Age and Ageing 2017; 46: 383–392
doi: 10.1093/ageing/afw247
Published electronically 7 January 2017
© The Author 2017. Published by Oxford University Press on behalf of the British Geriatrics Society.
This is an Open Access article distributed under the terms of the Creative Commons Attribution
Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-
commercial re-use, distribution, and reproduction in any medium, provided the original work is
properly cited. For commercial re-use, please contact journals.permissions@oup.com
REVIEW
Interventions to prevent or reduce the level
of frailty in community-dwelling older adults:
a scoping review of the literature
and international policies
MARTINE T. E PUTS
1
,SAMAR TOUBASI
1
,MELISSA K. ANDREW
2
,MAUREEN C. ASHE
3,4
,JENNY PLOEG
5
,
ESTHER ATKINSON
6
,ANA PATRICIA AYALA
7
,ANGELIQUE ROY
8
,MIRIAM RODRÍGUEZ MONFORTE
1
,
H
OWARD BERGMAN
9
,KATHY MCGILTON
8
1
University of Toronto—Lawrence S. Bloomberg Faculty of Nursi ng, Toronto, Ontario, Canada
2
Dalhousie University —Division of Geriatric Medicine, Halifax, Nova Scotia, Canada
3
University of British Columbia—Family Practice, Vancouver, British Columbia V5Z 1M9, Canada
4
University of British Columbia—Centre for Hip Health and Mobility -7 F-2635 Laurel St., Vancouver, British Columbia V5Z
1M9, Canada
5
McMaster University—School of Nursing, Hamilton, Ontario, Canada
6
University of Toronto—Gerstein Information Science, Toronto, Ontario, Canada
7
University of Toronto—Gerstein Information Science Centre, Toronto, Ontario, Canada
8
University Health Network—Toronto Rehabilitation Institute, Toronto, Ontario, Canada
9
Department of Family Medicine, McGill University—Geriatric Medicine, Montréal, Canada
Address correspondence to: M. T. E. Puts. Tel: (+1) 416 978 6059; Fax: (+1) 416 978 8200. Email:
martine.puts@utoronto.ca ,
martineputs@hotmail.com
Abstract
Background: frailty impacts older adults’ ability to recover from an acute illness, injuries and other stresses. Currently, a
systematic synthesis of available interventions to prevent or reduce frailty does not exist. Therefore, we conducted a scoping
review of interventions and international policies designed to prevent or reduce the level of frailty in community-dwelling
older adults.
Methods and analysis: we conducted a scoping review using the framework of Arksey and O’Malley. We systematically
searched articles and grey literature to identify interventions and policies that aimed to prevent or reduce the level of frailty.
Results: fourteen studies were included: 12 randomised controlled trials and 2 cohort studies (mean number of participants
260 (range 51–610)), with most research conducted in USA and Japan. The study quality was moderate to good. The inter-
ventions included physical activity; physical activity combined with nutrition; physical activity plus nutrition plus memory
training; home modifications; prehabilitation (physical therapy plus exercise plus home modifications) and comprehensive
geriatric assessment (CGA). Our review showed that the interventions that significantly reduced the number of frailty mar-
kers present or the prevalence of frailty included the physical activity interventions (all types and combinations), and preha-
bilitation. The CGA studies had mixed findings.
Conclusion: nine of the 14 studies reported that the intervention reduced the level of frailty. The results need to be inter-
preted with caution, as only 14 studies using 6 different definitions of frailty were retained. Future research could combine
interventions targeting more frailty markers including cognitive or psychosocial well-being.
Keywords: scoping review, frail elderly, interventions, physical activity, nutrition, older people
383
Downloaded from https://academic.oup.com/ageing/article/46/3/383/2863854 by guest on 20 August 2022
Introduction
With increasing age the risk of frailty increases. While there
are several definitions of frailty, it is seen as a multidimen-
sional concept with various indicators such as weight loss,
lack of physical activity and lack of strength [
1, 2]. A sys-
tematic review showed that the weighted prevalence of
frailty in community-dwelling older adults was 10.7% (ran-
ging between 4.0 and 59.1) [
3].
Previous research has shown that frail older adults are
moreatriskforadversehealthoutcomes, such as falls, mobil-
ity decline, hospitalisation, institutionalisation and increased
risk of death [
4]. Researc h highlights that frailty is potentially
reversible [
5]. Therefore, it is important to identify frailty to
prevent, reduce and postp one adve rse health con sequences
for older adults and their caregivers . There have been many
studies examining the prevalence of frailty and validity of sev-
eral frailty models. How ever, there has not been a comprehen-
sive review of interventions and/or policies that can prevent
or reduce the level of frailty in community-dwelling older
adults . Ideally, these interventions could be implemented by
primary healthcare providers and/or older adults and /or their
families/caregivers, and w ould focus on improving outcomes
that are important to older adults [
6]. A comprehensive review
of interventions to prevent and reduce frailty is also important
for healthcare professionals who will want to ensure that they
are well equipped to assist community-dwelling older adults
and their families/caregiv ers once frailty is identified.
