J Nutr Health Aging

ASSESSING THE CURRENT STATE OF COGNITIVE FRAILTY:

MEASUREMENT PROPERTIES
L. SARGENT1, R. BROWN2
1. Ph.D. Candidate at Medical University of South Carolina, Faculty of Virginia Commonwealth University, School of Nursing, Richmond, VA, USA; 2. Assistant Professor Research &
Education Librarian, Virginia Commonwealth University School of Nursing Affiliate Faculty, Tompkins-McCaw Library for the Health Sciences, Richmond, VA, USA. Corresponding
author: L. Sargent, Candidate at Medical University of South Carolina, Faculty of Virginia Commonwealth University, School of Nursing, Richmond, VA, USA, [email protected].

Abstract: Background: Currently, an estimated 25-30% of people ages 85 or older have dementia, with a
projected 115 million people worldwide living with dementia by 2050. With this worldwide phenomenon fast
approaching, early detection of at-risk older adults and development of interventions focused on preventing
loss in quality of life are increasingly important. A new construct defined by the International Consensus Group
(I.A.N.A/I.A.G.G) as «cognitive frailty» combines domains of physical frailty with cognitive impairment and
provides a framework for research that may provide a means to identify individuals with cognitive impairment
caused by nonneurodegenerative conditions. Using the integrative review method of Whittemore and Knafl.,
2005 this study examines and appraises the optimal measures for detecting cognitive frailty in clinical
populations of older adults. Methods: The integrative review was conducted using PubMed, CINAHL, Web of
Science, PsycInfo, and ProQuest Dissertations & Theses. From the total 185 articles retrieved, review of titles
and key words were conducted. Following the initial review, 168 articles did not meet the inclusion criteria for
association of frailty and cognition. Of the 18 fulltext articles reviewed, 11 articles met the inclusion criteria;
these articles were reviewed in-depth to determine validity and reliability of the cognitive frailty measures.
Results: Predictive validity was established by the studies reviewed in four main areas: frailty and type of
dementia MCI (OR 7.4, 95% CI 4.2-13.2), vascular dementia (OR 6.7, 95% CI 1.6-27.4) and Alzheimer’s
dementia (OR 3.2, 95% CI 1.7-6.2), frailty and vascular dementia (VaAD) is further supported by the rate of
change in frailty x macroinfarcts (r = 0.032, p < 0.001); frailty and the individual domains of cognitive function
established with the relationship of neurocognitive speed and change in cognition using regression coefficients;
individual components of frailty and individual domains of cognitive function associations inculded slow gait
and executive function (β -0.20, p < 0.008 ), attention (β -0.25 p < 0.008), processing speed (β -0.16, p < 0.008),
word recall (β - 0.18, p = 0.02), and logical memory (β = 0.04, p =0.04). Weak grip was predictive for changes
in executive function (β - 0.16, p =0.008). Physical activity was associated with changes in executive function
(β = -0.18, p= 0.02) and word recall (β = 0.17, p= 0.02), individual components of frailty and global cognitive
function were found in several studies which included grip strength (r = - 0.51, p < 0.001), gait speed (r = - 0.067,
p < 0.001), and exhaustion (β - 0.18, p < 0.008). Conclusions: This paper presents the first-known review of
the measurement properties for the cognitive frailty construct since the published results from the International
Consensus Group (I.A.N.A/I.A.G.G). Evidence presented in this review continues to support the link between
physical frailty and cognition with developing validity to support distinct relationships between components of
physical frailty and cognitive decline. Results call attention to inconsistencies in reporting of reliability, validity,
and heterogeneity in the measurements and operational definition for cognitive frailty. Further research is needed
to establish an operational definition and develop psychometrically appropriate clinical measures to construct an
understanding of the relationship between physical frailty and cognitive decline.

Key words: Cognitive decline, physical frailty, measurements, cognitive frailty

Introduction impairment (2–4).

The relationship between physical frailty and cognitive
With the number of individuals ages 80 and older on the impairment has become increasingly more apparent with recent
rise, the burden of dementia is expected to be one of the most studies suggesting that the two are interrelated. Efforts focused
daunting and costly consequences of longer life expectancies. on understanding the relationship may provide a means to
Early detection of at-risk older adults and the development of identify individuals with cognitive impairment caused by non-
interventions focused on preventing loss in quality of life are neurodegenerative conditions which might be reversible (2,
increasingly more important. Diagnosing dementia, especially 3). Although, frailty and cognitive impairment have been
in the early stages of the disease is difficult; many cases go shown to be related, both constructs have long been studied
undiagnosed even in the intermediate or more advanced stages separately (3). To address this gap, the International Consensus
(1). This is partly because dementia is a complex condition Group organized by the International Academy on Nutrition
that cannot be attributed to a single functional or cognitive and Aging (I.A.N.A) and the International Association of
domain and the need to better understand the underlying Gerontology and Geriatrics (I.A.G.G) convened on April 16th,
neuropathology contributing to non-aging related cognitive 2013 in an effort to identify domains of physical frailty and
Received September 28, 2015
Accepted for publication November 30, 2015
1
J Nutr Health Aging

