because of the
often unexpected suddenness of violent death, combined with violation, and often
intentionality or culpability associated with the death, those attached to the victim
are not only vulnerable to levels of distress that are characteristic of reactions to
non-violent deaths, but particularly prone to thoughts of remorse, retaliation, and
fears of recurrence related to the act of violent dying. Furthermore, following the
work of Janoff-Bulman, it has become widely accepted that fundamental assump-
tions people hold about themselves, the world, and the relation between these two
may be shattered following traumatic loss (Janoff-Bulman, 1992; Matthews &
Marwit, 2003), although recent evidence suggests that these effects may not be as
strong as has been claimed (Mancini, Prati, & Bonanno, 2011).
In addition, complicating features can include having to deal with legal/crime-
related matters and the media. The clinical effects of violent dying are, then,
substantive and dynamically divergent from those of natural dying and may be
associated with prolonged dysfunction, including complicated grief. These pat-
terns of reactions lead to important questions in the context of this book: Do
those who experience the loss of a loved one through violent death have a higher
likelihood of suffering from complicated forms of grief? If so, what is the nature
of difficulties associated with the grieving process among survivors of violent
death? Who among this subgroup are the ones most vulnerable to complications?
Can intervention help these persons to come to terms with their loss?
Although the clinical effects of violent dying appear indisputable, they are
difficult to quantify and are rarely included in standardized measures of grief or
noted in empirical studies. However, given the compelling clinical indications,
 CG after violent death  279
there is good reason to address the above questions scientifically and to evaluate
the body of relevant research, particularly to give directions for future investiga-
tion. Unfortunately for current purposes, but not surprisingly given the nature and
manifestations associated with violent death, the limited literature on bereave-
ment following this type of death has focused largely on posttraumatic stress
symptoms and disorder rather than complicated grief (e.g., Kaltman & Bonanno,
2003; Mancini, Prati, & Black, 2011; Murphy, 2008). However, violent death
has typically been understood to trigger two concurrent but distinct syndromes:
(1) separation distress as a response to the lost relationship (with feelings of
longing, etc.), and (2) traumatic distress in reaction to the manner of dying (with
re-enactment thoughts, etc.) (Rynearson & Sinnema, 1999). Following this dis-
tinction, the former can be understood as relating to (complicated) grief, the latter
to posttraumatic stress (disorder), suggesting the need for scientific understanding
of both types of reactions.
Given the focus of the whole book, in this chapter we examine the phenomena
and manifestations of complicated grief following violent death. We follow the
definition of complicated grief provided by Stroebe, Hansson, Schut, and Stroebe,
(2008):

a deviation from the (cultural) norm (i.e., that could be expected to pertain –
importantly – according to the extremity of the particular bereavement event)
in either (a) the time course or intensity of specific or general symptoms
of grief and/or (b) the level of impairment in social, occupational, or other
important areas of functioning. (p. 7)

Different forms of complicated grief have been identified in the scientific literature
with various labels frequently being attached to them, the main ones being pro-
longed or chronic, delayed/inhibited, and absent grief (see, for example, Chapter
5 in this volume). It becomes evident that we are talking then of “complicated
grief” in terms of a clinically relevant syndrome.
Consideration is first given to the concept of violent death in the context
of bereavement. Then, in the main part of the chapter, empirical literature on
complicated grief following violent death is critically assessed. The focus is on
well-designed, quantitative studies.1 The review covers bereavement following
different types of violent death, including studies of homicide, suicide, accident,
and natural death. It assesses what we know about complicated grief across these
violent and non-violent types of loss in terms of its prevalence and distinctive fea-
tures, risk factors, models/techniques for assessment, and intervention efficacy.
More general concerns about the state of research knowledge are also addressed.
Finally, we draw general conclusions and set a research agenda for the future.

On the definition of violent death

In the scientific literature, violent deaths have been defined as those resulting from
accidents, suicide, or homicide (e.g., Cleiren, 1991; Mancini, Prati, & Bonanno,
2011), and in many international classification systems these three alternative
280  E. K. Rynearson et al.
causes are listed alongside a fourth category, “natural death,” featuring illnesses.
The category of “fatal accident” includes vehicle crash, drowning, and natural
disaster (Currier, Holland, & Neimeyer, 2006). Grouping deaths due to these
three causes into the violent death category is usual in the trauma literature (e.g.,
Kaltman & Bonanno, 2003; Norris, 1992), although it must be noted that some
deaths in these categories may not be violent in the sense that they may not be
an expression of physical force against other persons (e.g., death from suicide
through overdose of sleeping pills, hospital accidents with anesthetics). Traumatic
loss is defined mainly in objective terms as a sudden and violent mode of death,
characterized by one of the above-mentioned three causes. Violent death is thus
more specific than another frequently used categorization including sudden, unex-
pected, and traumatic death (e.g., Fujisawa et al., 2010). Further to the former
categorization, Rynearson’s “3 V’s” – violent, violation (transgression), and voli-
tion (intentional or freely chosen on part of perpetrator or victim) – may pinpoint
useful defining properties of these types of death. Not all of the 3 V features may
be equally applicable to the different types of violent deaths (e.g., volition does
not seem to fit accidents as much as homicide).
Does such a definition into the above three types of violent death suffice? In
contemporary society one might consider further specification of the violent cat-
egories in terms of military attack, including genocide, and terrorism. However,
these massive types of loss are beyond the scope of this chapter, although from
a certain perspective these would fall under homicide.2 Furthermore, euthanasia
may merit separate consideration in the context of impact on bereavement, for
this too is a non-natural cause of death (though related to natural causes such as
cancer), associated with unique bereavement reactions. Like suicide deaths, those
following euthanasia take place with the agreement of the “victim.” However,
regardless of one’s opinion on euthanasia, such a death is difficult to include under
violent deaths.
A further note of caution is in order when drawing a simple distinction between
violent and non-violent (or “natural”) deaths. The words of Barry, Kasl, and
Prigerson (2002) express the point in relationship to bereavement:

Researchers commonly classify deaths as violent or non-violent according

to how the death occurred. Importantly, deaths perceived to be violent by
bereaved individuals may not be classified as violent according to the manner
in which earlier studies have defined violent deaths. Such may be the case
for a natural death that is accompanied by much pain and physical illness.
A death such as this may be perceived as violent by the surviving family
member. (p. 454)

Categorization may thus be inaccurate not only because of a failure to specify

subcategories of violent death, as mentioned above, but also because deaths may
be perceived differently from the formal cause by bereaved survivors. We need to
keep this in mind but at the same time we need to be aware of the risk that what
are generally regarded as rather clear concepts – such as violent and non-violent
causes of death – run the risk of becoming fuzzy, following such reasoning.
 CG after violent death  281

Violent death and complicated grief: assessment of the

scientific literature
Although there are promising directions in research on this topic, there are also
limitations in availability of sound empirical research on bereavement following
violent death. It will become evident in our review of the literature, which fol-
lows, that there are even fewer studies that stringently examine complicated grief
following violent death. Nevertheless, in our view, it is important to assess what
is known so far, so that future research can build on the available empirical litera-
ture. Furthermore, with this interest in mind too, and in contrast with some other
reviewers and researchers conducting empirical research (e.g., Currier, Holland,
Coleman, & Neimeyer, 2008; Vessier-Batchem & Douglas, 2006) we made the
decision to follow very stringent criteria for complicated grief, in line with the
scope of this volume.

Prevalence of complicated grief

With considerable consistency, studies have shown higher intensities of grief
following violent than non-violent causes (e.g., Currier et al., 2008; Dyregrov,
Nordanger, & Dyregrov, 2003; Mancini, Prati, & Black, 2011), but these do
not inform us about complicated grief (as a clinically relevant condition). As
Hardison, Neimeyer, and Lichstein (2005) commented: “Higher scores [on their
complicated grief scale] represent greater impairment . . . [providing] a continuous
measure of intensity of grief-related symptomatology rather than a classificatory
diagnosis of complicated grief disorder” (pp. 103–4; italics added).
When cut-off points on questionnaires have been used as indicators of compli-
cated grief (i.e., also not diagnostic categorization), those bereaved from violent
causes seem to have excessive rates compared with norms for the bereaved in
general. Although one must be cautious about inferring prevalences of com-
plicated grief from such sources (further application of diagnostic criteria and
professional clinical assessment is needed), some indication may be derived. For
example, Ghaffari-Nejad, Ahmadi-Mousavi, Gandomkar, and Reihani-Kermani
(2006) examined the prevalence of complicated grief (intensity) following the
Bam earthquake in Iran, which killed thousands. Scores over the established cut-
off point for complicated grief were present among 76% of their large sample of
respondents. Dyregrov et al. (2003) reported that 78% of parents following the
violent loss of their child (suicide or accident) scored above the cut-off levels
for complicated grief reactions 1.5 years post loss. The above prevalences are in
excess of rates found for complicated grief irrespective of mode of death (which
is typically within the range of 5–33%; see Forstmeier & Maercker, 2006).
However, the data are not totally conclusive, since some studies do not find higher
prevalence of complicated grief after violent death, although this may be because
population samples include only small numbers of people bereaved by violent
death (e.g., Kersting, Brähler, Glaesmer & Wagner, 2011).
Some researchers have drawn conclusions about the comparative impact of
the different modes of violent death (e.g., Cleiren, 1991; Currier et al., 2008;
282  E. K. Rynearson et al.
Dyregrov et al., 2003). For example, in terms of highest ICG scores, Currier et al.
(2008) reported homicide to be the most perturbing cause of death, followed by
suicide, accidents, natural sudden, and natural anticipated deaths. Dyregrov et al.
(2003) found similar differences in high levels of grief among bereaved parents
following the death of their child as a result of suicide or accident. Some stud-
ies have focused on intensity of symptoms following loss from specific types of
violent death compared with non-violent death, for example for suicide (Bailley,
Kral, & Dunham, 1999), providing insights into differential types of reactions,
but so far not comparing prevalences of complicated grief across types of death.

Distinctive features of complicated grief

Just as there are no scientifically stringent studies comparing the prevalence of
complicated grief reactions following violent as compared with non-violent causes
of death, so is there little sound investigation of distinctive features across these
modes of death. The literature on bereavement following homicide is a case in
point. There is frequent reference to the excruciating, long-lasting, and extremely
complex reactions following this type of loss (e.g., Horne, 2003; Pynoos & Nader,
1990). Causal statements (e.g., about mechanisms contributing to resilience)
are often made on the basis of small-scale qualitative studies of the homicide
bereaved (e.g., Burke, Neimeyer, & McDevitt-Murphy, 2010; Johnson, 2010).
Although these studies provide fine-grained descriptive accounts of what these
persons are encountering, it remains unclear to what extent the identified reactions
are characteristic of bereavement in general, or specific to homicide bereavement
in particular. Clearly, for this purpose, one needs to compare homicide bereaved
with other bereaved groups.
Are there any studies of complicated grief that have included control groups
of bereaved from non-violent death causes, to take us a step further toward dis-
cerning unique bereavement reactions following the different types of loss? To
our knowledge, only one study comes near to reaching the necessary criteria.
McClatchey, Vonk, and Palardy (2009) investigated the prevalence of “childhood
traumatic grief” (CTG) among bereaved children, using a cut-off score indicating
“clinically-significant frequency” (p. 312). Children who had lost a parent through
violent/sudden or through expected death were compared. Rather in contrast to
what one would expect from studies reviewed so far, the incidence of CTG did
not differ between these groups. However, it must be noted that this study did not
completely follow our definition for violent death: This category included sudden
non-violent deaths (heart attacks). This may partly explain why no differences
were found between the groups.

Risk factors
Not surprisingly, given the state of knowledge described above, no information
is available about risk (or protective) factors that may make some individuals
more (or less) vulnerable to complicated grief following a death through murder,
 CG after violent death  283
suicide, or accident. Future research needs to cover a broad range of risk/pro-
tective factors. So far, leads have been provided by researchers examining the
relationship between intensity of grief and sense making or meaning making and
found these factors to be particularly problematic following violent (compared
with non-violent) death circumstances (e.g., Currier et al., 2006, 2008). Likewise,
based on previous research (e.g., Wickie & Marwit, 2000), there are good reasons
to assume that the shattering of world assumptions should be systematically and
differentially related to mode of death. Furthermore, in a recent study by Mancini,
Prati, and Black (2011) self-worth was found to mediate the effects of violent
loss on posttraumatic stress symptoms and depression, but not on levels of grief.
However, extension beyond meaning making and world assumptions to other
intra- and interpersonal risk and protective factors is essential.

Conclusions
Our review of the empirical literature has revealed that there is remarkably little
sound empirical research on complicated grief following violent compared with
non-violent death. Such comparisons are essential to establish the unique conse-
quences of violent types of death. Quite consistently, studies have shown higher
intensities of grief following violent than non-violent causes, but these do not
inform us about complicated grief. Likewise, those bereaved from violent causes
seem to have highly excessive rates compared with norms for the bereaved in
general, but evidence is weak: We could not find a single well-controlled study
that compared complicated grief rates following violent versus non-violent death.
A research design to overcome these gaps in the literature would comprise a (pref-
erably) longitudinal comparison across violent and non-violent modes of death
of the prevalence and manifestations of complicated grief assessed by means of
clinical interviews. The ideal study would be large-scale, use a prospective design,
assess violence incrementally, not use cause of death and violence interchange-
ably, and consider the circumstances of the bereaved. A research agenda for the
future should include examination of risk/protective factors, map different pat-
terns of complications following different causes of violent death, and go beyond
diagnosis based on total symptom score, to consider complicated grief due to
some particular, idiosyncratic feature. Furthermore, we need to test the models
and strategies of psychotherapeutic intervention; examining the effectiveness of
these programs is critically important.
Even with such guidelines, the challenge remains for researchers and clini-
cians to decide who among the bereaved should be included in the complicated
grief category in future investigation of bereavement following violent death.
As stated earlier, our interest is in complicated grief as a clinically relevant syn-
drome. Simply using a continuous measure of intensity of grief symptoms that
indicates increasing impairment is not – in this context – informative (it simply
shows the intensity of grief-related symptomatology). Using a validated cut-off
point to ascertain the likelihood of complicated grief is at least a first step toward
establishing the presence of complicated grief. However, this by itself is not
284  E. K. Rynearson et al.
sufficient when we are striving to investigate complicated grief as a clinically rel-
evant syndrome. Thus, as indicated above, in our view, it is necessary for trained
professionals to conduct clinical interviews to establish “complicated grief.” The
criteria they use for determining this are also not set in stone, but are currently
likely to include use of a cut-off point on a validated grief questionnaire for initial
screening, making use of criteria proposed for the future DSM category system,
and further information from the bereaved person in the clinical interview(s).

Violent death and complicated grief: conceptual issues

We have already considered the definition of complicated grief in some detail.
However, other conceptual issues arise from our review. In particular, when sug-
gesting new directions for research, it is important to step back and consider whether
extension of current directions is sufficient, or whether we should be extending the
scope of our investigation. Although far from comprehensive, the following two
issues illustrate the sorts of extensions that we think deserve attention.

Different causes, different complications?

Finer-grained quantitative examinations of comparative patterns of complication
associated with the different types of violent death are completely lacking. Studies
that have provided qualitative or descriptive accounts of bereavement following
the different modes of death can be drawn on for identification of variables for
inclusion in future studies (see, for example, Armour, 2006, for comparison of
experiences among those bereaved following accidents, suicide, and homicide). It
is noteworthy that there are more qualitative studies focusing on homicide (e.g.,
Asaro, 2001a, 2001b; Burke et al., 2010; Clements & Burgess, 2002; Goodrum,
2005) and suicide (e.g., for reviews, see Jordan, 2008; Jordan & McIntosh, 2010;
Sveen & Walby, 2007) than on accidents. A notable exception is the study of
traffic accident survivors by Lehman, Wortman, and Williams (1987), although,
again, the focus is not on complicated grief but on other consequences such as
depression and general psychiatric symptoms. Thus, more studies specifically
on (complicated) grief reactions following accidental death circumstances are
needed too.
A few studies have looked in general (i.e., not cause of death specifically) at
the comorbidity of complicated grief with posttraumatic stress disorder, depres-
sion, or anxiety disorders (e.g., Morina, Rudari, Bleichhardt, & Prigerson, 2010)
but, to our knowledge, none have compared patterns of comorbidity following
violent versus non-violent causes.

Accumulation of symptoms versus idiosyncratic complications?

Complicated grief may be overlooked if the focus is limited to high accumulated
levels of symptoms or even diagnostic assessment, rather than identifying essential
and unique bereavement reactions relating to a particular mode of death (violent
 CG after violent death  285
and non-violent). In other words, assessment of complicated grief following a
violent death which is based on an initial score above a prescribed cut-off point
on a questionnaire and/or confirmed by diagnostic investigation may still exclude
persons who do suffer complications. Following violent death, complications in
the grieving process may have more to do with a specific aspect than elevation of
symptoms, or accumulation of symptoms, to reach diagnosis. For example, fol-
lowing suicide, a specific difficulty may have to do with the overwhelming feeling
that one should have prevented the death, or with the extreme strain of feeling one
has to keep the actual cause of death a secret. These types of thoughts are reported
more frequently following suicide than following other types of death (Bailley et
al., 1999). In this context, it is important to note that assessment of the level of
grief may vary according to whether general grief instruments or mode-of-death-
specific instruments are used (see Sveen & Walby, 2007, for suicide bereavement).

Psychotherapeutic intervention following violent death

The review of empirical evidence presented above identifies limited, yet valuable,
knowledge about complicated grief following violent death and areas for research
in the future. However, as will have become evident from the preceding review,
although such lines of investigation are potentially informative, they are unlikely
to provide a comprehensive picture, and certainly not one that is sufficient for
clinical purposes. They can usefully be complemented – and enriched – through
examination of principles of psychotherapeutic intervention and assessment. In
our view, there are good reasons to bring research and practice together to address
these topics, to provide further insight into complicated grief following violent
death, which will also contribute to the forthcoming research agenda.