Therefore, in this scoping review, our aim was to identify
relevant peer-reviewed and grey literature on interventions
and policies to prev en t and/or delay frailty in community-
dwelling older adults . Our research question was: Which inte-
rventions and policies focused on reducing frailty exist and to
what extent are these interventions effective? We have chosen
the scoping review methodology as it allows for inclusion of
quantitative and qualitative studies, grey literature and includes
a stakeholder consultation which we considered important for
future intervention development and uptake.
Methods
The scoping review methodology protocol has been pub-
lished [
7] but we describe it briefly below.
Review methodology
We used the scoping review methodological framework as
described by Arksey and O’Malley [
8] with the modifica-
tions recommended [
9, 10] Our six-stage scoping review
model included: (i) identifying the research question; (ii)
identifying relevant studies; (iii) selecting studies; (iv) chart-
ing data; (v) summarising and reporting the results and (vi)
consulting with stakeholders. We followed the reporting
guidelines for PRISMA statement [
11].
Search methods
Our inclusion criteria were: original publications of an inter-
vention to prevent or to reduce the level of frailty in
community-dw elling older adults aged ≥65 years; and/or
studies that included a wider age range were eligible if the
mean/median age of the study population was aged 65 years
and over and/or if they had included a subgroup analysis for
the population aged >65 years. Included studies needed to
provide a measure of frailty before and after the intervention.
Grey literature that had information on the effect of the inter-
ventions or policies to prevent or reduce frailty were eligible.
We searched for publications dated between January 2000
and February 2016 using key words and Medical Subject
Headings terms such as frail elderly or frailty AND interven-
tions or evaluation study randomised trial AND Aged or
65 and over or senior. Searches (including grey literature)
wereperformedwithnolanguagerestrictionsandcarried
out by three librarians on our team. See Appendix A for the
Medline search strategy used.
We searched the Cochrane Central Register of
Controlled Trials (CENTRAL), Cochrane Effective Practice
and Organisation of Care Group (EPOC), MEDLINE,
EMBASE, Cumulative Index to Nursing and Allied Health
(CINAHL), Allied and Complementary Medicine (AMED),
Psych INFO, Ageline, Sociological Abstracts, Web of
Science, Applied Social Sciences Index and Abstracts
(ASSIA), Database of Abstracts of Reviews of Effects
(DARE), Health Technology Assessment (Canadian Search
Interface) databases and reference lists of included studies.
We searched trial registries for potential studies in progress.
Study selection
We included studies through a two-step process (see
Figure
1 PRISMA flow chart). First, abstracts were screened
by two independent reviewers. Then, all potentially relevant
full articles were reviewed for study inclusion by two
reviewers. The grey literature was searched by the librarians
and reviewed by two reviewers.
Data abstraction
We used standardised data collection forms developed by the
research team. Data were abstracted by two reviewers inde-
pendently and compared. The information abstracted in-
cluded: characteristics of the study population, study design
details , frailty assessment used and levels of frailty of study
participants, intervention details and outcomes, analyses used.
Quality assessment
We assessed the quality of the included studies using the
Mixed Methods Appraisal Tool (MMAT) [
12]. The MMAT
is a quality assessment tool that is useful to assess qualitative,
quantitative and mixed method studies. Incomplete report-
ing, preventing reproducibility and transparency in the meth-
ods, and factors that can lead to selection bias such as low
response rates (<60%), contamination in trials, measurement
error, etc. can all lead to lower quality ratings. By using the
MMAT it allowed us to use one quality assessment tool to
review all studies. We used the MMAT to review study qual-
ity but we did not exclude any study based on the score.
M. T. E. Puts et al.
384
Downloaded from https://academic.oup.com/ageing/article/46/3/383/2863854 by guest on 20 August 2022
Data analysis
We summarised the results using a narrative descriptive
synthesising approach. There was heterogeneity in study
inclusion criteria, interventions, frailty assessment tools and
outcomes, thus a pooled analysis was not appropriate.
Consultation
We organised a stakeholder consultation (step 6 of the
framework) on 19 May 2016 including older adults, family
caregivers, representatives of provincial and national com-
munity organisations, senior organisations and primary care
organisations to provide feedback on the findings and to
develop next steps in research and practice. An older adult
and caregiver reviewed the results for accuracy. We sum-
marised their feedback and present it below.