ASSESSING THE CURRENT STATE OF COGNITIVE FRAILTY: MEASUREMENT PROPERTIES

cognition. Additionally, the consensus group recommended The criteria used to identify frailty often depend on the
formal assessments based on studies that supported findings operational definition. The commonly-known criterion is the
of an association between progressive physical frailty and “phenotypic” definition developed by the work completed
cognitive impairment in older adults. The new construct called in the Cardiovascular Health Study (CHS) (5, 11). The CHS
cognitive frailty (3), extends the physical frailty construct with phenotype includes decline in lean body mass, strength,
a formal cognitive assessment and a comprehensive assessment endurance, balance, walking performance, and low activity
of depressive symptoms. (5). It allows for a continuous scoring system versus a nominal
The construct cognitive frailty, will provide new system because it can capture the multidimentional nature of
opportunities for research, assist in further defining cognitive frailty. The components have concurrent and predictive validity
impairment related to physical causes, and promote with hazard ratios (HR) ranging from 1.82-4.46 (p < 0.05) for
interventions that lead to improved quality of life in older outcomes that include incident disease, hospitalization, falls,
adults. Multiple studies have been conducted to develop disability and mortality in community-dwelling older adults (5).
clinical screening tools for the detection of cognitive and Additionally, the CHS model has positive predictive validity
functional decline independently, with many clinical screening (PPV) in detection of physical limitations. The Edmonton Frail
instruments available to clinicians. However, the optimal Scale (EFS) includes evaluation of the social support domain
measures or combination of measures to accurately detect and has been validated with non-specialists with no formal
cognitive frailty in the clinical setting is unclear (3). As training in geriatric care (12). Construct validity for the EFS for
researchers attempt to deconstruct the relationship between detection of physical performance was statistically significant
physical frailty and cognitive impairment, the emphasis must be (r= - 0.58, p = 0.006, n=21) along with inter-rater reliability
placed on evaluating the strength of the psychometric tests used (k = 0.77. p = 0.0001) and internal consistency (Cronbach α
to evaluate the new construct. The purpose of this integrative = 0.62)12. However, the use of the EFS for the detection of
review was to examine the literature to determine progress in cognitive impairment (r = - 0.005, p = 0.801, n=30) was not
the establishment of validity and reliability for the measurement statistically significant (12).
of cognitive frailty. Other validated frailty instruments with unique operational
definitions have been described in the literature: the Frailty
Operational and Theoretical Definitions Index (FI), Clinical Frailty Scale, Study of Osteoporotic
Fractures (SOF), SPPB (gait speed, repeated chair stands, and
Establishing a comprehensive understanding of the new tandem balance tests) validated in the Established Population
construct cognitive frailty requires a critical review of what is for Epidemiologic Studies of the Elderly (EPESSE), and
known about the consensus on operational definitions and tools Tilburg Frailty Indicator (TFI) which includes three frailty
used to study frailty and cognitive impairment individually. domains (physical, psychological and social) (13–16). Several
frailty assessment tools are time consuming, not practical
Frailty except for research purposes, and have slightly different
The first definition of frailty was proposed in 1988 (6), but measurement properties. The literature reflects the lack of
since that time the international community has come to no consensus and ongoing debate about how to operationalize a
consensus on a definition of the term or an assessment tool definition for frailty (17).
to measure the condition (7). The International (I.A.N.A.)
Task Force on Frailty identified 17 cohort-based definitions, Cognitive Impairment
all using different frailty assessment tools. More recently, The theoretical and operational definition for the progressive
Rodríguez-Mañas et al, 2013 attempted to achieve consensus loss of memory unrelated to the normal aging process has
for an operational definition using a Delphi process, which been controversial. Mild cognitive impairment (MCI) was
resulted in consensus on the value of screening for physical first proposed by Petersen et al, 1999 then revised with the
frailty in the following six domains: physical performance, International Working Group on Mild Cognitive Impairment
including gait speed and mobility, nutritional status, mental (19). MCI is the most commonly used term to describe a
health, and cognition. Because there is still a need to identify a progressive measurable change in memory that differs from
specific combination of clinical and laboratory biomarkers for healthy aging adults. The recommended criteria for MCI is
a diagnosis, an operational definition was not recommended self and/or informant report of memory impairment and/or
(8). Even though consensus has not been reached regarding an evidence of decline over time on objective tasks with preserved
operational definition of frailty, the theoretical definition, which activities of daily living, and minimal impairment in complex
is generally agreed upon, describes frailty as a multidimensional instrumental functions with no diagnosis of dementia (19).
geriatric syndrome with increased vulnerability to stressors as Resulting from the research on MCI the Diagnostic Statistical
a result of reduced capacity of different physiological systems Manual-5 (DSM-5) included a category of neurocognitive
with adverse health outcomes that include falls, disability, disorder and distinguishes between mild (mNCD) and major
hospitalizations, and mortality (7, 9, 10). (mNCD) neurocognitive disorders to describe the heterogeneity