Effects of psychotherapeutic interventions

Given that few studies have specifically isolated complicated grief reactions fol-
lowing violent deaths, it is not surprising that there is little information available
about the efficacy of intervention programs specifically on complicated grief fol-
lowing violent loss. Very few studies have even focused on other consequences of
intervention (e.g., on lowering posttraumatic stress symptom levels) for bereaved
persons after violent death. Most of the studies that have done so also include
people who have encountered other impactful events without lethal consequences
(e.g., Brom, Kleber, & Defares, 1989; Layne et al., 2008). These studies do find
positive results of the interventions put to the test, but results are not presented
for bereaved and other victims of these events separately, and no attention is paid
specifically to recovery from complicated grief.
A study conducted by Murphy (Murphy et al., 1998; for a recent overview, see
Murphy, 2008), however, specifically focused on parents who have lost a child
through suicide, homicide, or accident. In this study, the parents were randomly
assigned to a mixed problem- and emotion-focused intervention condition or a
non-intervention control group. Results show that mothers with high initial levels
286  E. K. Rynearson et al.
of distress improved more in the intervention condition than did similar mothers
in the control condition. By contrast, mothers with relatively low levels of distress
at baseline were worse after intervention than the control group. The number of
fathers participating in the study was small, but results did not indicate an effect
of the intervention for fathers. The impact specifically on complicated grief, how-
ever, was not investigated in this study (the grief measure was an unvalidated
scale developed by the investigators for this particular study).
A recent review by Szumalis and Kutcher (2011; for a further review, see
McDaid, Trowman, Golder, Hawton, & Sowden, 2008) summarized results for the
effectiveness specifically for postsuicide intervention programs, including their
impact on grief symptoms (though again not specifically on complicated grief).
Improvements in grief experiences, both short and long term, were reported.
These results are promising in that they suggest that intervention programs
may be effective in reducing distress, but it is evident that more research into the
effects of intervention after (specific types of) violent bereavement is needed too,
and for subgroups undergoing complications in their grieving.
Turning to specific techniques of assessment: At this time, to the best of our
knowledge, there are five manualized, time-limited, focused interventions specifi-
cally designed for non-natural dying. Three of these (Cohen et al., 2006; Layne
et al., 2008; Salloum, 2008) have been designed for children and adolescent
outpatients and two (Rynearson et al., 2006; Murphy, 2008) for adults. Although
evidence so far seems to suggest decrease of grief symptoms and trauma distress,
further methodologically sound investigations are needed to establish efficacy
(only the Murphy study met the necessary criteria for current purposes). Four of
the interventions applied combined techniques from CBT and narrative therapy.
CBT principles included structured, time-limited agendas (10–12 individual or
group sessions), relationship-based collaboration, clarification of connections
between thoughts, feelings, and behaviors, affirmative guidance, relaxation
exercises, modeling, and teaching techniques of imaginative exposure. We illus-
trate this by describing next the two adult approaches developed by Murphy and
Rynearson.
The support groups designed by Murphy and colleagues (e.g., Murphy, 1996;
Murphy, Baugher, Lohan, Schneidermann, & Herrwagen, 1996; Murphy et al.,
1998) included problem-focused and emotion-focused support. This program was
professionally led, and designed specifically for parents who had lost a child to
violent death. Twelve sessions of 2 hours were held, one each week. The first hour
in each session (apart from the first and last, which were data-collecting sessions)
was dedicated to the problem-focused support dimension of providing informa-
tion and building coping skills, addressing areas such as managing cognitive and
emotional responses; health issues; parental role loss; legal concerns; partner and
family relationship concerns; feelings toward others; and expectations for the
future. The second hour, of more direct emotion-focused support, assisted parents
to share their experiences; obtain feedback to help reframing of aspects to do with
the death and its consequences; and receive emotional support.
 CG after violent death  287
Narrative therapy (Rynearson, 2010; see also Currier & Neimeyer, 2006,
and Currier et al., 2008, for their related meaning making approach) contrasts
in some respects with the problem- and emotion-oriented support program of
Murphy and colleagues. Strategies based on the narrative approach encourage
the retelling of the living and dying story of the deceased with a restorative goal
of creating a more plausible and coherent retelling of the narrative imagery, of
re-enactment, promoting alternative outcomes and a transcendent perspective.
Rynearson (2010) recently described his narrative therapy approach in some
detail. According to this perspective, in the case of violent death, storytelling is
distorted, focusing intensely, even obsessively, on a re-enactment of the dying
(further complicated by the public and legal processes surrounding this type
of death). Fundamental to this approach is the understanding that established
clinical assessment principles (based on narrative analysis) and associated con-
structs for guiding therapeutic interventions give insight into complicated grief
following violent death. It is understood that, in order to assess the impact of
the death on the bereaved person, the clinician needs to listen to and help the
client to revise his or her personal account of the loss and its aftermath. This
is likely to be necessary for bereavement in general, but it may be even more
important after violent death. Characteristic of narratives of bereavement after
violent death are themes of horror and helplessness and topics such as remorse,
re-enactment, retaliation, and retribution that distort the dying story structure
(Currier & Neimeyer, 2006).
The procedures of Murphy and Rynearson may overlap, insofar as participants
presumably shared their narratives of the living and dying of the deceased spon-
taneously during the support group of Murphy and colleagues. There are also
apparent differences between the Murphy and Rynearson (and other) approaches
(the Murphy program has no sessions of direct exposure and retelling of the trau-
matic dying re-enactment, whereas the Rynearson one does). However, none of
the manualized interventions cites a specific corrective mechanism, and that is
presumably because the explanation of treatment effects is non-specific. It seems
plausible that the various interventions are successful because they are based
upon the common principles of stress moderation, reconstructive exposure, and
meaningful re-engagement, which are basic to time-limited trauma or grief treat-
ments. Three main goals of such interventions are (see also Rynearson, Correa,
Favell, Saindon, & Prigerson, 2006):

1 The moderation of distress (through a confiding relationship, a safe setting,

psychoeducation and stress reduction strategies) that fosters mastery of per-
sonal safety and autonomy.
2 Exposure and reconstructive processing of the dying and grieving narra-
tive through an active procedure of reliving the narrative fixation (through
imaginary verbal and non-verbal retelling). This would foster coherence and
motivation for re-engagement by revising the teller’s role (identity) within
the narrative.
288  E. K. Rynearson et al.
3 Meaningful re-engagement with valued, vital activities and relationships
within the family and community in an altered identity that honors the
transformation.

Apart from the Murphy investigation (which included examination of grief

reduction but not of clinically assessed complicated grief), we know of no studies
that have put such approaches to strict scientific test. Nor do we know of any that
have compared these assessment techniques in relationship to the effectiveness
of an intervention based on their protocol. Specific investigation of their impact
on complicated grief for bereaved persons following violent loss is also lacking.
These are all matters for future investigation. A useful strategy would be to follow
the example of Shear, Frank, Houck, and Reynolds (2005), who conducted a study
of the efficacy of intervention for bereaved persons with complications in their
grieving, comparing the efficacy of two different scientifically based programs.
They assigned their clients either to traditional interpersonal psychotherapy or to
a treatment program that followed the principles of a specially derived protocol,
based on the dual-process model (see Stroebe & Schut, 1999), called compli-
cated grief treatment, and examined the course of their grief over time (the latter
treatment program was associated with faster and better adjustment). Naturally,
inclusion of a non-intervention control group to compare the treatment conditions
with natural recovery trajectories would be advisable; this was not feasible in the
Shear et al. (2005) study. For our current interests, it would be useful to extend
such intervention efficacy examination of its impact on those bereaved following
the violent death of a loved one, specifically. It is useful to note in this context
that Asukai, Tsuruta, and Saito (2011) recently conducted a pilot study using a
modified version of Shear and colleagues’ complicated grief treatment for a small
sample of Japanese women bereaved by violent death. The results were promis-
ing, with reporting of a significant reduction of grief symptoms but, as for the
Shear et al. study, a non-intervention control condition was not included.

General conclusions
To convince governments and funding agencies of the importance of supporting
those dealing with the violent death of a relative, one needs, first, to demonstrate
that these survivors encounter greater and/or different extreme difficulties than
do other bereaved persons and, second, to show how professional intervention
can actually help reduce suffering associated with this type of death. However,
our review of scientific evidence on complicated grief following violent death
revealed remarkably little sound knowledge to date in terms of recovery from
complicated grief through intervention. However, research is moving toward
addressing issues surrounding the prevalence of complicated grief and comparing
violent and non-violent causes. Although more fine-grained research is needed,
results do suggest violent death to be a risk factor for complicated grief. We high-
lighted new research directions, ranging from prevalence (e.g., good comparative
studies of impact, focusing on symptomatology and complicated grief “caseness”)
 CG after violent death  289
to intervention efficacy studies (to elucidate what works best for whom, following
specific types of violent death). As illustrated above, promising research along
these lines is already being conducted (e.g., Fujisawa et al., 2010; Kersting et al.,
2011). We outlined how different approaches, including the narrative approach
to clinical assessment and intervention, can fuel future research, and how such
approaches provide guidelines for the treatment of complicated grief experienced
by some bereaved persons following the violent death of a loved one. We hope
that researchers and practitioners can work together toward building a solid
knowledge base, thereby improving the evidence base of care for these bereaved
persons.

Notes
1 It is beyond the scope of this chapter to cover qualitative investigations of compli-
cated grief following violent death. In fact, most qualitative studies also highlight
posttraumatic stress rather than complicated grief reactions.
2 For an example of empirical research in this category, see Schaal, Jacob,
Dusingizemungu, and Elbert (2010). For a review of the consequences of disasters
on individuals, families and communities, see Bonanno, Brewin, Kaniasty, and La
Greca (2010).

References
Armour, M. (2006). Violent death: Understanding the context of traumatic and stigmatized
grief. Journal of Human Behavior in the Social Environment, 14, 53–90.
Asaro, M. (2001a). Working with adult homicide survivors, Part I: Impact and sequelae of
murder. Perspectives in Psychiatric Care, 37, 95–101.
Asaro, M. (2001b). Working with adult homicide survivors, Part II: Helping family mem-
bers cope with murder. Perspectives in Psychiatric Care, 37, 115–136.
Asukai, N., Tsuruta, N., & Saito, A. (2011). Pilot study on traumatic grief treatment pro-
gram for Japanese women bereaved by violent death. Journal of Traumatic Stress, 24,
470–473.
Bailley, S., Kral, M., & Dunham, K. (1999). Survivors of suicide do grieve differently:
Empirical support for a common-sense proposition. Suicide and Life-Threatening
Behavior, 29, 256–271.
Barry, L., Kasl, S., & Prigerson, H. (2002). Psychiatric disorders among bereaved persons:
The role of perceived circumstances of death and preparedness for death. American
Journal of Geriatric Psychiatry, 10, 447–457.
Bonanno, G., Brewin, C., Kaniasty, K., & La Greca, A. (2010). Weighing the costs of
disaster: Consequences, risks, and resilience in individuals, families, and communities.
Psychological Science, 11, 1–49.
Brom, D., Kleber, R., & Defares, P. (1989). Brief psychotherapy for posttraumatic stress
disorders. Journal of Consulting and Clinical Psychology, 57, 607–612.
Burke, L., Neimeyer, R., & McDevitt-Murphy, M. (2010). African American homicide
bereavement: aspects of social support that predict complicated grief, PTSD, and
depression. Omega, 61, 1–24.
Cleiren, M. (1991). Adaptation to bereavement. Leiden: DSWO Press.
Clements, P., & Burgess, A. (2002). Children’s responses to family member homicide.
Family Community Health, 25, 32–42.
290  E. K. Rynearson et al.
Cohen, J. A., Mannarino, A. P., & Staron, V. (2006). A pilot study for modified cognitive–
behavioral therapy for childhood traumatic grief. Journal of the Academy of Child and
Adolescent Psychiatry, 45, 1465–1473.
Currier, J., Holland, J., & Neimeyer, R. (2006). Sense-making, grief, and the experience of
violent loss: Toward a mediational model. Death Studies, 30, 403–428.
Currier, J., Holland, J., Coleman, R., & Neimeyer, R. (2008). Bereavement following
violent death: An assault on life and meaning. In Stevenson, R., & Cox, G. (Eds.)
Perspectives on violence and violent death (pp. 175–200). Amityville, NY: Baywood.
Currier, J. M., & Neimeyer, R. A. (2006). Fragmented stories: The narrative integration
of violent loss. In Rynearson, E. K. (Ed.), Violent death: Resilience and intervention
beyond the crisis (pp. 85–100). New York: Taylor & Francis.
Dyregrov, K., Nordanger, D., & Dyregrov, A. (2003). Predictors of psychosocial distress
after suicide, SIDS, and accidents. Death Studies, 27, 143–165.
Forstmeier, S., & Maercker, A. (2006). Comparison of two diagnostic system for compli-
cated grief. Journal of Affective Disorders, 99, 203–211.
Fujisawa, D., Miyashita, M., Nakajima, S., Ito, M., Kato, M., & Kim, Y. (2010). Prevalence
and determinants of complicated grief in general population. Journal of Affective
Disorders, 127, 352–358.
Ghaffari-Nejad, A., Ahmadi-Mousavi, M., Gandomkar, M., & Reihani-Kermani, H.
(2006). The prevalence of complicated grief among Bam earthquake survivors in Iran.
Archives of Iranian Medicine, 10, 525–528.
Goodrum, S. (2005). The interaction between thoughts and emotions following the news of
a loved one’s murder. Omega, 51, 143–160.
Hardison, H., Neimeyer, R., & Lichstein, K. (2005). Insomnia and complicated grief symp-
toms in bereaved college students. Behavioral Sleep Medicine, 3, 99–111.
Horne, C. (2003). Families of homicide victims: Service utilization patterns of extra- and
intrafamilial homicide survivors. Journal of Family Violence, 18, 75–82.
Janoff-Bulman, R. (1992). Shattered assumptions: Towards a new psychology of trauma.
New York: Free Press.
Johnson, C. (2010). When African American teen girls’ friends are murdered: A qualitative
study of bereavement, coping, and psychological consequences. Families in Society,
91, 364–370.
Jordan, J. (2008). Bereavement after suicide. Psychiatric Annals, 38, 1–6.
Jordan, J., & McIntosh, J. (2010). Grief after suicide: Understanding the consequences
and caring for the survivors. New York: Routledge.
Kaltman, S., & Bonanno, G. (2003). Trauma and bereavement: Examining the impact of
sudden and violent deaths. Anxiety Disorders, 17, 131–147.
Kersting, A., Brähler, E., Glaesmer, H., & Wagner, B. (2011). Prevalence of complicated
grief in a representative population-based sample. Journal of Affective Disorders, 131,
339–343.
Layne, C.  M., Saltzman, W.  R., Poppleton, L., Burlingame, Pasalic, A., Durakovic, E.,
et al. (2008). Effectiveness of a school-based psychotherapy for war-exposed adoles-
cents: A randomized controlled trial. Journal of the Academy of Child and Adolescent
Psychiatry, 47, 1048–1062.
Lehman, D., Wortman, C., & Williams, A. (1987). Long-term effects of losing a spouse
or child in a motor vehicle crash. Journal of Personality and Social Psychology, 52,
218–231.
Mancini, A. D., Prati, G., & Black, S. (2011). Self-worth mediates the effects of violent
loss on PTSD symptoms. Journal of Traumatic Stress, 24, 116–120.
 CG after violent death  291
Mancini, A. D., Prati, G., & Bonanno, G. A. (2011). Do shattered worldviews lead to com-
plicated grief? Prospective and longitudinal analyses. Journal of Social and Clinical
Psychology, 30, 184–215.
Matthews, L. T., & Marwit, S. J. (2003). Examining the assumptive world views of parents
bereaved by accident, murder, and illness. Omega, 48, 115–136.
McClatchey, I., Vonk, M., & Palardy, G. (2009). The prevalence of childhood traumatic
grief: A comparison of violent/sudden and expected loss. Omega, 59, 305–323.
McDaid, C., Trowman, R., Golder, S., Hawton, K., & Sowden, A. (2008). Interventions
for people bereaved though suicide: Systematic review. British Journal of Psychiatry,
193, 438–443.
Morina, N., Rudari, V., Bleichhardt, G., & Prigerson, H. (2010). Prolonged grief disorder,
depression, and PTSD among bereaved Kosovar civilian war survivors: A preliminary
investigation. International Journal of Social Psychiatry, 56, 288–297.
Murphy, S. (1996). Parent bereavement stress and preventive intervention following the
violent deaths of adolescent or young adult children. Death Studies, 2, 441–452.
Murphy, S. (2008). The loss of a child: Sudden death and extended illness perspectives. In
Stroebe, M., Hansson, R. O., Schut, H., & Stroebe, W. (Eds.). Handbook of bereave-
ment research and practice: Advances in theory and intervention (pp.  375–395).
Washington, DC: APA.
Murphy, S., Baugher, R., Lohan, J., Schneidermann, J., & Herrwagen, J. (1996). Parents’
evaluation of a preventive intervention following the sudden violent deaths of their
children. Death Studies, 20, 435–468.
Murphy, S., Johnson, C., Cain, K., Das Gupta, A., Dimond, M., & Lohan, J. (1998).
Broad-spectrum group treatment or parents bereaved b the violent deaths of their 12- to
28-year-old children: A randomized, controlled trial. Death Studies, 22, 209–235.
National Centers for Disease Control. (2009, April 17). Vital Statistics Report, 57, 14.
Norris, F. (1992). Epidemiology of trauma: Frequency and impact of different potentially
traumatic events on different demographic groups. Journal of Consulting and Clinical
Psychology, 60, 409-418.
Parkes, C.  M. (1993). Psychiatric problems following bereavement by murder or man-
slaughter. British Journal of Psychiatry, 162, 49–54.
Pynoos, R. S., & Nader, K. (1990). Children’s exposure to violence and traumatic death.
Psychiatric Annals, 20, 334–344.
Rando, T. (1993). Treatment of complicated mourning. Champaign, IL: Research Press.
Rynearson, E.  K. (1994). Psychotherapy of bereavement after homicide. Journal of
Psychotherapy Practice and Research, 3, 341–347.
Rynearson, E. K. (2001). Retelling violent death. New York: Brunner-Routledge.
Rynearson, E.  K. (2010). The clergy, the clinician, and the narrative of violent death.
Pastoral Psychology, 59, 179–189.
Rynearson, E. K., Correa, F., Favell, J., Saindon, C., & Prigerson, H. (2006). Restorative
retelling after violent dying. In Rynearson, E. K. (Ed.), Violent dying: Resilience and
intervention beyond the crisis (pp. 195–216), New York: Taylor & Francis.
Rynearson, E.  K., & Sinnema, C.  S. (1999). Supportive group therapy for bereavement
after homicide. In Blake, D., & Young, B. H. (Eds.), Group treatment for post traumatic
stress disorder (pp. 137–147). New York: Taylor & Francis.
Salloum, A. (2008). Group therapy for children experiencing grief and trauma due to homi-
cide and violence: A pilot study. Research and Social Work Practice, 18, 198–211.
Schaal, S., Jacob, N., Dusingizemungu, J.-P., & Elbert, T. (2010). Rates and risks for pro-
longed grief disorder in a sample of orphaned and widowed genocide survivors. BMC
Psychiatry, 10. doi: 10.1186/1471-244X-10-55.
292  E. K. Rynearson et al.
Shear, K., Frank, E., Houck, P., & Reynolds, C. (2005). Treatment of complicated grief:
A randomized controlled trial. Journal of the American Medical Association, 293,
2601–2608.
Stroebe, M. S., Hansson, R. O., Schut, H., & Stroebe, W. (2008). Handbook of bereavement
research and practice: Advances in theory and intervention. Washington, DC: APA.
Stroebe, M., & Schut, H. (1999). The dual process model of coping with bereavement:
Rationale and description. Death Studies, 23, 197–224.
Sveen, C. A., & Walby, F. (2007). Suicide survivors’ mental health and grief reactions: A
systematic review of controlled studies. Suicide and Life-Threatening Behavior, 38,
13–30.
Szumalis, M., & Kutcher, S. (2011). Post-suicide intervention programs: A systematic
review. Canadian Journal of Public Health, 102, 18–29.
Vessier-Batchem, M., & Douglas, D. (2006). Coping and complicated grief in survivors of
homicide and suicide decedents. Journal of Forensic Nursing, 2, 25–32.
Wickie, S., & Marwit, S. (2000). Assumptive world views and the grief reactions of parents
of murdered children. Omega, 42, 101–113.
Part VI

Conclusions
21 Complicated grief
Assessment of scientific knowledge
and implications for research and
practice
Margaret Stroebe, Henk Schut, and
Jan van den Bout

Our objective in compiling this edited volume has been to provide an up-to-date,
state-of-the-art account of scientific research on complicated grief (CG), one that is
hoped to be useful for researchers, practitioners, and policy makers alike. We have
included diverse contributions, representing contemporary research and thinking
from a variety of disciplines and perspectives. Scientific and societal issues have
been addressed throughout, and it will have become evident that our authors have
at times come to different conclusions on fundamental issues. In this concluding
chapter, we reflect on the research presented in this volume, to summarize devel-
opments, highlight implications, and indicate current understanding – as well as
gaps in our knowledge  – regarding CG. We try to draw together the different
lines of argument, so that readers can form their own conclusions about scientific
knowledge on CG and associated implications for research, practice, and policy.
We order discussion mainly according to the themes of each part of the volume,
given that these reflect our chosen scope. We cover a range of general issues sug-
gested by the contributions, and we highlight topics which merit further scientific
attention.