2
J Nutr Health Aging

THE JOURNAL OF NUTRITION, HEALTH & AGING©

in presentation for cognitive impairment (20). Theoretical Framework

The Mini-Mental Status Exam© (MMSE) is one of the most
commonly used screening tools for the assessment of MCI
and dementia in research and clinical settings. The MMSE
offers modest accuracy but has the best value for ruling-out a
diagnosis of dementia in community and primary care settings;
sensitivity (85.1%), specificity (85.5%), positive predictive
value (34.4%), and negative predictive value (98.5 %) (21).
Several cognitive screening instruments (CSI) are available,
yet many of them have not been validated for detecting early
cognitive impairment. Several CSI instruments have been
evaluated, specifically the Mini-Mental State Examination
(MMSE), the Six-Item Cognitive Impairment Test (6CIT),
the Montreal Cognitive Assessment (MoCA), the Test Your
Memory (TYM) test, and the Addenbrooke’s Cognitive Methods
Examination-Revised (ACE-R) for accuracy in diagnosing
dementia versus no dementia and mild cognitive impairment Through an extensive literature review, the variables of
versus no dementia (22). Using Cohen’s effect size, all of the interest were identified by asking the following questions: 1)
CSI instruments were effective for detection of dementia versus since the publication of the International Consensus Group
no dementia; the MoCA (1.45) performed best for detection of (I.A.N.A & I.A.G.G) what research has been conducted on
MCI and non-demented with medium ranges for the ACE-R cognitive frailty? and, 2) what psychometric measures are
(0.73), MMSE (0.69), 6CIT (0.65), and TYM (0.48) (22). being utilized to study the association of physical frailty and
cognitive decline? Even though research focused on validating
Cognitive Frailty tools for one or more of the phenotypes that make up frailty and
The International Consensus Group (I.A.A.A. /I.A.G.G.) dementia exists separately, the primary aim of this integrative
report addresses the need to focus research efforts on a clinical review was to identify psychometric tools used to measure the
condition characterized by the occurrence of physical frailty construct of cognitive frailty.
and cognitive impairment, in absence of overt dementia A systematic search of five databases was conducted
diagnosis or underlying neurological conditions (3). According along with a search of the reference lists for the retrieved
to the Consensus Group, cognitive frailty is considered a publications to identify published papers addressing the topic.
heterogeneous clinical syndrome in older adults with evidence The database searches were conducted in PubMed, CINAHL,
of: 1) physical frailty and cognitive impairment (Clinical Web of Science, PsycInfo, and ProQuest Dissertations &
Dementia Rating score of 0.5); and 2) exclusion of a clinical Theses. The articles included were limited to those written
diagnosis of Alzheimer’s Disease or other dementia (3). The in English and addressing an adult population. A full search
International Consensus Group suggested a list of possible strategy is provided in Appendix A. Medical Subject Headings
biological, clinical, and imaging markers for improving the (MeSH) and equivalent controlled vocabulary and keywords
detection for physical disability and neurodegenerative disease were utilized in each database as appropriate and yielded 1,119
(2, 3). The list was not intended to be complete, accurate articles with 1 article identified through other sources for a
or exhaustive; instead, the intent was to stimulate research total 1,120 articles. After deduplication of the initial list there
to further characterize a complex multidimensional geriatric were 723 articles to review. Figure 1, represents the number of
syndrome and encourage the development of preventive and articles reviewed in each step of the process.
therapeutic interventions (3). A review of titles and/or key word(s) of the 723 articles
Worsening cognitive impairment may increase the of English, Physical Frailty and Cognitive decline, and/or
occurrence of frailty, but frailty may be associated with Dementia. International articles were included to support
cognitive impairment (4). The mechanisms and the directional the application of the cognitive frailty construct from the
relationship behind the dynamic association of physical International Consensus Group (I.A.N.A/I.A.G.G). After
frailty and cognitive impairment presented in the theoretical reviewing all of the articles to ensure that the cognitive frailty
framework for cognitive frailty remains unexplained. In order construct was not addressed by earlier studies, a date limit of
to develop a deeper understanding, psychometric properties 2013 was set based on the publication of the cognitive frailty
for the instruments measuring cognitive frailty must be clearly construct by the International Consensus Group (I.A.N.A/
defined. I.A.G.G); this resulted in 185 articles. Following the review of
the abstracts and titles, 168 articles did not meet the inclusion
criteria for association of frailty and cognition. Of the 18 full-
text articles reviewed, 11 articles met the inclusion criteria;