The nature of complicated grief

A fundamental question that we wanted to explore in compiling this volume
was: What is CG? Is it a specific disorder or an overarching term for several
different disorders? Can we describe its defining features/symptoms? Although
there appears to be quite some similarity between authors’ conceptualizations,
what has emerged is that there is still a lack of a well-accepted, standard defini-
tion for CG or agreement on a set of diagnostic criteria (if it is to be considered
a mental disorder). There does seem to be considerable  – at least implicit  –
agreement that CG denotes a syndrome, that is, a pathological condition, one
that merits the attention of health care professionals. In this section, we focus
on the concept of CG per se; in the next, we discuss the related issue of CG as
a mental disorder.
296  Margaret Stroebe et al.

CG definition
Some authors have basically followed the lead of Prigerson and her colleagues,
conceptualizing CG as one specific disorder, prolonged grief disorder (PGD),
using the scale of assessment and criteria derived from her extensive body of
research, both of which have evolved over time. Others have criticized this
approach as being too narrow, potentially leaving out subtypes of CG (we return
to this below). Boelen, van den Hout, and van den Bout, following PGD, define
CG as:

a clinical condition that encompasses specific grief reactions (including

separation distress, difficulties accepting the loss, and moving on without
the lost person) that cause significant distress and disability at least 6 months
after the death occurred. Symptoms of PGD are distinct from normal grief,
bereavement-related depression, posttraumatic stress disorder (PTSD), and
other anxiety symptoms and syndromes, and, if left untreated, associated
with significant impairments in health and quality of life. (Chapter 16)

Others adopt definitions along the lines of the definition set out in Chapter 1:

a clinically-significant deviation from the (cultural) norm (i.e., that could be

expected to pertain, according to the extremity of the particular bereavement
event) in either (a) the time course or intensity of specific or general symp-
toms of grief and/or (b) the level of impairment in social, occupational, or
other important areas of functioning. (Stroebe, Hansson, Schut, & Stroebe,
2008, p. 7)

Boerner and colleagues add further specifications to this latter definition, incor-
porating avoidance processes, self-redefinition problems, and difficulties forming
new relationships. Rando has perhaps the broadest definition, covering four types
of CG presentations: symptoms, syndromes, diagnosable mental or physical dis-
orders, and death.
Hopefully, as researchers continue to work toward further conceptual clarity,
a robust definition (and operationalization) and possibly typologies/subtypes of
CG will emerge.

Normal versus CG
To increase understanding of CG, it has sometimes been considered in relation
to normal grief (NG; some prefer to label this uncomplicated grief). For exam-
ple, Dyregrov and Dyregrov discuss patterns of NG and CG in children (noting
that this distinction is actually much more difficult to make for children). Three
defined features that were mentioned above to characterize CG are deviations/dif-
ferences (from NG) (1) in intensity, (2) in quality, and (3) in duration. These have
been subject to critical scrutiny in this volume; we consider each of the three next.
 Implications for research and practice  297
Questions about intensity relate to whether CG is simply a higher (or even
lower) level of symptoms or a distinct syndrome with different symptoms.
Wakefield notes that some bereaved people have to deal with greater and deeper
challenges than others, and that this does not make an intense response auto-
matically into CG. Along similar lines, Rynearson, Schut, and Stroebe argue that,
even if higher scores on a CG scale represent greater impairment (indicated on
a continuous measure of intensity of grief symptoms), this does not constitute a
classificatory diagnosis of CG. The question then is: What is classificatory CG?
Again, different aspects have been considered, relating to intensity. O’Connor,
from a physiological perspective, writes that it is still unknown whether CG rep-
resents merely acute grief in a bereaved person whose process of adaptation has
been interrupted, or a wholly other process than non-complicated adaptation. She
mentions in conclusion that most affective disorders are better described on a
continuum than as discrete categories and that CG may well turn out to be similar
in this regard. The physiological perspective may have potential to answer this
question, but O’Connor cautions that we should not expect a one-to-one corre-
spondence between any particular physiological or neurobiological marker and
CG. However, “by measuring these markers, we may see what contributes to poor
adaptation or what the physiological predictors of CG are” (Chapter 15).
If CG phenomena/symptoms are qualitatively different from NG this would
lend plausibility to the claim that CG symptoms are “pathognomonic for a disor-
der with a distinct etiology” (Chapter 8). Wakefield points to statements of other
researchers that support this view but argues that most have failed to find distinct
symptomatology. So Wakefield, although not denying that some grief disorders
might exist, queries whether CG has distinctive, pathognomonic symptoms sepa-
rating it from normal grief; there is according to him little support for a categorical
conceptualization of NG and PGD. According to Burke and Neimeyer, contem-
porary research suggests that grief can be evaluated on a continuum ranging from
low-level normative grief to a severe grief disorder, but these authors cautioned
that lack of a genuine cut-off point, at which grief responses are considered in
need of treatment, necessitates the use of personal or consensual judgments about
a given griever’s level of impairment and distress. Equally, viewing NG and CG as
symptomatically different is associated with difficulties too. For example, Burke
and Neimeyer argue that, if some individuals grieve in a diagnosably disordered
manner (i.e., with different symptomatology from NG), there is potential for them
to be socially and personally stigmatized.
Is CG distinguished from NG by incorporating a longer duration of basically
normal symptoms? Wakefield argues that interminability (at 6 or 12 months) is
not a valid criterion for CG, describing it as a scientific myth. He cites Bowlby,
making the point that NG is of much longer duration than is generally acknowl-
edged and arguing that to forget this warning pathologizes normal grief. By con-
trast, Boerner and colleagues identify two trajectories indicative of CG, namely
their so-called chronic post-loss distress and chronic pre- and post-loss distress
trajectories, which they contrasted with uncomplicated patterns described as resil-
ient and improved trajectories.
298  Margaret Stroebe et al.
It becomes clear from our chapters that understanding of the role of intensity,
quality, and duration of symptoms in CG can (and needs to) be deepened in future
research and that different disciplines, from philosophers to neurophysiologists to
(cultural) psychologists, have a part to play in this endeavor.

Subtypes/variety of CG patterns
A lot of discussion has surrounded the issue of whether there are subtypes of
CG, and what form(s) these might take. We noted that the PGD/CG approach
of Prigerson and colleagues has sometimes been considered too narrow (e.g., by
some clinicians, as noted by Rando). In the context of considering subtypes of
CG, Boelen and Prigerson’s statement deserves consideration: “criteria . . . should
cover the many different forms the clinical picture of PGD/CG may take” (Chapter
7). Thus it becomes evident that this team of researchers acknowledges variety,
if not explicitly subtypes. CG can comprise different combinations of symptoms,
leaving room for the identification of subtypes (although it is more difficult to see
how absent grief could be included within the PGD framework).
Furthermore, many researchers have identified subtypes of NG and CG. For
example, attachment theory describes insecurities of attachment that mirror onto
subtypes of CG. Interestingly, these subtypes were described long before attach-
ment theory came to have such an influence on the bereavement research field.
Mikulincer and Shaver describe how attachment insecurities are involved in
complicated patterns of grief. Dyregrov and Dyregrov mention subtypes of CG
among children, based on Bowlby’s classification, but note the lack of recognition
of trauma, a subtype that they add. Likewise for adults, Raphael, Jacobs, and Looi
identify traumatic grief as the coexistence of grief and trauma phenomena, and
the different reactive phenomena that may follow these different stressor experi-
ences. Bonanno and colleagues built up a research program to investigate types
of grieving in a fine-grained manner, and distinguish chronic grief from chronic
depression. In addition to describing different CG trajectories, Boerner and col-
leagues take understanding beyond the one pattern of NG (previously understood
as moving from a period of distress to recovery). Exceptions to that pattern used
to be considered CG, but these investigators showed how other patterns can also
be uncomplicated, notably a pattern that resembled resilience.
One of the most debated issues in subtypes concerns absent/delayed/inhibited
grief. Most consider it to be a CG phenomenon, but others have emphasized that
such absence need not always represent CG. Some research teams have reported
little empirical confirmation of this subtype of CG, particularly the team of
Boerner and colleagues – who actually nevertheless acknowledge the existence of
a subtype of delayed/absent grief (they also stress that exhibiting hardly any grief
reactions can be a form of NG). Mikulincer and Shaver make a strong theoretically
based case for two subtypes of CG, including absent grief, based on Bowlby’s
attachment theory analysis: Whereas attachment anxiety is said to underlie chronic
mourning, attachment-related avoidance contributes to the absence of grief. There
is acknowledgement that apparent absence of grief may reflect a real absence of
 Implications for research and practice  299
distress. This used to be difficult to distinguish from CG but Mikulincer and Shaver
report their sophisticated experiments, which have supported this distinction well.
Furthermore, they provide empirical evidence linking attachment insecurities to
CG. Their conclusion was unequivocal: “Overall, these findings emphasize that
researchers should take seriously Bowlby’s (1980) cautions about assessing grief
responses among avoidant people, because ‘in all studies except those using the
most sophisticated of methods, it is easy to overlook such people and to group
them with those whose mourning is progressing in a genuinely favorable way’ ”
(Chapter 14, quoting Bowlby, 1980, p. 211). Dyregrov and Dyregrov discuss the
phenomenon of postponed grief in children, an aspect that may be rather special
among children, being related to the fact that they lack emotional tolerance. These
authors suggest that children attempt to regulate their grief in tolerable doses and
use more avoidance than adults.
In sum, there is a need for scientists to come to agreement on the existence of
various subtypes of CG and to develop methods of assessment that could better
identify and map these (particularly, absent grief).

Subgroup differences in CG
Clearly, one size does not fit all. Throughout the book it has become evident that
there are variations in manifestations and phenomenology, and in appropriate
assessment for and treatment of CG across different groups of bereaved people.
Perhaps most strikingly, this is true for different cultures, as Rosenblatt’s chapter
has made amply clear, as in the opening sentence of Chapter 3: “Psychiatric diag-
nostic categories and psychiatric standards for what is normal and healthy and
what is not are saturated with the standards of Western culture.” Rosenblatt has
raised questions about the possibility of a universal definition of CG. Although
separation distress is recognized by many as a universal emotional response, fol-
lowing Rosenblatt’s line of reasoning, since there are no universal manifestations
of grief, there cannot be a universal definition of CG. So should CG usage/explo-
ration be restricted to Euro-American cultures? This would seem a deplorable
state of affairs, not least because of the pluralistic nature of society in the twenty-
first century. Scientific investigation should strive for culturally appropriate
understanding of CG; extension is needed to incorporate a worldwide perspective.
This point extends to treatment issues. For example, as Kissane points out, the
application of family therapy will require different approaches in countries where
family traditions are strong and decision making family centered. Finally and
importantly, Rosenblatt’s cultural perspective makes one aware that we cannot
consider scientific research on CG in a vacuum. For example, we need to be aware
that psychological treatment sometimes takes place in the context of economic,
political, or environmental turmoils.
The chapters selected to cover two within-culture subgroups, bereavement
of children and adolescents and that following violent death, serve to illustrate
the uniqueness of CG in different subgroups. For example, children form more
misinterpretations than adults, as they lack life experience and direct access to
300  Margaret Stroebe et al.
information about what happened. It seems highly plausible, especially in the
light of the research on adults by Boelen and colleagues, that misinterpretations
would be closely linked with high risk of CG. This needs further investigation in
children. It remains of concern too that children are not mentioned in the propos-
als for a new grief disorder.
Evidently, there is considerable room for expansion of research with regard to
understanding CG within specific subgroups of bereaved people.

Prevalence of CG
Given the difficulties in defining CG, distinguishing it from NG, demarking dif-
ferent subtypes and establishing patterns of similarities across subgroups, it is
understandably difficult to talk about prevalence of CG in any simple terms. Yet,
with some consistency, it has been reported that CG occurs in only a significant
minority of individuals. Reported prevalences vary considerably (because of dif-
ferences in types of loss, sample characteristics, criteria for assessment, etc.). Not
surprisingly perhaps, the percentage of psychiatric outpatients who meet criteria
for CG is higher than for the bereaved in general (Chapter 19). Indications are
that those bereaved following a violent death have still higher prevalences of CG
(Chapter 20), but more studies are needed (much more research following this
type of death has been on PTSD). Unresolved is the issue whether these higher
prevalences are to be conceptualized as reflecting intense, lengthy NG or really
indicating CG (and detailing still unspecified forms of CG). Suggestions are that,
for other specific subgroups too, prevalence is likely to be much higher, but in
general there is reasonable consistency between the research on the prevalence
of CG and that on resilience. Boerner and colleagues report that most bereaved
persons are resilient. Most suffer from normal grief and some have consistent
minimal distress.
In our view, one must be extremely cautious in making statements about the
prevalence of CG, or in generalizing from any one set of prevalence figures  –
which are frequently presented in terms of simple percentages – to other samples
or populations, particularly in view of the lack of agreement on precisely how to
define and/or operationalize CG.

Different conceptual/theoretical approaches

Different theories and models have been employed in this volume to understand
the phenomena and manifestations of grief and grieving, and to guide clinical
treatment and various sorts of intervention.
One of the most frequently cited theories for understanding phenomena and for
guiding intervention for CG is attachment theory. It figures prominently in diverse
chapters, not surprisingly, given that it is a theory of interpersonal relationships
and that bereavement involves the loss of a close relationship. Mikulincer and
Shaver describe the relevance of this theory’s basic postulates, relating it spe-
cifically to CG. In particular, they demonstrate how theory can be embedded in
empirical research, making use of all kinds of methodologies and paradigms.
 Implications for research and practice  301
This theory is also drawn on by Dyregrov and Dyregrov for understanding CG in
children. Raphael and colleagues link CG to disruption of the attachment to the
deceased, suggesting an ongoing, complicated attachment of the bereaved to the
deceased person. The stressor is regarded as being the disruption of the attach-
ment, its loss. O’Connor draws on attachment theory too, emphasizing the role
of the attachment figure in physiological as well as psychological regulation. She
describes the attachment-specific stress response involving physiological systems
when separated from the attachment figure, relating these reactions specifically
to CG. Kissane and colleagues draw on this theory to frame their family therapy
perspective.
Cognitive (stress) approaches of various sorts also figure prominently. These
range from more general ones such as Boelen and colleagues’ theoretical basis
and specific model following CBT principles for CG (treatment), to identification
of specific cognitive processes, such as Watkins and Moulds’s focus on repetitive
thought (RT). Other approaches within this category that have provided frame-
works for understanding CG include meaning construction (e.g., Kissane et al.)
and the narrative (e.g., Rynearson et al.) perspective. Stress theories have been
employed at different levels and for various purposes, including traumatic stress
theory (Raphael et al.) and general stress response theory (O’Connor).
Contrasting with the cultural perspective of Rosenblatt are psychiatric/medi-
cal model approaches (e.g., Raphael et al.) and physiological/neurological/bio-
logical/genetic ones (e.g., O’Connor; Cooper). To illustrate, in the latter category,
O’Connor points out that underlying aspects of the body’s stress response show
promise in distinguishing CG from non-CG or CG from PTSD. Her work dem-
onstrates that CG is a physiological as well as psychological reaction (and that
different theories can be called on at the same time). She argues that this stress
response may specifically include CG symptoms. If these symptoms are mediated
by attachment, she reasons, then understanding the physiology and neurobiology
of attachment will assist in treating the CG response to bereavement. Such lines
of research are still in their infancy; much needs further exploration, including
answers to questions such as: What genetic vulnerabilities constitute a risk of CG
after bereavement (see O’Connor)? Is CG a biological dysfunction (see Cooper)?
It will have become evident that some investigators employ more than one
theoretical approach and that the CG field in general is characterized by diversity
in theoretical approaches. Indeed, a continuation of theoretical pluralism may be
helpful or even essential in the coming years, given the complexity of the nature
of CG, as already indicated.

CG as a disorder and diagnostic categorization

Another fundamental question in this volume has been whether or not CG should
become a formal mental disorder in the DSM. What are the arguments for and
against this? Many disciplines, from philosophical to psychiatric ones, contribute
to this debate. The issue is particularly topical now, given that  – as we go to
press – CG is a candidate for inclusion in the new edition of the Diagnostic and
Statistical Manual of the American Psychiatric Association (DSM-5).
302  Margaret Stroebe et al.

Definition of CG as a mental disorder

Some authors have reflected on the concept of mental disorder in general, in their
efforts to consider CG as such a category. Providing a general perspective, Cooper
raises issues from her philosophical perspective: What is a (mental) disorder?
What is a dysfunction? She presents different viewpoints: on the one hand the
key question is not whether CG is a mental disorder, but whether it can helpfully
be treated by health care professionals; on the other hand, she also maintains that
CG could plausibly be considered a disorder, which was in line with her general
conclusion.
Psychiatric accounts are naturally linked with the DSM system. DSM-IV
defines a mental disorder as a

clinically significant behavioral or psychological syndrome or pattern that

occurs in an individual and that is associated with present distress (e.g., a
painful symptom) or disability (i.e., impairment in one or more important
areas of functioning) or with a significantly increased risk of suffering death,
pain, disability, or an important loss of freedom. (APA, 1994, p. xxi)

Boelen and Prigerson basically follow the DSM definition of diagnostic disorder,
basing their conclusions (that empirically based PGD can be defined as a formal
disorder) on five taxonomic principles for establishing the validity of a mental
disorder; they argue that these criteria are met and that it should therefore go
into DSM. Wakefield, however, basing his arguments on the same taxonomic
principles, presents other lines of reasoning (e.g., that DSM presents impairment
as a necessary condition for disorder, not one that is sufficient by itself) and con-
cludes that PGD cannot be seen as a mental disorder in terms of DSM. However,
he argues that some form(s) of complicated grief exists, remarking that, since
any biological response can malfunction, it is plausible that some grief disorders
exist. However, in his view, these other potential grief disorders have not yet been
detected.
We return to the above debate in the following section, but here we would
like to stress that, although our authors are in substantial agreement about the
existence of CG as a mental disorder, it must be kept in mind that most bereaved
people do not suffer from it. Even though acute grief is extremely painful and
debilitating, it usually does not need clinical intervention.

Current DSM status: criteria and problems

Currently, there is lack of agreement on criteria for CG inclusion in the DSM-5
system. Boelen and Prigerson include the three different sets of criteria that have
been proposed, namely, for PGD, CG, and bereavement related disorder, in their
Table 7.1. A lot of discussion continues, as reflected in the pages of our volume.
Empirical validation of the different systems is a major concern, with Boelen
and Prigerson pointing to the extensive research basis for PGD/CG, but also rec-
ognizing the need for further study (e.g., search for diagnostic algorithms that
 Implications for research and practice  303
best distinguish those who are at risk from those who are not; examination across
heterogeneous groups of bereaved). Strong voices of dissent about PGD are also
heard. For example, Rando rejects a category based on PGD alone, regarding it as
a subtype in an overarching category.
A major issue is still whether or not CG should go into DSM at all. Boelen and
Prigerson argue that PGD meets the definition of a mental/psychiatric disorder
and should therefore be included in DSM system. Others support this position,
arguing along the lines that, if other disorders such as PTSD are in the DSM
system, why should CG not be (see van den Bout & Kleber)? Cooper’s conclusion
is not so unequivocal (there are arguments on both sides about considering it a
diagnostic category). Wakefield’s title makes his position clear: The proposal to
add a category of CG to DSM is conceptually and empirically unsound, despite
the fact that he regards it as being one of the most thoroughly studied proposals
in DSM history. In his view, the proposed diagnostic criteria identify conditions
that are not due to psychiatric dysfunctions, but are instances of lengthy, intense
normal grieving. Nevertheless, he is respectful of the research efforts of the
Prigerson group (and others), which he acknowledges as having added substan-
tially to understanding symptoms and trajectories of grief and grieving.
Other issues concern specific, highly important details, for example the three
sets of criteria cover different durations of bereavement (the APA DSM-5 work-
group suggests 12 months; others 6 months’ duration at least). PGD/CG criteria
have also changed over the past couple of decades (e.g., differing symptoms) as
investigators have striven to standardize these (making comparison of results of
studies across time difficult). Some problems about CG as a potential new category
of mental disorder apply to other established psychiatric disorders too (see Boelen
& Prigerson), and others point to unique difficulties. For example, Wakefield criti-
cizes the DSM-5 workgroup’s label bereavement related disorder on the grounds
that it is insufficiently specific (many disorders other than disordered grief itself
can be related to bereavement), arguing for the use of the label CG (but he still
points to difficulties distinguishing it from normal grief). Wakefield also argues
for “far more stringent diagnostic criteria than those proposed, if massive false
positive diagnoses are to be avoided” (Chapter 8).
Again, subgroup differences are an issue in the context of DSM inclusion.
For example, Dyregrov and Dyregrov point out that DSM would need to reflect
the uniqueness of children’s grief and that an adult diagnosis should not be used
inappropriately for children. In particular, this relates to the fact that children
have immature systems for emotional and cognitive regulation, and that they are
dependent on adults who may also be grieving (p. 13). Likewise, cultural back-
ground needs to be taken into consideration when reflecting whether the (mental
health) language of DSM and the proposed criteria are applicable (see Rosenblatt).
As Rosenblatt illustrates, assumptions across cultures differ. What we consider
deviant may not be so in a different culture; CG as we see it may not even be
considered a complication. Boelen and Prigerson argue for testing in heterogene-
ous groups, supporting a global grounding for PGD/CG, but is DSM (and are the
proposed criteria) sensitive enough to the full range of cultural differences?
As this volume goes to press, it remains uncertain whether – or, if so, according
304  Margaret Stroebe et al.
to what criteria – CG (or another label) will go into DSM-5. On the one hand,
there are convincing arguments that this should take place; on the other hand,
there are equally good reasons for caution, which leads us to our next point.

(Dis)advantages of the diagnostic category of CG

What are the (dis)advantages of including CG as a category of mental disorder
in the DSM system? Again, there has been much debate but a short overview
must suffice here. Boelen and Prigerson consider benefits and harms, and van den
Bout and Kleber look to the lessons to be derived from the inclusion of PTSD in
DSM. Advantages include the facilitation of empirical research and recognition of
suffering of a significant minority of bereaved having difficulties in their process
of recovery. Boelen and Prigerson argue that it would imply not medicalization of
something normal but a normalization of something that mostly is not but some-
times is indeed pathological. Clients with CG would receive needed treatment
more easily (see van den Bout & Kleber). On the negative side, Wakefield warns
against pathologizing normal grief, using a case study to argue that diagnosis and
treatment may not be the best option in every case and could even “ ‘derail’ such
individuals from the hard work they need to do to change their circumstances and
themselves to create a new life” (Chapter 8). However, others have noted that a
DSM disorder provides the possibility, but not the requirement, to start treatment.
There is need for careful weighing up of the pros and cons of CG entering
DSM-5 as a category of mental disorder. Proponents should not sweep disadvan-
tages under the carpet; opponents need to realize the consequences of exclusion
of such a category for bereaved persons who suffer from CG symptomatology.

CG and other disorders

Various issues relating to CG in the context of other disorders have been raised.
First, is CG a distinct disorder or variant of some other condition? Many perspec-
tives have contributed to our understanding of this issue (e.g., Rando from a clini-
cal orientation, as a major issue in her chapter; Cooper from a philosophical one).
Cooper argued that multiple answers to this question might be justified: For some
purposes it is helpful to consider complicated grief alongside other conditions; for
others it might best be considered separately. Empirical studies have added much
to this debate; for example, Raphael and colleagues point out that recent research
has confirmed that CG, depression, and PTSD are separate syndromes. Again,
O’Connor points to the potential of including neurobiological and physiological
markers in study designs, as these markers can sometimes be used to discriminate
between disorders such as depression and CG, or PTSD and CG, “even when a
clinical gestalt may be murky” (Chapter 15). Rather differently, an important dis-
tinction Boerner and colleagues make is between chronic grief (post-loss onset)
and chronic depression (pre-loss onset), raising the questions: Should there be two
distinct categories of disorder, or is bereavement only an exacerbating factor for
the chronic depressive group?
 Implications for research and practice  305
A related matter concerns CG in the context of (i.e., parallel to) other con-
ditions/disorders. Our authors confirm what has become well established in the
scientific literature: that there is a range of health consequences associated with
bereavement, of which, in most researchers’ minds, CG is one (with the exception
of Rando, who considers other consequences such as physical illness and mortal-
ity as types of presentation of CG). Authors attest to a range of consequences
including but not limited to mental and physical health debilities/disorders and
social dysfunctions. Among children and adolescents, for example, as reviewed
by Dyregrov and Dyregrov, there are both short- and long-term consequences,
ranging from increasing mental health problems, decline in school performance,
social withdrawal, and behavioral problems to somatic complaints. A minority
experience more severe problems, even mortality and increased risk of depression
in adulthood. Finally, adding to the already complex picture, some investigators,
including Burke and Neimeyer, view CG as a risk factor for other psychological
and physical health problems (cardiovascular illness, substance abuse, depres-
sion, anxiety, and overall life disruption).
Not surprisingly, given the range of consequences of bereavement, there is
also comorbidity. For example, Raphael and colleagues focus on CG and PTSD
in adults; Dyregrov and Dyregrov do so among children. Raphael and colleagues
detail the reactions to the different stressors of loss and trauma, with different eti-
ologies and different implications for management (among adults and children).
They clearly separate the two, including a demarcation of the differing trigger-
ing events (see Tables 10.1 and 10.2: phenomena of posttraumatic reactions and
bereavement, and other signs of each reactive process). They also provide empiri-
cal evidence to back up their analyses.
Taken together, these lines of research indicate that bereavement can exac-
erbate other mental disorders and that comorbidity with CG can pertain. One
ongoing concern in this context is when and how to treat coexisting disorders.
Another is the role of pharmacology in therapeutic management. For example,
Raphael and colleagues note that, for CG and PTSD comorbidity, this remains to
be established.

Risk factors, processes, and mechanisms

We have stressed that most people adjust to bereavement in the course of time, but
that a minority experience CG. Research on risk factors, processes, and mecha-
nisms underlying CG all help us understand differential vulnerabilities to CG
between bereaved persons.

Risk factor research

Research on risk factors has much to offer in terms of scientific understanding
of CG and its application. For one thing, it would seem important for health care
professionals to know the characteristics and circumstances that may increase
the likelihood of CG. Identification of some of these (not all) may enable health
306  Margaret Stroebe et al.
care professionals to address some of the issues in the time prior to bereavement,
as Boerner and colleagues point out, for example, to help find additional sup-
port when caregiving begins to be overwhelming; to guide severely distressed
bereaved toward appropriate care services; and to make referrals to clinicians who
can diagnose CG and provide intervention tailored to the individual’s specific
needs.
However, this is a difficult area of research, one that is “rife with complexity,”
in Burke and Neimeyer’s words (Chapter 11). Rather than reviewing the full range
of risk factors here (indeed, Burke and Neimeyer have done that extensively), it
is perhaps useful to highlight these difficulties. These relate to many different
aspects. There are multiple factors covering different types of risk, and there are
also protective factors (e.g., Dyregrov and Dyregrov) and resilient characteristics
(Boerner et al.). Just to illustrate the range: On the one hand, as Dyregrov and
Dyregrov write about bereaved children, a good family climate will be protec-
tive whereas a negative family climate is associated with risks. Post-loss factors
emerge as critical: If death leads to massive changes in the child’s daily environ-
ment, the possibility of negative consequences increases; good parental or primary
care capacity and discipline are protective. They point to the need to strengthen
parental capacity (for intervention too). On the other hand, Kissane and colleagues
describe a different type of risk factor: They identify families at risk versus those
that are well functioning using a Family Relationships Index. So-called sullen and
hostile families were found to be at high risk of complications.
Some variables are poorly defined. Moreover, variables frequently interact
with each other to further increase risk. There are control group problems and
issues to do with causality. Added to these are the facts that some factors are
changeable/modifiable (and therefore psychotherapeutically relevant) whereas
others are not; that there is a huge range within the general risk category, from
mechanisms to protective variables; that some variables are not at all static (e.g.,
belief in an afterlife; bereavement is possibly a crisis time for that); or that there
may be third factors operating, ones that have not been taken into account. Also,
as Burke and Neimeyer point out, some factors may equally predict CG and other
disorders and symptoms (depression, suicidality). Are there universal risk factors,
or to what extent are they impacted by different cultural, economic, and politi-
cal contexts (see Burke & Neimeyer; Rosenblatt)? Also, cross-sectional studies
frequently form the basis for statements about risk factors for CG (no causality
statements are then possible). Burke and Neimeyer point to the need to distinguish
between correlates or consequences of CG and genuinely prospective predictors
per se, and they identify potential and confirmed risk factors in their review.
An important issue touched on above concerns whether the risk factors are
grief-specific or generic. To what extent are the risk factors identified by Burke
and Neimeyer, such as social support, insecure attachment styles, or neuroticism,
reflective of the general associations between these variables and ill-health, which
they are known to have, or to what extent are they unique predictors among the
bereaved, predicting pathology over and above the level of association found in
general? To unravel this, one needs to compare patterns of risk for bereaved with
non-bereaved controls. Obviously grief is non-existent in non-bereaved groups,
 Implications for research and practice  307
but proximal measures (e.g., depression) make comparisons with non-bereaved
people possible. One can, for example, compare depression for (non-)bereaved
men and women and establish the relative excesses between men and women.
What is the way forward in this complex area of research, to establish who is
susceptible to CG? One step that Burke and Neimeyer suggest is a meta-analytical
review of risk factor effect sizes, one that would include other relevant bereave-
ment outcomes (PTSD etc.), different study designs, and diverse samples. Clearly,
this is no easy task, but there may now be sufficient empirical basis to attempt it.

Processes and mechanisms

Scientific investigation in recent years has witnessed a move toward more micro-
level examination of cognitive processes (on CG), as illustrated by Watkins and
Moulds’s focus on repetitive thinking (RT) and Golden’s on autobiographical
memory (AM) and overgeneral memory bias (OGM) processes in CG. These
processes are examples of transdiagnostic processes, which are becoming more
and more familiar in the field of CBT. Repetitive thought (Watkins & Moulds)
encompasses processes that are relevant to normal and complicated grief, the con-
sequences of RT varying according to whether its content is positive or negative,
concrete or abstract. So there are subtypes with distinct functional consequences
(negative and abstract with maladaptive/pathological functioning). Watkins and
Moulds examined RT’s specific role in integrating the loss into existing mental
structures and detailing earlier unspecified operationalizations such as “working
through.” Golden examined AM and OGM specifically in relation to CG, showing
that OGM bias is present in individuals with CG. Such processes have attracted
considerable attention in recent years and undoubtedly represent an important
line of future research as scientists use newly available techniques and apply lab
methods, particularly to unravel more about underlying mechanisms associated
with or underlying CG.
A range of additional processes and mechanisms have come under investigation
in relationship to CG. O’Connor noted that self-regulation at the psychological as
well as physiological level may be important in coping with pangs of grief and
assist in acceptance of the death of an attachment figure. Avoidance (symptoms)
has been examined not only in absent grief but in chronic grief (PGD) too (see
Mikulincer & Shaver; O’Connor). Some knowledge about processes and mecha-
nisms of CG has emerged from research inspired by CBT theorizing. In the work
of Boelen and colleagues, three interrelated processes are identified: (a) insuf-
ficient elaboration and integration of the loss within autobiographical memory;
(b) negative thinking; and (c) anxious and depressive avoidance behaviors. These
authors reason that these three processes interact and play a mediating role in
CG. Wagner’s Internet intervention for CG contrasts with that of Boelen and col-
leagues but is also based on a CB framework. Three modules were included: self-
confrontation, cognitive reappraisal, and social sharing. Quite different processes
emerge from other treatment perspectives, such as the use of resilience as a family
strength and group processes, in the family and psychotherapy group (Kissane et
al.). Rynearson and colleagues’ treatment included processes furthering mastery
308  Margaret Stroebe et al.
of personal safety, confronting the death through “reliving,” and developing an
altered identity through re-engagement with activities and relationships.
Physiological processes/mechanisms are a growing focus of research.
O’Connor provides evidence for a physiological co-regulation basis of CG. She
reviews empirical work supporting the role of physiology and neurobiology in
CG, drawing on her conceptual analysis of two types of physiological stress:
(a) the general stress response (fight-or-flight), similar to CG criteria, including
efforts to avoid thoughts of the deceased, associated with failure to integrate
the reality of the death, leading to continuously realized acute grief; and (b) the
attachment-specific stress response driven by loss of rewarding aspects of attach-
ment, analogous to CG criteria including yearning for the deceased.
Investigators of processes underlying CG in this volume all point to the need
for further empirical testing of their hypotheses. They recommend the use of mul-
tiple methods and extension beyond self-report questionnaires (social cognition
techniques; diary-keeping methods), and incorporating experimental research to
test the proposed directions of causality between variables. Research needs to
establish whether these are the crucial pathways or central mechanisms in influ-
encing clients’ emotional problems, or only epiphenomena.

Treatment of complicated grief

Two topics strike us as particularly noteworthy in the context of professional treat-
ment for CG. First, details to do with the conceptual basis of the programs them-
selves need consideration. Second, efficacy of intervention needs to be addressed.

Principles, paradigms, and procedures

Psychotherapeutic treatments for CG have been described in this book. How have
they gone about treatment? What changes do the treatment programs aim to bring
about? Just as investigators have identified different processes and mechanisms
(described above), so are there differences in the principles, paradigms, and pro-
cedures adopted in interventions.
Not surprisingly, Watkins and Moulds suggest training individuals with CG to
be more concrete or shift to more adaptive forms of RT, whereas Golden argues
that OGM bias should be targeted in therapy, presenting evidence to support this
conclusion. The treatment program of Boelen and colleagues includes different
CBT interventions to achieve various aims: (a) to integrate the loss with existing
knowledge, (b) to identify and alter unhelpful thinking patterns, and (c) to replace
unhelpful avoidance strategies with more helpful ones. Wagner reviews Internet
research on treating CG, covering various forms, from text-based approaches with
therapist feedback to self-help treatments without guidance. Her own interven-
tion follows a treatment program originally developed for PTSD, using structured
writing assignments. Advantages of Internet over traditional methods have been
discussed, including the possibility to interact with others any time one wants;
anonymity and no geographic boundaries; a lower threshold (perhaps for men,
 Implications for research and practice  309
who are less inclined to accept psychotherapy?); the availability of social sup-
port without physical interaction and costs; and time for the therapist to reflect.
However, it may not suit all clients; the potential for misunderstandings is possibly
greater; dealing with a crisis such as suicidality may be more difficult; and there is
a need for careful diagnostic procedures before participation (see Wagner).
Although most therapy for CG has been directed toward individuals, we included
two very different group perspectives. First, over a number of decades, Kissane
and his colleagues have developed a program of family therapy, following the
understanding that loss does not occur in a social vacuum but is shared by people,
commonly the family. It is also different from other treatments in that it is initiated
during anti-cancer treatment before (and continuing through) bereavement, adding
the possibility of preventing the development of CG, by identifying problem fami-
lies, and also (as in other approaches) of ameliorating it. Kissane and colleagues
do not take the stand that family therapy should replace individual modes but argue
that it may be an important adjunct to them. These investigators explain that the
hypothesized mechanism of change is the strengthening of family bonds. Second,
Piper and Ogrodniczuk describe cost-effective time-limited short-term therapy
groups, two different models, specifically for patients who meet CG criteria. They
describe two types of group therapy, interpretive (to enhance patient insight about
repetitive conflicts and traumas associated with their losses) and supportive (the
therapist gratifies the patient, provides guidance, and provides non-interpretive
interventions), and describe their ongoing research on the effectiveness of these.
Different treatment paradigms seem appropriate for different subgroups.
Raphael and colleagues describe assessment and management of CG with
trauma syndromes as comorbidity, giving examples of programs for treating such
comorbidity, for adults and children. Dyregrov and Dyregrov mention the need
to strengthen parental capacity for children, indicating that there may be unique
elements at stake in the case of children with CG.
Other approaches reflect theoretical analyses (and their related empirical
findings). For example, Mikulincer and Shaver provide an attachment theory
perspective on grief resolution and counseling, proceeding on the assumption
that, if attachment insecurities are risk factors for CG, then regaining a sense of
attachment security (e.g., security priming or provision of a secure base) should
ameliorate CG.
The concluding comments on the previous section apply here too. In the con-
text of therapy, we need to establish the specific processes or mechanisms that
bring about favorable outcomes of intervention programs (e.g., Boelen and col-
leagues ask “whether or not CBT indeed produces alleviation of distress because
it lessens negative cognitions and avoidance”; Chapter 16).

Efficacy of intervention for CG

There has been pressure for evidence-based treatment (including RCTs) in recent
decades, and some of the main players in this domain have contributed to this
book. The results of the effectiveness of psychotherapeutic treatment for CG have
310  Margaret Stroebe et al.
been quite positive. Piper and Ogrodniczuk and Wagner reviewed studies (the
latter including recent meta-analysis) of efficacy of intervention programs in gen-
eral, concluding that they are effective for CG. Similar conclusions were drawn
by Wagner for Internet studies, and she gave evidence for effectiveness of her own
program for CG. Nevertheless, limitations were pointed out: Little is as yet known
about the mechanisms through which the structured writing or written disclosure
in these programs leads to improvement in CG, a feature that others identified too
(e.g., Piper & Ogrodniczuk). Kissane and colleagues have been examining the
efficacy of family intervention in improving family communication and report
some indication that family therapy reduced the rate of development of CG
preventively. This is an ongoing program of research in which efficacy is being
further investigated. Boelen and colleagues were able to provide some evidence
for the postulated underlying processes as well as the effectiveness of CT treat-
ment for CG. Piper and Ogrodniczuk performed different trials and compared the
two models of group treatment, documenting patterns of benefits for some groups.
Rynearson and colleagues drew attention to the paucity of research on efficacy of
intervention following the extreme case of bereavement following a violent death,
in which professional treatment may be more needed than following non-violent,
timely deaths.
There are considerable difficulties in conducting such research, as Piper and
Ogrudniczuk describe. For example, follow-up investigation of the efficacy of
an intervention program at later points in time is essential (e.g., symptoms may
increase, before any benefits of the treatment become evident), but this not always
done. There are different models and ways to evaluate efficacy for reducing CG.
Usually this includes treatment and a waiting-list control condition, whereas an
attention-placebo control condition would be better but is not always feasible.
In our view, it is essential that high-quality studies be conducted to establish
the efficacy of clinical intervention for CG. Providing sound evidence that our
therapeutic techniques are effective  – even economically advantageous  – can
help convince governments and funding agencies of the importance of supporting
bereaved people who need help.