3
J Nutr Health Aging

ASSESSING THE CURRENT STATE OF COGNITIVE FRAILTY: MEASUREMENT PROPERTIES

these articles were reviewed in-depth to determine validity and probable/possible diagnosis of dementia (26, 27). Although
reliability of the cognitive frailty measures. several studies reported baseline cognitive status, scores were
Data extraction, was used to identify the psychometric not always considered in the statistical model. This finding
properties based on the measurements provided in the article may be important because baseline cognition can decrease
or if the criteria could be found in the original longitudinal the association between frailty and all dementia outcomes;
study as referenced by the author. The level of evidence was association between frailty and dementia was stronger with
appraised for each study using the Center for Evidence Based higher baseline scores (HR 1.78, 95% CI 1.14-2.78) than those
Medicine Levels of Evidence (23). Studies were evaluated with lower baseline cognitive scores (HR 0.79, 95% CI 0.50-
with a systematic approach and rated based on their strength 1.26 p value for interaction = 0.02) (28).
of evidence. The operational definitions for both frailty and
cognition were reported separately to highlight the combination Figure 1
of tools being used to study the relationship between physical Search Strategy Diagram
frailty and cognition and report on measurement properties
and significant findings. A framework, presented in Table
1, was developed to report the operational definition criteria
being used for cognitive frailty based on impairment in the
physiological domains defined by The Interventions on Frailty
Working Group: mobility, balance, muscle strength, motor
processing, nutrition (often operationalized as nutritional status
or weight change/sarcopenia), cognition, endurance (including
feelings of fatigue and exhaustion), and physical activity (24).
Cognition was further defined in the framework based on
the use of neuropsychiatric testing and/or a clinical cognitive
assessment tool (i.e. MMSE or CDR) in the operational
definition. To accompany these results, and to help with
replication of the work, the search strategy and data extraction
results have been made available online.

Results

The association between phsycial frailty and cognitive

decline was established in cross-sectional and longitudinal
studies before the International Consensus Group (I.A.N.A/
I.A.G.G) proposed the definition of cognitive frailty in 2013
(25). Additionally, evidence presented in this review supports
the link between physical frailty and cognitive decline with
developing validity to support distinct relationships between
components of physical frailty and cognition in community-
dwelling older adults. Table 2 presents a comparison of
the screening tools used by the ten studies included in this Cross-sectional studies
review and those proposed by the International Consensus Six cross-sectional studies examined the association of
Group (I.A.N.A/I.A.G.G) as a framework for evaluating the frailty and cognitive decline using a modified CHS criterion (5).
development and validation of an operational definition for Functional status evaluations were added in several studies (26,
cognitive frailty. 29, 30) and co-morbidies, age, gender, BMI, and depression
None of the researchers explicity described using a were often considered in the covariate analysis (26, 27, 31). The
theoretical framework; however, all the studies discussed cross-sectional studies relied on clinincal evaluations including
components of cognitive frailty in relation to the International MMSE, executive tests, gait speed, grip strength, weight loss,
Consensus Group’s (I.A.N.A/I.A.G.G) proposed definition. and psychological markers (Table 2). Few of the studies used
All 11 studies examined the correlation of physical frailty biomarkers, and only one used imaging in the operational
and cognitive impairment. Additionally, six studies definition (30).
examined rate of change in frailty scores in associaton to
rate of deterioration of cognitive scores. Participants were Cohort study
non-demented at baseline in all but two studies, including One cohort study examined the associations between frailty
baseline amnestic Mild Cognitive Impairment (aMCI) and a and cognitive decline over 12 months (32). The study used

4
J Nutr Health Aging

THE JOURNAL OF NUTRITION, HEALTH & AGING©

Table 1
Operational Definitions of Cognitive Frailty

Reference Mobility/ Strength Balance Motor Pro- Nutrition/ Endurance/ Physical Neuropsy- Clinical
Gait Speed cessing Weight loss Fatigue Activity chiatric Cognitive
Testing Assessment
Tool¥
Shimada et X X X X X X X
al. 2013
Kulmala et X X X X X X
al. 2014
Buchman et X X X X X X
al. 2014
Rolfson et X X X X X X X X
al. 2013*
Oosterveld X X X X X X X
et al. 2014
McGough X X X X X X
et al. 2013
Alencar et X X X X X X X
al. 2013
Gray et al. X X X X X X X
2013
Solfrizzi et X X X X X X X X
al. 2013
Robertson X X X X X X X
et al. 2014
Han et al. X X X X X X
2014
*Rolfson et al. (2013) used 3 operational definitions: CHS, Edmonton Frail Scale, and Frailty Index; ¥ Clinical Cognitive Assessment Tool was defined as use of any of the following:
MMSE, MoCA, CDR, ADAS-Cog or CASI

the CHS criterion (5) with the addition of a functional status Validity
evaluation and tested the MMSE and Clinical Dementia Rating For all the studies in this review, criterion validity was
Scale (CDR). The study did not control for chronic diseases or examined for performance of the operationalization of various
depression. Additionally total sample size (n=182) was small, cogntive frailty measurements. Predictive and discriminant
affecting power for individual classifications of frailty (non- validity was commonly reported as odds ratio (OR) or
frail n=43, pre-frail n=104, frail n=35) (30). hazard ratio (HR); two studies used Pearson correlations and
multiple linear regression models to establish associations
Longitudinal studies between components of physical frailty and cognitive function.
Results from four longitudinal studies were published after Predictive validity was established by investigating frailty
2013. A modified CHS criterion (5) was used in three of and rate of change in cognition or correlation of frailty and
the studies. One study used more than one frailty instrument cognitive decline. Discriminant validity was established by
to determine if the relationship between neurocogntive analyzing the relationship between measures of frailty (frail,
speed (NCS) and frailty was affected by how frailty was pre-frail, and robust) and type of demenia (MCI, clinically
operationalized (33). The use of biomarkers, clinical markers, diagnosed dementia, vascular dementia, and Alzheimer’s) (26,
and imaging varied among studies. The use of biomarkers 28, 30, 32). All of the studies evaluated community-dwelling
and imaging was more commonly used in the longitudinal older adults for which the CHS frailty measures are validated
studies than cohort and cross-sectional studies (Table 2). (5). Only one study compared more than one operational
Functional status evaluation was added in one study (34) and defintion of frailty: CHS, FI, and EFS (33). Heterogeneity was
co-morbidities were considered in the analysis for all of the present in the objective measures, and the terminology-specific
studies. language for the components of the CHS frailty construct often
varied from the validated CHS criteria (5).