Conclusions
In this closing chapter, we have highlighted both the advancements in science as
well as limitations in knowledge about CG that have emerged from the foregoing
chapters of this book. We have indicated directions for future research in this area.
So what about the future perspective in general? In our view, this can build on
the multidisciplinary approach to CG, as represented in the pages of our volume.
There would be advantages to extending this multidisciplinary approach to make
it a truly interdisciplinary one, whereby researchers would collaborate and share
their particular vantage points, working toward a common knowledge base to
acquire deeper understanding of CG. Indeed, some chapters already show evi-
dence of following such interdisciplinary lines, as exemplified in their integration
of different types of theoretical perspectives and multi-method approaches.
 Implications for research and practice  311
This interdisciplinary effort should, we think, not only take effect on a purely
scientific level, but also involve practice (and at times even other societal stake-
holders such as policy makers). In most publications on CG (ours is for the most
part no exception to this) the central idea has been that scientists can provide
knowledge that can be subsequently applied in practice (and in society more
generally). However, the channel of scientific communication and inspiration in
the CG field needs to go both ways: Research needs to look toward practice (and
societal concerns more generally) for much of its impetus too. To illustrate this
from the therapy area: Some clinicians may consider a particular technique to
be effective, but it may not be evident precisely why this intervention works.
Researchers can take note of the therapeutic principles adopted in the therapeutic
approach, and probe further to discover underlying processes that may explain
why the approach or technique is actually effective. So, it is necessary not only
to listen to researchers in order to build on our CG knowledge base, but to listen
to clinicians as well, and to try to understand what they are in fact doing and then
unravel what it is that makes their techniques effective.
Following an interdisciplinary approach and such dual-direction strategies as
that described above will, in our view, lead to greater wisdom concerning the
phenomena and manifestations of complicated grief, and help to provide health
care professionals with a scientifically grounded foundation for conducting their
work with bereaved persons.

References
APA. (1994). Diagnostic and statistical manual of mental disorders (4th edn.). Washington,
DC: American Psychiatric Association.
Bowlby, J. (1980). Attachment and loss, vol. 3. Sadness and depression. New York: Basic
Books.
Stroebe, M. S., Hansson, R. O., Schut, H., & Stroebe, W. (2008). Handbook of bereave-
ment research and practice: Advances in theory and intervention. Washington, DC:
American Psychological Association.
Author index