5
J Nutr Health Aging

ASSESSING THE CURRENT STATE OF COGNITIVE FRAILTY: MEASUREMENT PROPERTIES

Heterogeneity was present in the objective measures for (β – 0.18) (36) were predictive for changes in global cognition.
cognitive assessment and neuropsychiatric testing. Two studies Psychological markers were frequently used for the
assessed global cognition with the MMSE (30, 34), four used assessment of endurance, fatigue, or depression. However,
the MMSE and domain specific neuropsychiatric testing (26, variability existed in the type of assessment scale used and how
29, 32, 33), three used only domain neuropsychiatric testing the psychological marker was operationalized. Psychological
(27, 28, 35), and one assessed global cognition with both the markers were typically used to either assess endurance for
MMSE and MoCA with domain specific neuropsychiatric fatigue in the CHS criteria (29, 35) or considered as a covariate
testing (36). The Cognitive Dementia Rating scale (CDR) had in the statistical analysis (27, 28, 32, 34). Variability in the
no predictive validity with evidence of no difference between psychological markers can be seen in Table 2 and online
frailty and cognition (relative risk = 2.1; p = 0.393) (32). The material.
National Center for Geriatrics and Gerontology-Functional
Assessment tool (NCGG-FAT) had good test-retest reliability Reliability
with moderate to high external validity (Person r= 0.496 to Due to the heterogeneity in the objective measures for
0.842). The MMSE continues to be the most commonly used frailty, reliability was not consistently examined for cognitive
clinical cognitive assessment tool for operationalizing cognitive frailty. The limited reliability and variability in the operational
frailty (25); concurrent validity (Pearson r = 0.776; p < 0.001) measurements used for the CHS frailty criteria add challenges
and reliability test-retest (Person r = 0.827; p = 0.001) (37) with to establishing an operational definition for cognitive frailty.
neuropsychiatric testing predictive and discriminate validity Motor performance was the only measurement for which
is established by the rate of change in MMSE and CHS frailty validity and reliability was established (34).
criterion (32).
Predictive validity was established in four main areas: 1) Feasability
frailty and type of dementia: MCI (OR 2.0; p= <0.001) and Instrumental assessments for cognitive frailty are currently
(OR 7.4, 95% CI 4.2-13.2) (29, 30); vascular dementia (OR time-consuming, expensive, require extensive training, and
6.7, 95% CI 1.6-27.4) and (HR 2.68, 95% CI 1.16-7.17) (30, the clinical translation properties are not clear. The addition
34); and Alzheimer’s dementia (OR 3.2, 95% CI 1.7-6.2), (HR of biomarkers and imaging potentiates the complexity of
1.08, 95% CI 0.74-1.57), and (HR 0.62, 95% CI 0.20-1.89) the feasability for measures and complicates the process for
(28, 30, 34). The relationship between frailty and vascular detection of cognitive frailty in the clinical setting.
dementia (VaAD ) is further supported by the rate of change in
frailty x macroinfarcts (r= 0.032, p < 0.001) (35). Evidence of Discussion
convergent validity exists between dementia and non-dementia
types with findings to support the associations between frailty The findings from this review continue to support evidence
and non-Alzheimer’s dementia (OR 2.57, 95% CI 1.08-6.11). for the association between physical frailty and cognitive
2) Frailty and the individual domains of cognitive function decline. However, while cross-sectional studies have detected
was identified by evaluating the relationship of neurocognitive a relationship, further studies are needed to determine causal
speed and change in cognition using regression coefficients (33) pathways (38). Studies continue to use different combinations
and evaluation of the MMSE subdomains. Individual domains of measurement instruments for cognitive frailty, but are
of cognitive function were found to be gender specific (31). measuring similar domains of physical frailty and cognition.
Predictive validity was dependent on the frailty operational Based on the findings in this review the CHF criteria with
definition; Frailty Index (FI) and NCS (OR 0.87, 95% CI 0.81- measures of mobility/gait speed, strength, nutrition/weight
0.95) compaired to the modified CHS and EFS which found no loss, endurance/fatigue, and physical activity, neuropsychiatric
correlation with neurocognitive speed (33). testing and a cognitive assessment tool was the most common
3) Individual components of frailty and individual domains operational definition (Table 1). Further testing of the
of cognitive function associations inculded slow gait and cognitive frailty construct should attempt to provide validity
executive function (β -0.20), attention (β -0.25), processing and reliability for objective measures and scales which are
speed (β -0.16) (36), word recall (β -.0.18, p = 0.02), and based on self-report. Self-report scales must prove to be stable
logical memory (β = 0.04, p =0.04) (27). Weak grip was over time (test-retest reliability), and those administered by
predictive for changes in executive function (β – 0.16, p several individuals need to exhibit good inter-rater reliability.
=0.008) (27). Physical activity was associated with changes in Additionally, inclusion of a theoretical framework will provide
executive function (β = -0.18, p= 0.02) and word recall (β = a structure for generating cumulative knowledge on which
0.17, p= 0.02) (27). interventions can be based.
4) Individual components of frailty and global cognitive Studies are starting to deconstruct the relationship
function were found in several studies (27, 28, 34–36). between the components of physical frailty and cognitive
Individual components included grip strength (r = - 0.51, p < decline. Unravelling of the complex cognitive frailty
0.001), gait speed (r = -0,067, p < 0.001) (35), and exhaustion construct will refine the operational definition and improve an