Adams, K.M. 29 Bisconti, T.L., Bergeman, C.S., and Boker,

Adelson, N. 32 S.M. 58–9
Admoni, S. 199 Bisson, J.I., Brayne, M. et al. 124
Ainsworth, M.D.S. and Eichberg, C.G. 249 Black, D. and Urbanowicz, M.A. 249
Akiskal, H. 16 Bluck, S. et al. 176
Allumbaugh, D.L. and Hoyt, W.T. 237, Boelen, P.A. xi, 6, 8, 85–97, 186, 221–32,
238, 272 296, 298, 300
Anderson, W. et al. 134 Boelen, P.A. and van den Bout, J. 22, 48,
Andersson, G. 235 87
Andersson, G. et al. 241, 242 Boelen, P.A., van den Bout, J., and van den
Aneshensel, C.S. et al. 57, 58, 59 Hout, M.A. 70, 152, 153, 156, 158n3,
Appelo, M.T. et al. 126 165, 194, 170, 171
Armour, M. 47, 284 Boelen, P.A. and van den Hout, M.A. 229
Asaro, M. 284 Boelen, P.A., van den Hout, M.A., and van
Asukai, N., Tsuruta, N. and Saito, A. 288 den Bout, J. 119, 163, 169, 178, 179,
Averill, J.R. and Nunley, E.P. 34 181, 185, 193
Boelen, P.A., Huntjens, R.J.C. et al. 183
Baddeley, A.D. 176, 179 Boelen, P.A., de Keijser, J. et al. 88, 172,
Bailley, S., Kral, M. and Dunham, K. 282, 221, 231, 238
285 Boelen, P.A., Keijers, L., and van den
Balk, D.E. et al. 36 Hout, M.A. 223, 231
Barak, A., Hen, L. and Boniel-Nissim, M. Boelen, P.A. and Klugkist, I. 230
235 Boelen, P.A. and Prigerson, H.G. 213
Barker, J. 29 Boelen, P.A., van de Schoot, R. et al. 129
Barry, L., Kasl, S. and Prigerson, H. 280 Boelen, P.A., Stroebe, M. et al. 48
Bartels, A. and Zeki, S. 211 Boerner, K. xi, 55–65, 297
Bauer, J. and Bonanno, G.A. 171 Boerner, K., Schulz, R., and Horowitz, A.
Bayer, R. 14 63
Beck, A.T. 222, 229, 230 Boerner, K., Wortman, C.B., and Bonanno,
Benedek, D. and Ursano, R. 135 G.A. 57, 59, 60
Bennett-Levy, J. et al. 227, 228 Bolton, D. 20, 21, 24n1, 122
Berant, E., Mikulincer, M., and Shaver, Bonanno, G.A. xi, 5, 55–65, 120, 195
P.R. 196 Bonanno, G.A., Boerner, K., and Wortman,
Berger, M. and Riemann, D. 211 C.B. 43, 47, 61
Bergner, A. et al. 241 Bonanno, G.A. and Field, N.P. 57, 59, 73
Berridge, K.C., Robinson, T.E., and Bonanno, G.A. and Kaltman, S. 104, 145,
Aldridge, J.W. 206 163, 164, 167
Berry, D.S. and Pennebaker, J.W. 239 Bonanno, G.A., Kelter, D. et al. 57
Birch, L.S. and Davidson, K.M. 180 Bonanno, G.A. and Mancini, A.D. 145
 Author index  313
Bonanno, G.A., Moskowitz, J.T. et al. 57, Callahan, J. 150, 152
58 Camarena, B., Cruz, C. et al. 213
Bonanno, G.A., Neria, Y. et al. 87, 88, 89, Campbell, J.D., Trapnell, P.D. et al. 223
94, 209 Caplan, P.J. 27
Bonanno, G.A., Papa, A., and O’Neill, K. Carr, D.S. 150, 152
163, 167 Carr, D. 44, 45
Bonanno, G.A., Papa, A. et al. 103, 110, Carver, C.S. and Scheier, M.F. 167–8
150, 152, 162, 163, 164, 170, 171, 229 Cassidy, J. and Kobak, R.R. 191
Bonanno, G.A., Wortman, C.B., Lehman, Cassidy, J., Shaver, P.R. et al. 199, 200
D.R. et al. 58, 59, 61, 62, 153, 154, 207 Cerel, J., Fristad, M.A. et al. 68
Bonanno, G.A., Wortman, C.B., and Chan, E.K., O’Neill, I. et al. 256
Nesse, R.M. 58, 61, 62, 164, 165, 171, Charmaz, K. and Milligan, M.J. 27, 28, 30,
249, 250 32, 34
Boorse, C. 13, 14, 15, 19 Chentsova-Dutton, Y. et al. 57
Boorse, C. and Wakefield, E. R. 16 Christensen, H., Griffiths, K.M. et al. 242
Borkovec, T.D., Ray, W.J., and Stöber, J. Clarke, G. et al. 242
169 Cleiren, M. 151, 152, 153, 279, 281
Borkovec, T.D. and Roemer, L. 169 Clements, P. and Burgess, A. 284
Boschen, M.J. 115 Coffino, B. 69
Boss, P. 250, 251 Cohen, J.A. 231
van den Bout, J. xi, 3–9, 115–26, 221–32, Cohen, J.A., Goodman, R.F. et al. 72
295–311 Cohen, J.A., and Mannarino, A. 130, 139
Bower, J.E., Kemeny, M.E. et al. 162, 163, Cohen, J.A., Mannarino, A., and Starlon,
166, 167 V. 75, 138
Bowlby, J. 45, 47, 56, 71, 72, 163, 190–6, Cohen, J.A., Mannarino, A.P., and
198, 200, 205, 223, 249, 297, 298, 299; Knudsen, K. 75
concern for misconstrual of potential Cohen, J.A., Mannarino, A.P., and Staron,
severity and duration of normal grief V.R. 72, 286
100–1 Coleman, R.A. and Neimeyer, R.A. 153
Bradley, R., Greene, J. et al. 238 Conway, M.A. 178
Brave Heart, M.Y.H. and DeBruyn, L.M. Conway, M.A. and Pleydell-Pearce, C.W.
30 176, 178, 179, 180, 183, 222
Brennan, K.A., Clark, C.L., and Shaver, Cook, J.E. and Doyle, C. 242
P.R. 191 Cooper, R. xii, 5, 13–24, 122
Brent, D., Melhelm, N. et al. 69, 70, 75, Creamer, M., Burgess, P., and Pattison, P.
135, 136 249
Breslau, N. et al. 119 Cribb, G., Moulds, M.L., and Carter, S.
Brewin, C.R., Dalgleish, T., and Joseph, 170
S. 180 Cronbach, L.J. and Meehl, P.E. 273
Brom, D., Kleber, R., and Defares, P. 285 Currer, C. 30
Brom, D. and Kleber, R.J. 120 Currier, J.M., Holland, J.M. and Neimeyer,
Brom, D., Kleber, R.J., and van den Bout, R.A. 74–75, 280
J. 115 Currier, J.M., Holland, J. et al. 151, 280,
Brown, E.J. et al. 69, 70 281, 282, 283, 287
Brown, S.L., Nesse, R.M. et al. 153, 154 Currier, J.M. and Neimeyer, R.A. 287
Bryant, R.A., Sutherland, K., and Guthrie, Currier, J.M., Neimeyer, R.A. and Berman,
R.M. 177 J.S. 237, 238, 272
Buckley, T., McKinley, S. et al. 109, 208
Burke, L.A. xii, 7, 145–58, 297 Dalgleish, T. 222, 229
Burke, L.A., Neimeyer, R.A., and Dalgleish, T. and Power, M.J. 180, 181
McDevitt-Murphy, M.E. 156, 157, 282, Dalgleish, T., Rolfe, J. et al. 179
284 Dalgleish, T., Tchanturia, K. et al. 179
Burnett, P., Middleton, W. et al. 130, 147 Dalgleish, T., Williams, J.M.G. et al. 178
Byrne, G. and Raphael, B. 130 Davidson, K., Foa, E. et al. 122
314  Author index
Davis, C.G., Lehman, D.R. et al. 165 Folkman, S. and Moskowitz, J.T. 250
Davis, C.G., Nolen-Hoeksema, S., and Foote, C.E. and Frank, A.W. 32, 34
Larson, J. 166 Forstmeier, S. and Maercker, A. 281
Davis, C.G., Wortman, C.B. et al. 171 Fortner, B.V. 272
Davis, N.L. 157 Fraley, R.C and Bonanno, G.A. 196, 197,
Davison, K.P., Pennebaker, J.W. and 198
Dickerson, S.S. 236 Fraley, R.C., Garner, J.P., and Shaver, P.R.
de Keijser, J. 125 195
Deckert, J. et al. 213 Fraley, R.C. and Shaver, P.R. 191, 195
Depue, R.A. and Morrone-Strupinsky, J.V. Frances, A. 17
205, 206 Freud, S. 129
Deutsch, H. 47 Fujisawa, D. et al. 87, 88, 280, 289
Dickson, J.M. and Bates, G.W. 177
van Doorn, C., Kasl, S.V. et al. 198 Galea, S. et al. 131
Dowdney, L. 68 Gamino, L.A., Sewell, K.W., and
Dowdney, L. et al. 68, 71, 76 Easterling, L.W. 151, 154
Dumont, I. and Kissane, D.W. 253 del Gaudio, F. xii, 8, 248–59
Dupré, J. 22, 23 Gerra, G. et al. 206, 207
Dyregrov, A. xii, 6, 68–77, 296, 298, 299 Ghaffani-Nejad, A. et al. 281
Dyregrov, A. et al. 74 Gibbs, B.R. and Rude, S.S. 177
Dyregrov, K. xii, 6, 68–77, 296, 298, 299 Gilbert, P. 185
Dyregrov, K., Nordanger, D. and Gillath, O. et al. 195
Dyregrov, A. 70, 281, 282 Golden, A.-M. xii–xiii, 7, 176–86
Golden, A.-M., and Dalgleish, T. 129
Easterling, L.W., Gamino, L.A. et al. 154 Golden, A.-M., Dalgleish, T., and
Edwards, B. and Lavery, V. 251 Mackintosh, B. 177, 178, 179, 180,
Ehlers, A. 165, 169, 228 182, 183, 184, 185
Ehlers, A. and Clark, D.M. 177, 222, 229 Goldsmith, B. et al. 146, 150
Ehlers, A., Mayou, R.A. and Bryant, B. Goldstein, J., Alter, C.L., and Axelrod, R.
165 248, 249
Ehring, T. et al. 169 Gone, J.P. 32
Ehring, T., Szeimies, A.-K. and Schaffrick, Good, B.J. and DelVechio, M.-J. 33
C. 167 Goodheart, C.D., Kazdin, A.E., and
Elkin, I., Shea, T. et al. 270 Sternberg, R.J. 273
Elliot, A.J. and Reis, H.T. 225 Goodman, R.F. et al. 71, 72
Engel, G. and Wilkinson, 20 Goodrum, S. 284
Greenberg, M.A. 162
Fabrega, Jr., H. 27, 33 Gross, J. 63
Faschingbauer, T.R., Zisook, S., and De Gündel, H. et al. 210
Vaul, R. 87 Gunnar, M. and Quevedo, K. 70
Fava, G.A. et al. 270 Gyurak, A. Gross, J.J., and Etkin, A. 73
Feigelman, W., Gorman, B.S. et al. 236
Feigelman, W., Jordan, J.R., and Gorman, Haine, R.A. et al. 69, 76
B.S. 137, 151 Hardison, H., Neimeyer, R., and Lichstein,
Feynman, Richard 105 K. 281
Field, N.P. and Bonanno, G.A. 229 Hartz, G. 41
Field, N.P and Filanosky, C. 48 Harvey, A. et al. 177
Field, N.P. and Sundin, E.C. 196, 197 Hayes, S.C. et al. 224
Field, N.P. and Wogrin, C. 156 Hazan, C., Gur-yaish, N., and Campa, M.
Figley, Charles 116 205
Foa, E.B. and Jaycox, L.H. 239 Hazan, C. and Zeifman, D. 192
Foa, E.B and Kozak, M.J. 169, 170 Hebert, R.S. et al. 63
Foa, E.B. and Rothbaum, B.O. 194, 224, Hebert, R.S., Schulz, R. et al. 63–4
226 Heinrichs, M. et al. 207
 Author index  315
Helm, B. 17 Kalantari, M. et al. 77
Hendriks, M.C., Rottenberg, J., and Kaltenthaler, E. et al. 242
Vingerhoets, A.J. 208 Kaltman, S. and Bonanno, G.A. 64, 279,
Hensley, P. 137 280
Herbert, R.S., Dang, Q., and Schulz, R. 63 Kato, P.M. and Mann, T. 237, 238, 272
Hofer, M.A. 205, 212 Keane, T.M., Marshall, A.D., and Taft,
Hofer, M.A. and Horowitz, M. 207 C.T. 119
Hofer, M.A., Wolff, C.T. et al. 206 Keesee, N.J., Currier, J.M., and Neimeyer,
Hogan, N. 74 R.A. 146, 150, 151
Hogan, N., Worden, J.W., and Smidt, L. Kelly, L. 237
87 Kendler, K. 19
Hogan, N., Worden, J.W. et al. 41 Kendler, K., Myers, J. et al. 19
Hollan, D.W. 28 Kendler, K.S. 125
Holland, J.M., Neimeyer, R.A. et al. 87, Kersting, A., Brähler, E. et al. 87, 281, 289
103, 145, 147 Kersting, A., Dorsch, M. et al. 241
Horne, C. 282 Kersting, A., Kroker, K. et al. 154, 213,
Horowitz, M.J. 117 236, 241
Horowitz, M.J., Bonanno, G., and Holen, Kirmayer, L. 33
A. 41, 85 Kissane, D.W. 8, 248–59, 299
Horowitz, M.J., Marmar, C.R. et al. 266 Kissane, D.W. and Bloch, S. 248, 252, 254
Horowitz, M.J., Siegel, B. et al. 41, 55, 85, Kissane, D.W., Bloch, S. et al. 248, 249
90, 102, 106, 209 Kissane, D.W. and Hooghe, A. 252
Horowitz, M.J., Wilner, N., and Alvarez, Kissane, D.W., Lichtenthal, W., and
W. 270 Zaider, T. 249
Horwitz, A. 15 Kissane, D.W, McKenzie, M. et al. 248
Horwitz, A. and Wakefield, J. 15, 109 Klass, D. 31
van den Hout, M.A. xiii, 8, 221–32, 296 Klass, D. et al. 29
van der Houwen, K., Schut, H. et al. 236, Kleber, R.J. xiii, 6, 115–26
241, 242, 244 Knutson, B., Adams, C.M. et al. 210
van der Houwen, K., Stroebe, M. et al. Kosfeld, M., Heinrichs, M. et al. 207
152, 179, 185, 229, 230, 236, 237 Kovacs, G.L., Sarnyai, Z., and Szabo, G.
206
Insel, T.R. 206 Kristensen, P., Weisaeth, L., and Heir, T.
Ito, T. et al. 165 133
Kumar, S.M. 185
Jacobs, J. xiii, 6, 129–39, 298 Kuyken, W. and Brewin, C.R. 179, 180
Jacobs, J.R. and Bovasso, G.B. 69 Kuyken, W. and Dalgleish, T. 177
Jacobs, S.C 41, 45, 60
Jacobs, S.C., Kasl, S.V. et al. 208 Landau, J. 251
Jacobs, S.C., Mason, J. et al. 206 Lang, A. and Gottlieb, L. 149
Jacobson, N.S., Follette, W.C. and Lange, A., Rietdijk, D. et al. 235, 239, 241
Ravenstorf, D. 270 Lange, A., van de Ven, J.P. et al. 239
Jacobson, N.S., Martell, C.R. and Langner, R. and Maercker, A. 213
Dimidjian, S. 225, 228 Larson, D.G. and Hoyt, W.T. 272
Jacobson, N.S. and Ravenstorf, D. 270 Latham, A. and Prigerson, H. 48, 157
Janoff-Bulman, R. 162, 224, 250, 278 Laurie, A. and Neimeyer, R.A. 150, 152
Janoff-Bulman, R. and Berg, M. 250 Layne, C.M., Kaplow, J., and Pynoos, R.S.
Jerga, C., Shaver, P.R. and Wilkinson, R.B. 74
196, 197 Layne, C.M., Saltzman, W.R. et al. 75,
Johnson, C. 282 285, 286
Johnson, J.G. et al. 90 Layne, C.M., Savjak, N. et al. 74
Jones, M. et al. 47 Lehman, D., Wortman, C., and Williams,
Jordan, J. 284 A. 284
Jordan, J. and McIntosh, J. 284 Lepore, S.J. et al. 163, 164
316  Author index
Li, L.W. 57 Melhem, N. et al. 69, 135, 136
Li, Y. xiii, 8, 248–59 Mezzich, J.E. 36
Lichtenthal, W.G. and Cruess, D.G. 77 Michael, S. and Snyder, C. 165
Lichtenthal, W.G., Cruess, D. et al. 22 Middleton, W., Burnett, P. et al. 59
Lichtenthal, W.G., Cruess, D.G., and Middleton, W., Moylan, A. et al. 56, 130
Prigerson, H.G. 103, 146, 164 Middleton, W., Raphael, B. et al. 45, 48,
Lieberman, S. 248, 258 50
Lifton, R. 116 Mikulincer, M., Dolev, T., and Shaver, P.R.
Lilienfeld, S. and Marino, L. 20 196
Lim, M.M. and Young, L.J. 206 Mikulincer, M., Hirschberger, G. et al. 199
Lin, K.K. et al. 69, 70, 73 Mikulincer, Mario xiv, 7, 190–200, 223,
Lobb, E.A. et al. 146 298, 299
Looi, J.C.L. xiii, 6, 129–39, 298 Mikulincer, M. and Orbach, I. 193
Luecken, L.J. 68, 69 Mikulincer, M., Shaver, P.R., and Horesh,
Lund, D.A. et al. 57 N. 199
Luoma, J. and Pearson, J. 48 Moberly, N.J. and MacLeod, A.K. 184
Lutz, C. 28 Mol, S.S. et al. 118
Momartin, S., Silove, D. et al. 150, 151
de Maat, S., Dekker, J. et al. 270 Monk, T.H. et al. 211
Maccallum, F. and Bryant, R.A. 183, 184, Moore, S.A. and Zoellner, L.A. 177
185, 186, 230 Moos, R.H. and Moos, B.S. 251
McClatchey, I.S., Vonk, M.E., and Palardy, Morina, N., Rudari, V. et al. 134, 284
G. 75 Morina, N., Von Lersner, U., and
McDaid, C., Trowman, R. et al. 286 Prigerson, H.G. 87, 88
McDermott, O.D. et al. 211 Moulds, M.L. xiv, 7, 162–72
McDevitt-Murphy, M.E. et al. 146 Moulds, M.L., Kandris, E. et al. 170
McHugh, P.R. and Treisman, G. 116 Murphy, S. 279, 285, 286
Maciejewski, P. et al. 106 Murphy, S., Baugher, R. et al. 286, 287
McKenna, K. and Bargh, J. 236 Murphy, S. Johnson, C. et al. 285, 286
McKenna, K. and Seidman, G. 236 Murphy, S.A., Johnson, L.C. et al. 64
McKenzie, K.R. 250
MacLatchey, I., Vonk, M., and Palardy, G. Nager, E.A. and De Vries, B. 197
282 Neimeyer, R.A. xiv, 7, 145–58, 183, 250,
McNally, R.J. 117, 119 272, 297
McNally, R.J. et al. 177, 183 Neimeyer, R.A., Baldwin, S.A., and
Magnavita, J.J. 273 Gillies, J. 150, 183
Malkinson, R. and Bar-Tur, L. 28 Neimeyer, R.A., Burke, L. et al. 156
Mancini, A.D. xiii–xiv, 5, 55–65 Neimeyer, R.A. and Currier, J. 44
Mancini, A.D., Bonanno, G.A. and Clark, Neimeyer, R.A., Harris, D. et al. 156
A. 60 Nelson, E.E. and Panksepp, J. 206
Mancini, A.D., Prati, G. and Black, S. 64, Neria, Y. and Litz, B. 129
279, 281, 283 Neria, Y. et al. 133
Mancini, A.D., Prati, G. and Bonanno, Neria, Y., Nandi, A., and Galea, S. 125
G.A. 278, 279 Nolen-Hoeksema, S. 162, 178
Mann, J.R. et al. 241 Nolen-Hoeksema, S., McBride, A. and
Martell, C.R., Addis, M.E., and Jacobsen, Larson, J. 162, 164
N.S. 169 Nolen-Hoeksema, S., Parker, L.E., and
Martin, L.L. and Tesser, A. 163 Larson, J. 164, 171
Maschio, T. 28 Nolen-Hoeksema, S., Wisco, B.E. and
Matthews, L.T. and Marwit, S.J. 278 Lyubomirsky, S. 169
Mawson, D. et al. 230 Nordenfelt, L. 16
Mayou, R.A., Ehlers, A., and Bryant, B. Norris, F. 280
165, 224 Norris, F. and Wind, L. 131
Mealey, L. 16
Megone, C. 16 O’Connor, M.-F. xiv, 8, 204–15, 297
 Author index  317
O’Connor, M.-F. et al. 88, 209 Prigerson, H.G., Shear, M.K. et al. 41, 85,
Ogrodniczuk, J. xiv, 8, 263–75 90, 96, 103
van der Oord, S. et al. 77 Prigerson, H.G. and Vanderwerker, L.C.
Osterweis, M., Solomon, F., and Green, F. 125
47, 56 Prigerson, H.G., Vanderwerker, L. and
Ott, C.H. 157 Maciejewski 22, 41, 42, 48, 106
Prisciandaro, J.J. and Roberts, J.E. 214
Palmqvist,B., Carlbring, P., and Andersson, Pynoos, R.S., Frederick, C. et al. 130
G. 242 Pynoos, R.S. and Nader, K. 282
Panksepp, J., Knutson, B., and Burgdorf, Pynoos, R.S., Nada, K. et al. 130
J. 206 Pynoos, R.S., Steinberg, A., and Brymer,
Panksepp, J., Nelson, E., and Bekkedal, M. 130, 139
M. 206
Park, C. and Halifax, J. 156 Radden, J. 17
Parkes, C.M. 48, 56, 250, 278 Raes, F., Williams, J.M.G., and Hermans,
Parkes, C.M. and Prigerson, H.G. 109 D. 185
Parkes, C.M. and Weiss, R.S. 45, 47, 104, Ramponi, C., Barnard, P., and Nimmo-
105, 192 Smith, I. 177
Paul, N. and Grosser, G. 249 Ramsay, R.W. 25, 124, 224
Pearlman, L., Wortman, C. et al. 47 Rando, T.A. xv, 5, 40–51, 157, 278, 296,
Pearlman, M.Y. Schwalbe, K.D. and 298
Cloitre, M. 68 Raphael, B. xv, 6, 129–39, 298
Pennebaker, J.W. 77 Raphael, B., Martinek, N., and Wooding,
Pennebaker, J.W. and Seagal, J.D. 169 S. 119, 131
Pennebaker, J.W. et al. 239 Raphael, B. and Minkov, C. 130
Pfeffer, C. et al. 68 Raphael, B., Minkov, C. et al. 48
Piper, W.E. xv, 8, 263–75 Rauch, P.K. and Muriel, A.C. 249
Piper, W.E., Debbane, E.G. et al. 274 Redmond, A. 34
Piper, W.E., Joyce, A.S. et al. 268 Redmond, A. and Pratt, L. 34
Piper, W.E., McCallum, M., and Azim, H. Reynolds, C.F. III et al. 87
263, 268 Reznek, L. 16
Piper, W.E., Ogrodnic, J.S., and Riches, G. and Dawson, P. 69
Weideman, R. 264 Richman, K. 16
Piper, W.E., Ogrodnic, J.S. et al. 263, 265, Robins, E. and Guze, S.B. 22, 86
269 Rockland, L.H. 268
van Praag, H.M. 116, 121 Rohlof, H. 36
Pratt, L. 34 Roisman, G.I., Tsai, J.L., and Chiang,
Prigerson, H.G. xv, 6, 41, 43, 55, 85–97, K.H.S. 194
110, 112, 254, 256, 298 Romans, S.E. and Clarkson, R.F. 208
Prigerson, H.G., Ahmed, I. et al. 150–2 Rosch, E. 20
Prigerson, H.G., Bierhals, A.J. et al. 102, Rose, S.D. 271
103, 157, 215n2 Rosen, G.M. and Lilienfeld, S.O. 117, 118,
Prigerson, H.G., Frank, E. et al. 85, 89, 120
145, 146, 208 Rosenblatt, Paul C. xv–xvi, 5, 16, 27–37,
Prigerson, H.G., Horowitz, M.J. et al. 68, 299
71, 86, 87, 90, 95, 99, 102, 108, 109, Rosenthal, P.A. 248
110, 115, 145, 146, 164, 197, 208, Rosner, R., Kruse, J., and Hagl, M. 74
215n1, 221 Rostila, M. and Saarela, J.M. 68, 69
Prigerson, H.G. and Jacobs, S.C. 41, 85, Rotheram-Borus, M.J., Stein, J.A., and
86, 87, 147, 209 Lin, Y.Y. 249
Prigerson, H.G. and Maciejewski, P.K. Rotheram-Borus, M.J., Weiss, R. et al. 249
101, 102, 103, 107, 177, 256 Rottenberg, J., Hildner, J.C., and Gotlib,
Prigerson, H.G., Maciejewski, P.K. et al. I.H. 182
74, 85, 86–7, 87, 90, 101, 102, 103, Rowa-Dewar, N. 237
135, 146, 147, 209, 211, 213, 231 Rubin, S.S. 48, 192
318  Author index
Rubin, S.S., Malkinson, R. et al. 41, 50 Shear, K., Jackson, C. et al. 136, 146
Ruwaard, J. et al. 235 Shear, K., Monk, T. et al. 100–1, 137, 207,
Rynearson, E.K. xvi, 9, 135, 278–89, 297 229
Rynearson, E.K., Correa, F. et al. 286, Shear, K. and Mulhare, E. 103
287–8 Shear, K. and Shair, H. 101, 185 207, 223
Rynearson, E.K. and McCreery, J. 129, Shear, K., Simon, N. et al. 86, 89, 91, 94,
135 95, 99, 104, 106, 115, 207, 208
Rynearson, E.K. and Sinnema, A. 279 Shear, K. et al. 112
Silverman, G.K., Johnson, J.G., and
Saldinger, A., Cain, A., and Porterfield, Prigerson, H.G. 196
K. 70 Silverman, P.R. and Worden, J.W. 68, 75,
Saldinger, A. et al. 73, 76 76
Saldinger, A., Porterfield, K., and Cain, Simon, N.M. et al. 91, 106, 138
A.C. 69 Sireling, L., Cohen, D., and Marks, I. 230
Saler, L. and Skolnick, N. 70 Small, N. 34
Salloum, A. 286 Small, N. and Hockey, J. 33
Salloum, A. and Overstreet, S. 74 Smyth, J., True, N., and Souto, J. 169
Sanders, C.M. 146 Solomon, R. 17
Sandler, I.N. et al. 76, 249 Spek, V. et al. 235
Savitz, D.A., Hertz-Picciotto, I. et al. 240 Spitzer, R.J. 15, 124
Sbarra, D.A. and Hazan, C. 205, 206 Spitzer, R.J., First, M.B. and Wakefield,
Scheeringa, M.S. 71 J.C. 119, 124
Scheier, M.F. and Carver, C.S. 240 Spooren, D.J., Hendrick, H. and Jannes, C.
Schmiege, S.J. et al. 69 150, 151
Schmiege, S.J., Khoo, S.T. et al. 69 Stearns, P.N. 28, 35
Schnurr, P.P. 115 Steer, R., Ball, R. et al. 255
Schulz, R., Boerner, K. et al. 63 Steer, R., Brown, G. et al. 255
Schulz, R., Mandelson, A.B. and Haley, Stein, D.J. et al. 88, 89, 108, 121
W.E. 57, 58, 61, 63 Stein, N., Folkman, S. et al. 166, 171
Schut, H. xvi, 3–9, 278–89, 295–311 Stöber, J. and Borkovec, T.D. 169
Schut, H.A.W. and Stroebe, M.S. 44, 124, Stone, A. 23
238 Stroebe, M.S. xvi, 3–9, 278–89, 295–311
Schut, H.A.W., Stroebe, M.S. et al. 237, Stroebe, M.S, Boelen, P. et al. 48, 162, 170
238, 242, 272 Stroebe, M.S., Gergen, M.et al. 18, 33
Schwartzberg, S.S. and Janoff-Bulman, Stroebe, M.S., Hansson, R.O. et al. 45, 47,
R. 229 72, 193, 279, 296
Segal, Z.V., Williams, J.M.G., and Stroebe, M.S and Schut, H.A.W. 22, 27,
Teasdale, J.D. 185 36, 47, 50, 51, 103, 163, 192, 249, 288
Segerstrom, S.C. et al. 167 Stroebe, M.S., Schut, H.A.W. and Stroebe,
Segerstrom, S.C., Stanton, A.L., and W. 198, 208
Shortridge, B.E. 162 Stroebe, M.S., Schut, H.A.W. et al. 47, 48
Shaffer, D. et al. 111 Stroebe, M.S., van Son, M. et al. 40, 41,
Shapiro, E.R. 31, 250 102, 125
Shatan, C. 116 Stroebe, W. and Schut, H. 146
Shaver, P.R. xvi, 7, 190–200, 298, 299 Stroebe, W., Schut, H., and Stroebe, M.
Shaver, P.R. and Fraley, R.C. 192 138, 139, 146, 163, 170
Shaver, P.R. and Tancredy, C. 249 Suler, J. 243
Shaver, P.R., Schachner, D.A. and Sumner, J.A., Griffith, J.W., and Mineka,
Mikulincer, M. 193 S. 177
Shear, K. 48, 75 Sveen, C.A. and Walby, F. 284, 285
Shear, K., Boelen, P.A., and Neimeyer, Szumalis, M. and Kutcher, S. 286
R.A. 221, 225, 226, 228, 231
Shear, K., Frank, E. et al. 55, 88, 137, 138, Tafoya, N. and Del Vecchio, A. 30
145, 172, 221, 226, 230, 231, 238, 288 Tarakeshwar, N. et al. 150
 Author index  319
Taylor, L.K. et al. 73 Werman, D.S. 268
Teasdale, J.D. 222 Whitaker, D.S. and Lieberman, M.A. 249,
Teasdale, J.D. and Fogarty, S.J. 176 250
Tedeschi, R.G. and Calhoun, L.G. 166 White, M. 258
Tremblay, G.C. and Israel, A.C. 68, 69, Wickie, S. and Marwit, S. 283
70, 76 Wierzbicka, A. 30
Treynor, W., Gonzalez, R., and Nolen- Wijngaards-de Meij, L. et al. 57, 153, 196,
Hoeksema, S. 166 197, 225
Trivedi, M.H. et al. 270 Wikan, U. 28
Tucker, G.J. 122 Wilfley, D.E. et al. 266, 271
Tucker, M. and Oei, T.P.S. 274 Wilkinson, S. 19, 20
Williams, H.L. Conway, M.A., and Cohen,
Vachon, M.L.S. et al. 57, 64 G. 176
Vallacher, R.R. and Wegner, D.M. 168 Williams, J.M.G. 178, 179
Vanderwerker, L.C. and Prigerson, H.G. Williams, J.M.G., Barnhofer, T., et al. 177,
154 178, 179, 180, 182
Vanderwerker, L.C., Jacobs, S.C. et al. 196 Williams, J.M.G., Chan, S. et al. 177
Vessier-Batchem, M. and Douglas, D. 281 Williams, J.M.G. and Broadbent, K. 177
Vicary, A.M. and Fraley, R.C. 237, 243 Williams, J.M.G. and Dritschel, B.H. 178
Viederman, M. 105 Williams, J.M.G., Stiles, W.B., and
Shapiro, D. 179
Wagner, B. xvi, 8, 235–44 Williams, J.M.G., Teasdale, J.D. et al. 185
Wagner, B., Knaevelsrud, C., and Williams, W.V. and Polak, P.R. 249
Maercker, A. 221, 226, 230–1, 236, Wilson, D. 16
238, 239, 240, 243 Wittouck, C. et al. 237, 238
Wagner, B. and Maercker, A. 139, 240, Worden, J.W. 50, 69, 70, 71, 72
242, 243 Worden, J.W., Davies, B. and McCown,
Wagner, B., Schulz, W., and Knaevelsrud, D. 68
C. 243 Worden, J.W. and Silverman, P.R. 68, 69
Wakefield, J.C. xvi–xvii, 14, 15, 16, 24n1, Wortman, C.B. and Boerner, K. 57–8, 59,
99–112, 297 60, 61, 64
Walter, T. 18, 32, 34 Wortman, C.B. and Silver, R.C. 43, 56, 57,
Waskowic, T.D. and Chartier, B.M. 197, 59, 64, 125
198
Watkins, E.R. xvii, 7, 162–72 Yamada, K. et al. 166
Watkins, E.R., Baeyens, C.B. and Read, R. Young, A. 117
171, 172 Young, L.J. and Wang, Z. 206
Watkins, E.R. and Baracaia, S. 168 Yule, W. et al. 77
Watkins, E.R., Moberly, N.J. and Moulds,
M.L. 167 Zaider, T. and Kissane, D. 251
Watkins, E.R., Scott, J. et al. 172 Zaider, T.I. xvii, 8, 248–59
Watkins, E.R. and Teasdale, J.D. 168, 185 Zhang, B., El-Jawahri, A., and Prigerson,
Watkins, E.R., Teasdale, J.D. and H. 138
Williams, R.M. 185 Zhang, B., Mitchell, S.L. et al. 57, 58, 59
Watters, E. 34, 35 Zinzow, H., Rheingold, A. et al. 135
Wayment, H.A. and Vierthaler, J. 196, 197, Zisook, S. and Lyons, L. 264–5
198 Zisook, S. and Shear, K. 42, 104
Weiss, R.S. 192 Zisook, S. and Shuchter, S.R. 57, 59
Wellenkamp, J.C. 28 Zuckoff, A. et al. 137
Weller, E.B., Weller, R.A. et al. 75 Zuroff, D.C. and Blatt, S.J. 200
Subject index

abnormality 16, 56; abnormal attachment figure, role of 205, 206–7

environmental factors, responses to attachment insecurities and disordered
108; abnormal grief reaction 57, 90, patterns of grief 190–200; active
100, 124–5 suppression 195; Adult Attachment
accumulation of symptoms, idiosyncratic Interview 194; anxious attachment and
complications and 284–5 chronic mourning 193–4; attachment
active suppression 195 anxiety 191; attachment anxiety,
acute grief: and CG, symptoms of 103, attachment reorganization and 196–7;
104–5; symptoms of 207 attachment avoidance 191; attachment
adaptive function 176 behavioral system 190; attachment
adjustment disorder (AD) related to bonding with therapist, therapeutic
bereavement 99–100; criteria for 94–5 change and 200; attachment insecurities
Adult Attachment Interview 194 and disordered patterns of grief 193–8;
affective phenomena, CG and 131, 132 attachment orientations 191; attachment
American Indians, mental health work perspective on grief resolution and
with 32–3 counseling 198–200; attachment
American Journal of Psychiatry 121 perspective on loss and grief 191–2;
American Psychiatric Association (APA) attachment security 198–9; attachment
35, 41, 99, 100, 101, 102, 108, 115, theory 190–1; avoidant attachment
116, 118, 180 197; avoidant attachment, absence of
anxious and depressive avoidance 224–5 grief and 194–6; chronic mourning
anxious attachment and chronic mourning 193; cognitive loading 195–6; despair
193–4 191–2; detachment 191–2; dispositional
anxious avoidance 224, 228; anxious attachment security 198–9; dual-
preoccupation with and avoidance process models 192; empirical
of memories, vacillation between evidence on links between attachment
55; elaboration of loss and negative insecurities and complicated grief
cognitions, interactions between and (CG) 196–8; internal working models
mediating roles of 225–6 190; methodological problems 197;
arousal phenomena 131, 132 negative beliefs, pervasive presence of
assessment: management and 136–9; 194; painful memories, effect of 193–4;
techniques for 286 pair-bond relationships 192; preemptive
attachment anxiety: attachment insecurities defenses 195; protest 191–2;
and disordered patterns of grief 191; relationship-specific avoidance 197–8;
attachment reorganization and 196–7 reorganization in bereavement process
attachment avoidance 191 192; responses to intense distress
attachment behavioral system 190 191–2; self-representations 195–6
attachment bonding with therapist, attachment security 198–9
therapeutic change and 200 attachment-specific stress response 206–7
 Subject index  321
attachment style, role in grieving 153 avoidant attachment: absence of grief and
attachment theory: attachment 194–6; attachment insecurities and
insecurities and disordered patterns disordered patterns of grief 197
of grief 190–1; family therapy for CG
249–50; physiological mechanisms and Bali, grieving people in 28–9
neurobiology of CG 205 behavioral syndrome or pattern in
autobiographical memory (AM), repetitive individuals, PGD and 86
thought (RT) and 168–9 belief factors 154
autobiographical memory (AM) processes bereavement: bereavement phenomena
in CG 176–86; adaptive function 133; bereavement related disorder
176; autobiographical memory (AM) 92–3, 95; engagement with bereaved
176–7; Autobiographical Memory Test families 252; improvement during
(AMT) 177, 179, 180, 181, 182, 183, 61–2; pathology and 131–3; report
184; Autobiographical Memory Test on state-of-the-art knowledge
(AMT), methodology and limitations about (Institute of Medicine, 1984)
of 178; Biographical Memory Test 56; traumatic stress reactions and
(BMT) 180, 181, 182; Biographical bereavement reactions, phenomena of
Memory Test (BMT) – Deceased 130–1
181–2, 184, 186; CaRFAX (capture and bereavement research: areas of focus for,
rumination, functional avoidance, and differences in 44; attachment theory
executive control) model 178; categoric in 205–7; classification of deaths
memories 178; clinical relevance of 280; clinical practice and research
OGM bias 184–6; directive function findings, inconsistencies between
176; encoding 176; executive control 44–5; community of 3–4; neurobiology
and OGM bias 179–80; extended and 209–11; physiological and
memories 178; functional avoidance neurological variables in 8; planning
and OGM bias 179; future research of interventions for CG 199–200; risk
186; Grieving Mindfully (Kumar, S.M.) factors of complicated grief (CG) in
185; immunity and OGM bias 180–3; 157; subtypes of CG patterns, influence
memory and affect, domain of 176–7; on 298
memory retrieval, impaired ability in Bioethics, President’s Council on 17
177; MEmory Specificity Training Biographical Memory Test (BMT) 180,
(MEST) 185; overgeneral memory 181, 182
(OGM) bias 177; overgeneralization of Biographical Memory Test (BMT) –
memories, bias toward 177; personal Deceased 181–2, 184, 186
goals and autobiographical memory, biological dysfunction, Boorse’s
relationship between 184; psychological perspective on 14, 15, 19
mechanisms in OGM bias in relation bodies and minds of humans 14
to CG research 178–84; recurrent brief group therapies for CG 263–75;
intrusive memories 180; retrieval 176; avoidance 264, 271; clinical trials
rumination and OGM bias 178–9; self 267–9; comparative trial 268;
and OGM bias, CG research studies composition trial 268–9; conditions
and 183–4; self memory system (SMS) created in 269; control trial 268;
179, 183; self-representative function effectiveness and lasting benefits 269–
176; social function 176; storage 176; 70; evidence-based treatment 273–4;
transdiagnostic process 177 follow-up types and patient outcomes
Autobiographical Memory Test (AMT) 270–1; future activities, suggestions
177, 179, 180, 181, 182, 183, 184; for 274–5; high-QOR interpretive
methodology and limitations of 178 groups 268–9, 275; Impact of Event
avoidance: brief group therapies for CG Scale (Horowitz) 264; interpersonal
264, 271; clarity regarding CG, lessons psychotherapy (IPT) 266, 271;
from clinical practice 48–9; phenomena interpretive and supportive therapies,
of 131, 132; repetitive thought (RT) as differences between 265; interpretive
169–70 group therapy 267–8; interpretive
322  Subject index
statements, illustration of 266; chronic mourning: attachment insecurities
interpretive therapy 265, 267; intrusion and disordered patterns of grief 193;
264, 271; low-QOR supportive groups uncomplicated and complicated
268–9, 275; mental status examination patterns of grief and 56
264; meta-analytical reviews and chronicity, evidence for 59–60
their critiques 271–3; mixed-QOR clarity regarding CG, lessons from clinical
groups 269; prevalence of CG 264–5; practice 40–51; avoidance 48–9;
short-term group therapies, treatment clinical observation and empirical
with 263–4; supportive statements, investigation, discrepancies between
illustration of 266–7; supportive 43–5; complicated grief, two functions
therapy 265, 267; task overload 263; of 48–9; complicated grief (CG),
technical manuals, feedback and 265–6; current professional understanding
therapeutic differences 265; Treatment of 40–1; complicated grief (CG) as
Induced Deterioration Effect (TIDE) clinical phenomenon 43–5; complicated
272; validation 273 grief (CG) as distinct diagnostic entity
41–3; complicated grief symptoms 46;
CaRFAX (capture and rumination, complicated grief syndromes 46–7;
functional avoidance, and executive complicated mourning, four forms
control) model 178 of 45–8; comprehensive conceptual
categoric memories 178 model, need for 50–1; connection 48–9;
cause of death: as predictor of bereavement death, manifestation of complicated
outcomes 64; subsequent grief and 151; grief in 48; diagnosable mental or
violent death, causes and complications physical disorder 47–8; DSM-5 40–1;
of 284 fallacy of PGD as “equivalent” to
Changing Lives of Older Couples (CLOC) complicated grief 42–3; “Grief and
Study: risk factors for complicated Bereavement: What Psychiatrists Need
grief (CG) 150, 152, 153, 154; to Know” (Zisook, S. and Shear, K.)
uncomplicated and complicated 42–3; grief and mourning, distinction
patterns of grief and 57, 58, 61–2, 63 between 40; operationalized definition,
children, CG in 68–77; challenges need for 49–50; PGD as a diagnostic
for children, unique nature of 77; entity for complicated grief 41–2;
childhood bereavement, consequences prolonged grief disorder (PGD) 41
of 68–70; childhood bereavement, classification 22–3
risk and protective factors 69–70; clinical complications, future research
cognitive–behavioral therapy (CBT) directions and 157
75; DSM-5 68; DSM-5, diagnosis clinical effects after violent death,
for adults 71; emotional tolerance, quantification of 278–9
loss and lack of 73; Extended Grief clinical observation and empirical
Inventory 74; healthy grieving 70–1; investigation, discrepancies between
Hogan Sibling Inventory of Grief 43–5
74; informational climate, loss and clinical relevance: CG in terms of 279,
poverty of 73; interventions for grief 283–4; of OGM bias 184–6
in children 74–7; measurement of grief clinical significance of distress or
74; mother, loss of 69–70; normal disability 86–7
and complicated grief in children clinical trials 138, 264, 267–9, 270, 271–2,
70–4; parental capacity, strengthening 273–4
of 75–6; pathalogical mourning, clinical utility 89
Bowlby’s concept of 71–2; positive cognitive–behavioral therapy (CBT):
parenting, importance of 76; traumatic children, CG in 75; prolonged grief
experiences and loss 72–3; Traumatic disorder (PGD) and 221, 226–8;
Grief Inventory for Children 74; written research on CBT theory and treatment
disclosure interventions following for PGD 229–31; theoretical basis
trauma 77 of CBT for PGD 222–6; theoretical
chronic grief: depression and, distinction underpinnings of CBT for PGD
between 63; distress and 62–4 229–30; as treatment for PGD 230–1
 Subject index  323
cognitive loading 195–6 differences in 308–9; processes,
cognitive phenomena, CG and 131, 132 mechanisms and 307–8; protective
cognitive processing, meaning factors 306; repetitive thought (RT)
construction and 250 and implications for 172; resilient
comorbidity: of CG and PTSD 129; of CG characteristics 306; as risk factor,
and trauma syndromes 136–9; issue of predictive validity argument 107–11;
119 risk factors, research on 305–7;
comparative trial 268, 272 scientific knowledge base regarding,
complicated grief (CG): ameliorative contribution of 4; significant impact of
interventions 309; as categorical 204; subgroup differences in 299–300;
pathology, distinctive symptom subtypes/variety of CG patterns 298–9;
argument 103; in children 68–77; symptoms of 46; syndromes 46–7;
clarity regarding, achievement of treatment 308–10; variables, definition
40–51; as clinical phenomenon 43–5; of 306; see also concept of CG, lessons
comparison with other disorders from other cultures; risk factors for
304–5; complicated and uncomplicated complicated grief (CG)
patterns of grief, nature and prevalence complicated grief (CG) as a new DSM
of 55–65; concept of 27–37; mental disorder, lessons from PTSD
conceptual/theoretical approaches to for 115–26; comorbidity, issue of 119;
300–1; contemporary concern with complicated grief (CG), explorations
3–4; in context of other psychiatric of DSM-5 and 123–6; critical voices
disorders 129–39; criteria for 91–4; on the mental disorder PTSD 117;
cultural critique of concept of 31–3; dose–response relationship between
current DSM status, criteria and traumatic stressors and symptoms
problems 302–4; current professional 118; DSM, pragmatic system
understanding of 40–41; definition with drawbacks 121–3; effective
296; definition as mental disorder interventions, growth of 123–4;
302; as derailed grief, interminability medicalization, danger of 125;
argument 104–7; falsification of normal and abnormal adaptation to
105–7; description and definition of 3; trauma, blurring distinction between
diagnostic category of, (dis)advantages 125; overemphasis of complaints
of 304; as disorder, diagnostic 119–20; Post-Traumatic Stress Disorder
categorization and 301–5; disorder (PTSD), introduction as diagnostic
or not? 13–21; distinct and disabling category in DSM-III 115; PTSD
condition 85–6; as distinct diagnostic field, growth of 126; PTSD in DSM,
entity 41–3; distinct disorder? 21–3; historical note on 116–17; PTSD in
as distinct risk 145–6; distinctive DSM, research on 115; self-fulfilling
features of, violent death and 282; prophecy, introduction of disorder
efficacy of intervention for 309–10; as 124–5; simulation of complaints
explorations of DSM-5 and 123–6; 119–20; society and trauma/PTSD
fallacy of PGD as “equivalent” to 42–3; 120; stressor criterion 117–18, 125–6;
functions of 48–9; fundamental issues, stressor criterion, changes in 118–19;
differences of opinion about 4; grief- suffering, acknowledgment of 123
specific or generic risk factors 306–7; complicated grief (CG) in context of
interdisciplinary approach to study of, other psychiatric disorders 129–39;
need for 310–11; nature of 295–301; affective phenomena 131, 132; arousal
as a new DSM mental disorder, lessons phenomena 131, 132; assessment
from PTSD for 115–26; normal grief, and management 136–9; avoidance
CG compared with 296–8; paradigms phenomena 131, 132; bereavement
adopted in interventions, differences and pathology 131–3; bereavement
in 308–9; philosophical perspectives phenomena 133; cognitive phenomena
13–24; prevalence of 300; preventive 131, 132; comorbidity of complicated
interventions 309; principles adopted grief (CG) and PTSD 129; comorbidity
in interventions, differences in 308–9; of complicated grief (CG) and trauma
procedures adopted in interventions, syndromes 136–9; “pathologies of
324  Subject index
grief,” consideration of possibility lessons from clinical practice 48–9;
of 129; post-traumatic phenomena rational connection 17
133; prolonged grief disorder (PGD) conscious grieving, prolonged absence of
129; stressor exposures, PTSD 56
and bereavement 133–6; trauma constructive effects of RT 166–7;
syndromes, complicated grief (CG) constructive versus unconstructive
and 129; traumatic stress reactions and effects of RT, accounting for 167–71
bereavement reactions, phenomena of control theory account of RT 167–8
130–1 control trial 268, 272
complicated mourning, forms of 45–8 core body temperature 212
composition trial 268–9, 272 crime-related matters, violent death and
comprehensive conceptual model, need dealing with 278
for 50–1 cross-cultural variation in ideas about
computerized interventions for depression deviant grieving 28–9
235 cross-cultural variations in human
concept of CG, lessons from other cultures psychology 27
27–37; American Indians, mental crying, neurobiology of CG and 208
health work with 32–3; Bali, grieving cultural and situational insensitivity in
people in 28–9; clients from many mental health work 32–3
different cultures 33–5; cross-cultural cultural change and medical education
variation in ideas about deviant 34–5
grieving 28–9; cross-cultural variations cultural critique of concept of CG 31–3
in human psychology 27; cultural cultural formulation of CG 35–7
and situational insensitivity in mental cultural meanings, psychiatric diagnosis
health work 32–3; cultural critique and 33
of concept of CG 31–3; cultural cultural saturation of evidence 33
formulation 35–7; cultural saturation
of evidence 33; DSM-IV, cultural death: death- and bereavement-related
formulation and 35–6; DSM-V, cultural factors in CG 151–2; definition of
formulation and 36–7; dual-process violent death 279–80; manifestation
model 36; grief arises from discrete, of complicated grief in 48; peri-death
time-limited loss, assumption of 29–30; variables 151–2; see also causes of
grieving in all cultures is rather the death
same, assumption that 30–1; human delayed grief, evidence for 59
universals, assumptions of 33–4; Ifaluk depressive rumination 162
of Micronesia 28; individual is what despair 191–2
needs healing, assumption that 31, 32; detachment 191–2
Maisin people of Papua New Guinea deviant grieving, cross-cultural variation in
29; Major Depressive Disorder 36; ideas about 28–9
medical education, cultural change Diagnostic and Statistical Manual for
and 34–5; mourning for indigenous Mental Disorders (DSM) 85; argument
Australians 34; mourning in Poland that CG satisfies definition of mental
30–1; non-problematic symptoms disorder 102–3; pragmatic system with
of CG 28; problematic assumptions drawbacks 121–3; see also DSM-III;
underlying concept of CG 29–31; DSM-IV; DSM-V
psychiatric diagnosis and cultural diagnostic criteria, need for stringency in
meanings 33; recovery, Western 111–12
concept of 31; therapeutic treatment for diagnostic validators and clinical
bereavement importance of 35; Toraja utility, differentiation from “nearest
of Indonesia 28, 29; Western medicine, neighbors” with 89
influence of 34–5 diagnostic validity 88
connection: chronic grief and a directive function 176
dysfunctional worldview, connection discussions: family therapy for CG 258–9;
between 154; clarity regarding CG, prolonged grief disorder (PGD),
 Subject index  325
diagnostic category in DSM-5 95–7; empirical evidence on links between
review of risk factors for CG 155–7 attachment insecurities and CG 196–8
disorder: definitions of 13, 19; impairment encoding 176
as criterion for 101–2; as statistical engagement with bereaved families 252
deviance 102; Wakefield’s perspective ethical aspects of family-centered care 252
on 14–15, 16 Evidence-Based Psychotherapy
disorder conditions: descriptivist account (Goodheart, C.D., Kazdin, A.E. and
of 14; diagnosable mental or physical Sternberg, R.J.) 273
disorder 47–8; intuition of badness of evidence-based treatment for CG 273–4
disorders 14 Evidence-Based Treatment of Personality
dispositional attachment security 198–9 Dysfunction (Magnavita, J.J., ed.) 273
distress: chronic grief and 62–4; evolutionary adaptive disorder 16
moderation after violent death of 287; evolutionary dysfunction 15–16
responses to intense distress 191–2; executive control and OGM bias 179–80
separation distress 55; separation Extended Grief Inventory 74
distress, response to 214 extended memories 178
dopamine 206
dose–response relationship between Facebook 237
traumatic stressors and symptoms 118 family therapy for CG 248–59;
DSM-III: philosophical perspectives 19 attachment theory 249–50; challenges
DSM-IV: cultural formulation and 35–7; identified by the therapists 257–8;
DSM-IV-TR (APA, 2000) 85–6; cognitive processing and meaning
philosophical perspectives 19–20 construction 250; discussion 258–9;
DSM-V: children, CG in 68; clarity engagement with bereaved families
regarding CG, lessons from clinical 252; ethical aspects of family-centered
practice 40–1; cultural formulation care 252; family assessment and
and 36–7; diagnosis for adults 71; therapy, approach to 253–4; family
mental disorder, definition of 101–3, communication, improvement in 254–5;
108; potential benefits of adding new family therapy, reasons for 248–9;
syndrome should outweigh potential family therapy and CG 254–8; goals of
harms 89–90; proposals for 19, 99–100; family therapy 253; group adaptation
proposals of criteria for 90–5; see 250; harnessing family resilience
also prolonged grief disorder (PGD), 250–1; loss and family 248; promise in
diagnostic category in DSM-5 ameliorating and preventing depression
dual-process models: attachment 255–6; recognition of families in need
insecurities and disordered patterns of during palliative care 251–2; theoretical
grief 192; concept of CG, lessons from models guiding family therapy 249–51;
other cultures 36 therapy at home, conduct of 252–3;
duration threshold 104 therapy processes 256–7
functional avoidance and OGM bias 179
effectiveness: of bereavement functional impairment and quality of life
interventions 237–8; effective 109–10
interventions for PTSD, growth of functional neuroimaging 209–11
123–4; effects of psychotherapeutic future directions: brief group therapies for
interventions after violent death 285–8; CG 274–5; physiological mechanisms
efficacy of intervention for bereaved and neurobiology of CG 214–15
persons with complications in their
grieving, study of 288; and lasting gender, role in grieving 149–50
benefits of brief group therapies general stress response 205–6
269–70 Generalized Anxiety Disorder (GAD) 107,
elaboration of loss, anxious avoidance 125
and negative cognitions, interactions genetics 212–13
between and mediating roles of 225–6 grief: arising from discrete, time-limited
emotional tolerance, loss and lack of 73 loss, assumption of 29–30; grief
326  Subject index
trajectories 57–9; grief trajectories, informational climate, loss and poverty
predictors of 60–4; grief work as of 73
rumination hypothesis 163; love and Institute of Medicine 56
16–18; meaning, grief and search interdisciplinary approach, need for
for 153; measurement of grief in 310–11
children 74; as moral requirement internal working models 190
17; morbid patterns of 56; mourning Internet-based bereavement interventions
and, distinction between 40; natural and support 235–44; computerized
response of 145; “normal grief,” interventions for depression 235;