6
 Table 2
Use of biological, clinical, and imaging markers for cognitive frailty: International Consensus Group (I.A.N.A/I.A.G.G)
J Nutr Health Aging

Shimada et Kulmala et Buchman et Rolfson et Oosterveld McGough Alencar et Gray et al. Solfrizzi et Robertson Han et al.
al. 2013 al. 2014¥ al. 2014 al. 2013 et al. 2014 et al. 2013 al. 2013 2013 al. 2013 et al. 2014 2014

Biomarkers
Inflammatory markers (e.g. CRP, IL-6)
Beta-amyloid protein (aβ) X
aPOEε4 genotype X X
Anemia
Serum albumin X
Cholesterol Xβ
Vitamin D status
Clinical markers
MMSE X X X X X X X X X
Executive tests X X X X X X
ADAS-Cog X
CDR X X X

7
MoCA X
Gait speed X X X X X X X X X X X
Hand grip strength X X X X X X X X X X X
Weight loss X X X X X X X X X X X
Psychological marker: GDS X® X£ X€ X§ Xф XΩ Xᵏ X§
Actigraphy
Imaging
Dual energy
THE JOURNAL OF NUTRITION, HEALTH & AGING©

X-ray absorptiometry scans (DEXA)

Cerebral Computed tomography X X X
Cerebral Magnetic resonance imaging X X X
Functional MRI
Diffusion tensor imaging (DTI)
Tractography
Electrophysiological methods
Cognitive evoked potentials
¥ CT scan, MRI, and laboratory tests (not specified) were used to make a diagnosis of vascular dementia, Alzheimer’s disease, Lewy bodies, and dementia related to other medical causes;® Partial GDS scale; £ Psychological maker
evaluated with two questions from the Center for Epidemiologic Studies; € Psychological maker evaluated with the Edmonton Frail Scale; § GDS-15 scale; Ф GDS-15 scale and Cornell Depression Scale; Ω Psychological maker
evaluated with Center for Epidemiological Studies Depression; ᵏ GDS 30 scale; β Reported in original study.
J Nutr Health Aging

ASSESSING THE CURRENT STATE OF COGNITIVE FRAILTY: MEASUREMENT PROPERTIES

understanding of the clinical distinction between cognitive understanding of the directional relationship between physical
impairment due to physical frailty and an isolated neurological frailty and cognitive impairment, gender differences, and
condition. Disentangling the association between frailty and identify biomarkers to assist with detection of diagnosis and
cognitive decline requires the use of convergent validity to disease progression.
determine if the cognitive frailty construct is able to distinguish Acknowledgments: We would like to thank Elaine Amella, Ph.D., RN, FAAN, Martina
among between different types of dementia (e.g., Vascular, Mueller, Ph.D., and Mathew Gregoski, Ph.D., MS for their support.
Alzheimer’s, Lewy Body, and Parkinson’s dementia) (27). The
Conflict of interest: The authors have no conflict of interests to report
association of cognitive decline and frailty may be responsible
for part of the heterogeneity in the presentation of dementia. Ethical Standards: To reduce bias in this rigorous review the authors adhered to the
Whittemkore & Knafl., 2005 and PRISMA guidelines. This study did not use human
Movement toward evaluating specfic domains of cognitive subjects.
impairment such as executive functioning and psychomotor
speed versus a global assessment of dementia will facilitate References
an understanding of the implications for cogintive frailty.
However, the current lack of validity and reliability of a 1. NIH, WHO. Global health and aging. 2011:1-32. http://www.nia.nih.gov/sites/
cognitive frailty operational definition means that it is not default/files/global_health_and_aging.pdf. Accessed February 7, 2015.
2. Buchman AS, Bennett DA. Cognitive frailty. J Nutr Health Aging. 2013;17(9):738-
possible to recommend translation of measures to detect the 739. doi:10.1007/s12603-013-0397-9.
presence of risk factors that may predict cognitive frailty in the 3. Kelaiditi E, Cesari M, Canevelli M, et al. Cognitive frailty: rational and definition
from an (I.A.N.A./I.A.G.G.) international consensus group. J Nutr Health Aging.
clinical settings. 2013;17(9):726-734. doi:10.1007/s12603-013-0367-2.
A limitation of this review was the exclusion of studies that 4. Canevelli M, Kelaiditi E. The complex construct of mild cognitive impairment: Be
aware of cognitive frailty. J Frailty Aging. 2014;3(2):87-88.
did not address the cognitive frailty construct. In the future, a 5. Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a
review of the literature focused on individual physical function phenotype. J Gerontol A Biol Sci Med Sci. 2001;56(3):M146-M156. http://www.
measures may identify other markers associated with cognitive ncbi.nlm.nih.gov/pubmed/11253156. Accessed August 27, 2014.
6. Woodhouse, K. W., Wynne, H., Baillie, S., James, O. F. W., & Rawlins MD. Who
impairment. Further research with epidemiological and are the frail elderly? Q J Med. 1988;68(1):505-506.
population based studies that includes diverse ethnic and social 7. Van Kan GA, Rolland Y, Bergman H, Morley JE, Kritchevsky SB, Vellas B. The
I.A.N.A. task force on frailty assessment of older people in clinical practice. J Nutr
economic groups will help establish a better understanding Heal Aging. 2008;12(1):29-37. doi:10.1007/BF02982161.
of the prevalence of cognitive frailty. The majority of studies 8. Rodríguez-Mañas L, Féart C, Mann G, et al. Searching for an operational
definition of frailty: a Delphi method based consensus statement: the frailty
in this review either did not report ethnicity or the sample operative definition-consensus conference project. J Gerontol A Biol Sci Med Sci.
included a high proportion of white (88%-99%) females 2013;68(1):62-67. doi:10.1093/gerona/gls119.
(58%-80%). Only two studies provided a population-based 9. Fried LP, Ferrucci L, Darer J, Williamson JD, Anderson G. Untangling the Concepts
of Disability, Frailty, and Comorbidity: Implications for Improved Targeting
estimate of cognitive frailty with samples of 5,104 Japanese and Care. Journals Gerontol Ser A Biol Sci Med Sci. 2004;59(3):M255-M263.
(29) and 4,649 Irish community-dwelling older adults (36). doi:10.1093/gerona/59.3.M255.
10. Panza F, Solfrizzi V, Frisardi V, et al. Different models of frailty in predementia and
Understanding how demographics effect the measurement of dementia syndromes. J Nutr Health Aging. 2011;15(8):711-719. http://www.ncbi.
cognitive frailty are important since psychometric tools may nlm.nih.gov/pubmed/21968870. Accessed February 5, 2015.
11. Nguyen T, Cumming R, Hilmer S. A Review Of Frailty In Developing Countries.
be effected by populations which have higher rates frailty, Ageing Res Rev. 2013;20C(9):741-743. doi:10.1007/s12603-013-0398-8.
comorbidity, cardiovascular disease, poorer health, decreased 12. Rolfson DB, Majumdar SR, Tsuyuki RT, Tahir A, Rockwood K. Validity and
access to care, and low education and income (5). Inclusion reliability of the Edmonton Frail Scale. Age Ageing. 2006;35(5):526-529.
doi:10.1093/ageing/afl041.
of chronic diseases, such as depression and cardiovascular 13. Rockwood K, Stadnyk K, MacKnight C, McDowell I, Hébert R, Hogan DB. A brief
disease, as a part of the study design is an important part clinical instrument to classify frailty in elderly people. Lancet. 1999;353(9148):205-
206. doi:10.1016/S0140-6736(98)04402-X.
of describing other factors that may contribute to cognitive 14. Gobbens RJJ, van Assen MALM, Schalk MJD. The prediction of disability by self-
frailty over time. Additionally, adjustment for the presence reported physical frailty components of the Tilburg Frailty Indicator (TFI). Arch
Gerontol Geriatr. 2014;59(2):280-287. doi:10.1016/j.archger.2014.06.008.
of apolipoprotein (APOE) є4 alleles and other biomarkers 15. Ensrud KE, Ewing SK, Taylor BC, et al. Comparison of 2 frailty indexes for
(e.g. inflammatory makers, beta-amyloid protein) could help prediction of falls, disability, fractures, and death in older women. Arch Intern Med.
describe the pathophysiological mechanisms. 2008;168(4):382-389. doi:10.1001/archinternmed.2007.113.
16. Studenski S, Perera S, Wallace D, et al. Physical performance measures in the
The early detection of cognitive decline emphasizes a clinical setting. J Am Geriatr Soc. 2003;51(3):314-322. http://www.ncbi.nlm.nih.gov/
promising focus for the development of preventive and pubmed/12588574. Accessed August 13, 2015.
17. Pel-Littel RE, Schuurmans MJ, Emmelot-Vonk MH, Verhaar HJJ. Frailty: defining
therapeutic interventions. Current studies suggest the and measuring of a concept. J Nutr Health Aging. 2009;13(4):390-394. http://www.
importance in understanding both constructs separately as a ncbi.nlm.nih.gov/pubmed/19300888. Accessed March 1, 2015.
18. Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG, Kokmen E. Mild
way to deconstruct dissociable components, describe common cognitive impairment: clinical characterization and outcome. Arch Neurol.
pathologies, and develop a single operational definition which 1999;56(3):303-308. http://www.ncbi.nlm.nih.gov/pubmed/10190820. Accessed
would allow for targeted interventions. Ensuring validity and January 14, 2015.
19. Winblad B, Palmer K, Kivipelto M, et al. Mild cognitive impairment--beyond
reliability in the measures used is paramount if providers are to controversies, towards a consensus: report of the International Working Group on
identify individuals at risk for pathological non-normal aging Mild Cognitive Impairment. J Intern Med. 2004;256(3):240-246. doi:10.1111/j.1365-
2796.2004.01380.x.
changes and develop interventions to improve the quality of 20. Sachs-Ericsson N, Blazer DG. The new DSM-5 diagnosis of mild neurocognitive
life of older adults. Further research is needed to establish an disorder and its relation to research in mild cognitive impairment. Aging Ment
Health. 2015;19(1):2-12. doi:10.1080/13607863.2014.920303.
operational definition for cognitive frailty, develop a better