pattern of 64; prevalence and variety of effectiveness of bereavement
grief patterns 64; reactions to, earliest interventions 237–8; Facebook
taxonomies of individual differences 237; imaginative confrontation
in 55–6; RT and, evidence on 164–7; 239–40; Internet-based bereavement
RT and, theoretical perspective 163–4; interventions 236, 238–40, 243–4;
short-lived grief reactions 56–7; social Internet-based communication, new
support and 154; symptoms of 15; possibilities for 243; Internet-based
symptoms of, measurement of 283–4; discussion forums 235; Internet-based
vulnerability to 17–18; worldview, intervention for CG 238–40, 243–4;
grief and importance of 154; see randomized controlled trial results 240;
also uncomplicated and complicated Internet-based prevention program
patterns of grief 240; Internet-based therapeutic
“Grief and Bereavement: What interventions, advantages and
Psychiatrists Need to Know” (Zisook, disadvantages of 242–3; Internet-based
S. and Shear, K.) 42–3 treatment for parents after prenatal
grieving: in all cultures is rather the same, loss 240–1; meta-analyses and reviews
assumption that 30–1; healthy grieving of bereavement treatments 237–8;
70–1; for a loved one 16–17; marital mourning and loss support groups
dependency, role in 152; normal 236–7; online bereavement support
and pathological forms of grieving, groups 236–7; psychotherapeutic
comparisons between 56; race, role in interventions, development of new
150 forms of 235; self-help interventions
Grieving Mindfully (Kumar, S.M.) 185 for bereaved individuals 241–2; social
group adaptation 250 media, development of 235; Swiss
Bereaved Parents Study 236–7
happiness, idealization of 18 interpersonal factors in CG 154
harmfulness 14–15 interpersonal psychotherapy (IPT) 266,
health care professionals, roles of 64–5 271
healthy grieving 70–1 interpersonal regulators, withdrawal of 214
high-QOR interpretive groups 268–9, 275 interpretive and supportive therapies,
Hogan Sibling Inventory of Grief 74 differences between 265
homosexuality 14 interpretive group therapy 267–8
human psychology, cross-cultural interpretive statements, illustration of 266
variations in 27 interpretive therapy 265, 267
human universals, assumptions of 33–4 interventions: after violent death, aims
hypertension and heart rate 208–9 of 287–8; ameliorative interventions
hypothalamic-pituitary-adrenal (HPA) axis 309; computerized interventions
206 for depression 235; effectiveness of
bereavement interventions 237–8;
Ifaluk of Micronesia 28 for grief in children 74–7; paradigms
imaginative confrontation 239–40 adopted in interventions, differences
immunity and OGM bias 180–3 in 308–9; planning of interventions
Impact of Event Scale (Horowitz) 264 for CG 199–200; postsuicide
impairment as criterion for disorder 101–2 intervention programs 286; preventive
individual is what needs healing, interventions 309; principles adopted
assumption that 31, 32 in interventions, differences in 308–9;
 Subject index  327
procedures adopted in interventions, mental disorder: “mental/psychiatric
differences in 308–9; psychotherapeutic disorder,” definitional specificity and
intervention following violent death 89; philosophical perspectives 20–1;
285–8; psychotherapeutic interventions, prolonged intense grief, classification
development of new forms of 235; of 109; as Roschian concept 20
written disclosure interventions mental health work, cultural and
following trauma 77; see also Internet- situational insensitivity in 32–3
based bereavement interventions and mental status examination 264
support meta-analyses: brief group therapies for
intrapersonal factors in CG 153–4 CG, reviews and their critiques 271–3;
intrusion 264, 271 and reviews of bereavement treatments
intuition of badness of disorders 14 237–8
Inventory of Complicated Grief (ICG) 85 mixed-QOR groups 269
models: and theories of CG 207–8; and
kinship, role in grieving 152 theories of general bereavement 207
monoamine oxidase-A (MAO-A) 212–13
literature: review of empirical literature on morbid patterns of grief 56
risk factors for complicated grief (CG) mortality 68, 69–70, 109, 131, 166, 208,
145–58; scientific literature on violent 305
death, assessment of 281–4, 288; mother, loss of 69–70
studies of physiology and neurobiology mourning: by indigenous Australians 34;
of CG 208–13 and loss support groups 236–7; in
loss: elaboration and integration of 222–3; Poland 30–1
family and 248
The Loss of Sadness: How Psychiatry narrative therapy 287
Transformed Normal Sorrow into National Centers for Disease Control (US)
Depressive Disorder (Horwitz, A. and 278
Wakefield, J.) 15 negative beliefs, pervasive presence of 194
love and grief 16–18 negative cognitions: bereavement-
low-QOR supportive groups 268–9, 275 associated 153–4; elaboration of loss
and anxious avoidance, interactions
Maisin people of Papua New Guinea 29 between and mediating roles of 225–6
major depressive disorder (MDD) 21, 22, neurobiology 209–11
107, 134–5; concept of CG, lessons neuroticism, link with grief 153
from other cultures 36; philosophical non-problematic symptoms of CG 28
perspectives 19 normal and abnormal adaptation to trauma,
maladaptive abstract RT 171 blurring distinction between 125
maladaptive cognitions and assumptions, normal and complicated grief: in children
PGD and 227 70–4; determination of 18
marital dependency, role in grieving 152 normal and pathological forms of grieving,
maternal regulators, removal of 205 comparisons between 56
meaning, grief and search for 153 “normal grief,” pattern of 64
meaningful re-engagement after violent nucleus accumbens 210–11
death 288
measurement of grief in children 74 online bereavement support groups 236–7
medical education, cultural change and operationalized definition for CG, need for
34–5 49–50
medicalization of CG, danger of 125 opoids 206
memory and affect, domain of 176–7 overemphasis of complaints 119–20
memory processes, negative cognitions overgeneral memory (OGM) bias 177
and avoidance behaviors in oxytocin 206–7
development and maintenance of
222–6, 231 painful memories, effect of 193–4
memory retrieval, impaired ability in 177 pair-bond relationships 192
MEmory Specificity Training (MEST) 185 parental capacity, strengthening of 75–6
328  Subject index
pathological mourning, Bowlby’s concept response 205–6; genetics 212–13;
of 71–2 hypertension and heart rate 208–9;
“pathologies of grief,” consideration of hypothalamic-pituitary-adrenal (HPA)
possibility of 129 axis 206; interpersonal regulators,
peri-death variables 151–2 withdrawal of 214; maternal
persistent negative thinking 223–4 regulators, removal of 205; models
personal goals and AM, relationship and theories of CG 207–8; models
between 184 and theories of general bereavement
philosophical perspectives 13–24; 207; monoamine oxidase-A (MAO-
abnormality 16; biological dysfunction, A) 212–13; neurobiology 209–11;
Boorse’s perspective on 14, 15, 19; nucleus accumbens 210–11; opoids
bodies and minds of humans 14; 206; oxytocin 206–7; physiological
classification 22–3; complicated co-regulation 205; physiological stress
grief (CG), disorder or not? 13–21; responses 207–8; physiology and
complicated grief (CG), distinct neurobiology of CG, reasons for studies
disorder? 21–3; descriptivist account of 213; rapid eye movement (REM)
of disorder conditions 14; disorder, sleep 211–12; separation distress
definitions of 13, 19; disorder, response 214; sleep 211–12; studies of
Wakefield’s perspective on 14–15, physiology and neurobiology of CG
16; DSM-III 19; DSM-IV 19–20; 208–13
DSM-V, proposals for 19; evolutionary positive parenting, importance of 76
adaptive disorder 16; evolutionary Post-Traumatic Stress Disorder (PTSD)
dysfunction 15–16; grief as moral 6–7, 69, 71, 72, 76, 85, 88, 91, 107,
requirement 17; grieving for a loved 165–6, 204, 221, 237, 296; adaptation
one 16–17; happiness, idealization of to trauma 125; autobiographical
18; harmfulness 14–15; homosexuality memory (AM) and 177, 180, 181, 183;
14; intuition of badness of disorders 14; comorbidity of complicated grief (CG)
The Loss of Sadness: How Psychiatry and 129; complicated grief (CG) in
Transformed Normal Sorrow into context of 129–39; critical voices on
Depressive Disorder (Horwitz, A. 117; disorder as self-fulfilling prophecy
and Wakefield, J.) 15; love and grief 124–5; growth of field 126; historical
16–18; Major Depressive Disorder note on PTSD in DSM 116–17;
19; mental disorder 20–1; normal and introduction as diagnostic category in
complicated grief, determination of 18; DSM-III 115; research on PTSD in
President’s Council on Bioethics 17; DSM 115; society and 120; stressor
promiscuous realism, Dupré and 22–3; exposures, PTSD and bereavement
proportionality 18; psychic injury, 133–6
concept of 20; risk of grief, exposure to postsuicide intervention programs 286
18; Roschian concept, mental disorder posttraumatic phenomena 133
as 20; sadness 15; symptoms of grief predictive validity 107–9
15; vulnerability to grief 17–18; What preemptive defenses 195
Is Mental Disorder? (Bolton, D.) 20–1 President’s Council on Bioethics 17
physical disorders 109 prevalence of CG: brief group therapies
physiological mechanisms and for CG 264–5; violent death, CG after
neurobiology of CG 204–15; acute 281–2
grief, symptoms of 207; attachment problem-focused and emotion-focused
figure, role of 205, 206–7; attachment- support 286
specific stress response 206–7; prolonged grief disorder (PGD) 221–32;
attachment theory 205; complicated anxious and depressive avoidance
grief (CG), significant impact of 224–5; anxious avoidance, elaboration
204; core body temperature 212; of loss and negative cognitions,
crying 208; dopamine 206; functional interactions between and mediating
neuroimaging 209–11; future roles of 225–6; CBT as treatment
directions 214–15; general stress for 230–1; changing maladaptive
 Subject index  329
cognitions and assumptions 227; concept 89; PGD/CG as formal
clarity regarding CG, lessons from disorder? 86–90; potential benefits of
clinical practice 41; cognitive– adding new syndrome to DSM-5 should
behavioral therapy (CBT) 221, 226–8; outweigh potential harms 89–90;
complicated grief (CG) in context of prolonged grief disorder (PGD), criteria
other psychiatric disorders 129; as a for 90–1, 92–3; proposals of criteria
diagnostic entity for complicated grief for DSM-5 90–5; psychobiological
41–2; elaboration of loss, anxious dysfunction 87–8; social deviance or
avoidance and negative cognitions, conflicts with society, no involvement
interactions between and mediating of 88
roles of 225–6; fallacy of PGD as prolonged intense grief, classification of
“equivalent” to complicated grief 99–112; acute grief and CG, symptoms
42–3; as formal disorder? 86–90; of 103, 104–5; adjustment disorder
further research work, need for (AD) related to bereavement 99–100;
231–2; insufficient elaboration and argument that CG satisfies DSM’s
integration of loss 222–3; memory definition of mental disorder 102–3;
processes, negative cognitions and Bowlby, concern for misconstrual
avoidance behaviors in development of potential severity and duration
and maintenance of 222–6, 231; of normal grief 100–1; complicated
negative cognitions, elaboration of loss grief (CG) as categorical pathology,
and anxious avoidance, interactions distinctive symptom argument 103;
between and mediating roles of complicated grief (CG) as derailed
225–6; persistent negative thinking grief, interminability argument 104–7;
223–4; promotion of elaboration and falsification of 105–7; complicated
integration of loss 226–7; recognition grief (CG) as risk factor, predictive
as distinct disorder 221; reducing validity argument 107–11; conceptual
anxious and depressive avoidance 228; problems with predictive validity
research on CBT theory and treatment argument 107–9; diagnostic criteria,
for PGD 229–31; theoretical basis need for stringency in 111–12; disorder
of CBT for PGD 222–6; theoretical as statistical deviance 102; DSM-5
underpinnings of CBT for PGD 229–30 definition of mental disorder 101–3,
prolonged grief disorder (PGD), 108; DSM-5 proposals 99–100;
diagnostic category in DSM-5 85–97; duration threshold 104; functional
adjustment disorder (AD) related impairment and quality of life 109–10;
to bereavement, criteria for 94–5; impairment as criterion for disorder
behavioral or psychological syndrome 101–2; mental disorders of other kinds
or pattern occurring in individuals 109; mortality 109; physical disorders
86; bereavement related disorder 109; predictive validity 107–9; quality
92–3, 95; clinical utility 89; clinically of life 110; suicidal ideation 110–11
significant distress or disability as promiscuous realism, Dupré and 22–3
consequence of syndrome 86–7; proportionality 18
complicated grief (CG), criteria protest, attachment insecurities and 191–2
for 91–4; complicated grief (CG), psychiatric diagnosis, cultural meanings
distinct and disabling condition 85–6; and 33
Diagnostic and Statistical Manual psychic injury, concept of 20
for Mental Disorders (DSM) 85; psychobiological dysfunction 87–8
diagnostic validators and clinical psychological mechanisms in OGM bias in
utility, differentiation from “nearest relation to CG research 178–84
neighbors” with 89; diagnostic validity psychological syndrome or pattern in
88; discussion 95–7; DSM-IV-TR individuals, PGD and 86
(APA, 2000) 85–6; Inventory of psychotherapeutic intervention following
Complicated Grief (ICG) 85; “mental/ violent death 285–8
psychiatric disorder,” no definition psychotherapeutic interventions,
perfectly specifies boundaries for development of new forms of 235
330  Subject index
quality of life: functional impairment clinical complications and future
and 109–10; prolonged intense grief, research directions 157; death- and
classification of 110 bereavement-related factors 151–2;
discussion on review 155–7; distinct
race, role in grieving 150 risk, complicated grief (CG) as 145–6;
rapid eye movement (REM) sleep 211–12 gender, role in grieving 149–50; grief,
reconstructive processing after violent natural response of 145; initial analyses
death 287 147–8; interpersonal factors 154;
recovery, Western concept of 31 intrapersonal factors 153–4; kinship,
recurrent intrusive memories 180 role in grieving 152; limitations on
relationship-specific avoidance 197–8 review 156–7; marital dependency,
religion, grief and importance of 154 role in grieving 152; meaning,
reorganization in bereavement process 192 grief and search for 153; negative
repetitive thought (RT) 162–72; cognitions, bereavement-associated
autobiographical memoryand 153–4; neuroticism, link with grief
RT 168–9; as avoidance 169–70; 153; peri-death variables 151–2; race,
complicated grief (CG), implications role in grieving 150; relationship to
for 172; constructive effects of deceased 152; religious factors 154;
RT 166–7; constructive versus results 147–54; review description
unconstructive effects of RT, 146–7; review of empirical literature
accounting for 167–71; control theory 145–58; review procedure 147–54; risk
account of RT 167–8; depressive factors for common and complicated
rumination 162; grief and, evidence grief (CG) 149; sense making, grief and
on 164–7; grief and, theoretical search for 153; social support and grief
perspective 163–4; grief work as 154; spiritual life, grief and importance
rumination hypothesis 163; hypotheses of 154; subsequent analyses 148;
and relevant evidence, summary of survivor’s background 149–50; violent
170–1; maladaptive abstract RT 171; death 282–3; worldview, grief and
moderators of consequences of RT importance of 154
167; treatment implications 171–2; Roschian concept, mental disorder as 20
unconstructive effects of RT following rumination and OGM bias 178–9
bereavement 164–5
research: on CBT theory and treatment for sadness 15
PGD 229–31; further research work on scientific literature on violent death,
PGD, need for 231–2; future research assessment of 281–4, 288
on autobiographical memory (AM) self and OGM bias, CG research studies
processes in CG 186; psychological and 183–4
mechanisms in OGM bias in relation self-fulfilling prophecy, introduction of
to CG research 178–84; on PTSD in disorder as 124–5
DSM 115; risk factors for CG 305–7; self-help interventions for bereaved
on violent death as risk factor for CG, individuals 241–2
need for further study 283, 288–9; see self memory system (SMS) 179, 183
also bereavement research self-representations 195–6; self-
resilience: harnessing family resilience representative function 176
250–1; uncomplicated and complicated sense making, grief and search for 153
patterns of grief, nature and prevalence separation distress 55; response to 214
of 61 short-lived grief reactions 56–7
retrieval of memory 176 short-term group therapies, treatment with
risk factors for complicated grief (CG): 263–4
attachment style, role in grieving simulation of complaints 119–20
153; belief factors 154; cause of sleep 211–12
death, subsequent grief and 151; social deviance or conflicts with society,
Changing Lives of Older Couples no involvement of 88
(CLOC) Study 150, 152, 153, 154; social function, AM processes and 176
 Subject index  331
social media, development of 235 and avoidance of memories, vacillation
social support and grief 154 between 55; bereavement, report
spiritual life, grief and importance of 154 on state-of-the-art knowledge about
storage of memory 176 (Institute of Medicine, 1984) 56; cause
stressor criterion 117–18, 125–6; changes of death as predictor of bereavement
in 118–19 outcomes 64; Changing Lives of Older
stressor exposures, PTSD and bereavement Couples (CLOC) Study 57, 58, 61–2,
133–6 63; chronic grief and depression,
studies of physiology and neurobiology of distinction between 63; chronic
CG 208–13 grief and distress 62–4; chronic
suffering, acknowledgment of 123 mourning 56; chronicity, evidence
suicidal ideation 110–11 for 59–60; conscious grieving,
supportive statements, illustration of 266–7 prolonged absence of 56; delayed
supportive therapy 265, 267 grief, evidence for 59; distress,
survivor background 149–50 chronic grief and 62–4; grief reaction,
Swiss Bereaved Parents Study 236–7 earliest taxonomies of individual
symptoms of grief 15 differences in 55–6; grief trajectories
57–9; grief trajectories, predictors
task overload 263 of 60–4; health care professionals,
technical manuals, feedback and 265–6 roles of 64–5; improvement during
theory: basis of CBT for PGD 222–6; bereavement 61–2; morbid patterns
guiding family therapy 249–51; of grief 56; normal and pathological
underpinning CBT for PGD 229–30; forms of grieving, comparisons
see also attachment theory between 56; “normal grief,” pattern
therapy: attachment bonding with of 64; prevalence and variety of grief
therapist, therapeutic change and 200; patterns 64; resilience 61; separation
at home, conduct of 252–3; interpretive distress 55; short-lived grief reactions
and supportive therapies, differences 56–7
between 265; interpretive group
therapy 267–8; interpretive therapy validation 90, 108; brief group therapies
265, 267; narrative therapy 287; for CG 273; empirical validation 95,
processes in family therapy for CG 302–3
256–7; psychotherapeutic intervention violent death, CG after 278–89;
following violent death 285–8; short- accumulation of symptoms,
term group therapies, treatment with idiosyncratic complications or
263–4; supportive therapy 265, 267; 284–5; assessment, techniques for
therapeutic differences 265; therapeutic 286; causes and complications 284;
treatment for bereavement importance clinical effects, quantification of
of 35; see also brief group therapies 278–9; clinical relevance, CG in
for CG; cognitive–behavioral therapy terms of 279, 283–4; conceptual
(CBT); family therapy for CG issues 284–5; crime-related matters,
Toraja of Indonesia 28, 29 dealing with 278; definition of violent
transdiagnostic process 177 death 279–80; distinctive features
trauma syndromes, CG and 129 of CG 282; distress, moderation of
traumatic experiences and loss 72–3 287; effects of psychotherapeutic
Traumatic Grief Inventory for Children 74 interventions 285–8; efficacy of
traumatic stress reactions and bereavement intervention for bereaved persons with
reactions, phenomena of 130–1 complications in their grieving, study
Treatment Induced Deterioration Effect of 288; grief symptoms, measurement
(TIDE) 272 of 283–4; intervention, aims of
287–8; meaningful reengagement
uncomplicated and complicated patterns 288; narrative therapy 287; patterns
of grief, nature and prevalence of of complication 284; postsuicide
55–65; anxious preoccupation with intervention programs 286; prevalence
332  Subject index
of CG 281–2; problem-focused non-violent deaths, distinction between
and emotion-focused support 286; 280; violent deaths, trauma of 278
psychotherapeutic intervention vulnerability to grief 17–18
following violent death 285–8;
reconstructive processing 287; research Western medicine: influence of 34–5;
on violent death as risk factor for CG, recovery, concept of 31
need for further study 283, 288–9; risk What Is Mental Disorder? (Bolton, D.)
factors 282–3; scientific literature on, 20–1
assessment of 281–4, 288; violent and worldview, grief and importance of 154