8
J Nutr Health Aging

THE JOURNAL OF NUTRITION, HEALTH & AGING©

21. Mitchell AJ. A meta-analysis of the accuracy of the mini-mental state examination 30. Kulmala J, Nykänen I, Mänty M, Hartikainen S. Association between frailty
in the detection of dementia and mild cognitive impairment. J Psychiatr Res. and dementia: a population-based study. Gerontology. 2014;60(1):16-21.
2009;43(4):411-431. doi:10.1016/j.jpsychires.2008.04.014. doi:10.1159/000353859.
22. Larner AJ. Effect Size (Cohen’s d) of Cognitive Screening Instruments Examined 31. Han ES, Lee Y, Kim J. Association of cognitive impairment with frailty in
in Pragmatic Diagnostic Accuracy Studies. Dement Geriatr Cogn Dis Extra. community-dwelling older adults. Int Psychogeriatr. 2014;26(1):155-163.
2014;4(2):236-241. doi:10.1159/000363735. doi:10.1017/S1041610213001841.
32. Alencar MA, Dias JMD, Figueiredo LC, Dias RC. Frailty and cognitive impairment
23. OCEBM Levels of Evidence Working Group. The Oxford Levels of Evidence 2. among community-dwelling elderly. Arq Neuropsiquiatr. 2013;71(6):362-367.
Oxford Cent Evidence-Based Med. 2011. http://www.cebm.net/index.aspx?o=5653. doi:10.1590/0004-282X20130039.
Accessed February 7, 2015. 33. Rolfson DB, Wilcock G, Mitnitski A, et al. An assessment of neurocognitive speed in
24. Ferrucci L, Guralnik JM, Studenski S, Fried LP, Cutler GB, Walston JD. Designing relation to frailty. Age Ageing. 2013;42(2):191-196. doi:10.1093/ageing/afs185.
randomized, controlled trials aimed at preventing or delaying functional decline 34. Solfrizzi V, Scafato E, Frisardi V, et al. Frailty syndrome and the risk of vascular
and disability in frail, older persons: a consensus report. J Am Geriatr Soc. dementia: the Italian Longitudinal Study on Aging. Alzheimers Dement.
2004;52(4):625-634. doi:10.1111/j.1532-5415.2004.52174.x. 2013;9(2):113-122. doi:10.1016/j.jalz.2011.09.223.
25. Robertson DA, Savva GM, Kenny RA. Frailty and cognitive impairment--a review 35. Buchman AS, Yu L, Wilson RS, Boyle PA, Schneider JA, Bennett DA. Brain
of the evidence and causal mechanisms. Ageing Res Rev. 2013;12(4):840-851. pathology contributes to simultaneous change in physical frailty and cognition in old
doi:10.1016/j.arr.2013.06.004. age. J Gerontol A Biol Sci Med Sci. 2014;69(12):1536-1544. doi:10.1093/gerona/
26. Oosterveld SM, Kessels RPC, Hamel R, et al. The influence of co-morbidity glu117.
and frailty on the clinical manifestation of patients with Alzheimer’s disease. J 36. Robertson DA, Savva GM, Coen RF, Kenny R-A. Cognitive function in the prefrailty
Alzheimers Dis. 2014;42(2):501-509. doi:10.3233/JAD-140138. and frailty syndrome. J Am Geriatr Soc. 2014;62(11):2118-2124. doi:10.1111/
27. McGough EL, Cochrane BB, Pike KC, Logsdon RG, McCurry SM, Teri jgs.13111.
L. Dimensions of physical frailty and cognitive function in older adults with 37. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method
amnestic mild cognitive impairment. Ann Phys Rehabil Med. 2013;56(5):329-341. for grading the cognitive state of patients for the clinician. J Psychiatr Res.
doi:10.1016/j.rehab.2013.02.005. 1975;12(3):189-198. http://www.ncbi.nlm.nih.gov/pubmed/1202204. Accessed July
28. Gray SL, Anderson ML, Hubbard RA, et al. Frailty and incident dementia. J Gerontol 9, 2014.
A Biol Sci Med Sci. 2013;68(9):1083-1090. doi:10.1093/gerona/glt013.
29. Shimada H, Makizako H, Doi T, et al. Combined prevalence of frailty and mild
cognitive impairment in a population of elderly Japanese people. J Am Med Dir
Assoc. 2013;14(7):518-524. doi:10.1016/j.jamda.2013.03.010.