Results
Characteristics of included studies
In this review, 14,563 abstracts were retrieved, and after dedu-
plication 13,641 abstracts were reviewed by two independent
reviewers. One hundred and nine articles were selected for full
text review and 14 studies were included in this review. Twelve
studies were randomised controlled trials (RCTs) [
13–23]and
two studies were part of larger cohort studies [
24, 25](see
Supplementary data, Tables A and B, available at Age and
Ageing online). Seven studies were conducted in Asia [
5, 17–
19, 21, 24, 25], 3 in Europe [16, 20, 23], 3 in USA [13–15]
and 1 in Australia [
22]. Sample sizes varied from 51 to 610
with a mean of 260 participants and a total of 3,632 partici-
pants . The total and the mean number of participants for
RCTs were 2,475 and 206. The study mean age ranged from
70 [
5]to86years[23]. The percentage women in the study
ranged from 48% [
21] to 100% [18, 19].
Quality of the included studies
The quality assessment results are presented in
Supplementary data, Table C, available at Age and Ageing
online. The quality was moderate to good for most studies.
The response rate was not reported for the two cohort stud-
ies [
24, 25] and for the RCTs it varied between 39% [18]
and 100% [
20]. The randomisation, allocation concealment
and the level of blinding were not always described [
5, 14,
16, 20, 21]. For four studies it was not reported whether the
analysis was intent-to-treat [
14, 16, 18, 19]. Two studies had
Records identified through
database searching
(n = 14563)
Screening
Included Eligibility
Identification
Additional records identified
through other sources
(n = 1 )
Records after duplicates removed
(n = 13641)
Records screened
(n = 13641)
Records excluded
(n = 13536)
Full-text articles assessed
for eligibility
(n = 109)
Full-text articles excluded, with
reasons
(n = 53) no frailty outcomes
(n = 19) no frailty intervention
(n = 12) review/editorial
(n = 11) study protocol
Studies included in
qualitative synthesis
(n = 14 )
Studies included in
quantitative synthesis
(meta-analysis)
(n = 0 )
Figure 1. PRISMA FLOW CHART of the selection of included studies in the review.
Interventions to prevent or reduce the level of frailty in community-dwelling older adults
385
Downloaded from https://academic.oup.com/ageing/article/46/3/383/2863854 by guest on 20 August 2022
a drop-out rate greater than 20% and thus selection bias
may be present [
14, 16].
Interventions studied
The interventions included physical activity compared to con-
trol (four studies) [
13, 14, 16, 24]; physical activity combined
with nutrition compared to control (four studies) [
17–20];
physical activity plus nutrition plus memory training com-
pared to control (one study) [
5]; home modifications (one
study); prehabilitation physical therapy (PT; plus exercise plus
home modifications) (one study) [
15] and geriatric assessment
(three studies) [
21–23] (T a ble 1 and Supplementary data,
Table D , available at Age and Ageing online).
The physical activity interventions included group classes
in all nine studies [
5, 13, 14, 16–20, 24] with sessions ranging
in frequency from once weekly [
19] to 5 days per week [20].
Most included strength, balance, coordination, flexibility and
aerobic exercises provided by exercise professionals, which
were progressively increased based on the individuals’ com-
petency and performance. The studies that examined a nutri-
tional arm included interventions ranging from milk fat
supplementation [
18], supplementation using multi-fibre for-
mula enriched with iron, folate, vitamin B6, B12, D and cal-
cium [
5], education [19, 20] and cooking classes based on
healthy nutrition focused on the strengthening of muscles
through protein and vitamin D rich ingredients in addition
to supplementation of vitamin D [
19], and one study com-
bined nutritional consultation with exercise [
17].
The prehabilitation intervention included PT assessment,
environmental assess ment of the home and a prog res sive
competency-based exercise prog ramme. The study by
Gustafsson et al. had three arms in total including two
intervention ar ms, o ne g r oup that received a single p re-
ventative home visit and one arm that attended 4 weekly
educational meetings and a follow-up home visit. The
study by Cameron et al. [
22] evaluated a multifactorial
inter vention including a comprehensive geriatric assess-
ment and case management. Li et al. [
21] used a similar
inter vention to that of Cameron et al. The cognitive inter-
ventioninthestudyofNget al. [
5] consisted of a 2-hour
weekly training to stimulate shor t-term memory and
improve attention and inform ati on processing skill s, as
well as reasoning and problem-solving capacity for a 12-
week period followed by a booster session. T he s tudy by
Ng et al. [
5] in cluded a combination of exercise, nutrition
and cognitive interventions.
The control groups were given successful/healthy ageing
interventions (booklets and/or classes) [
13, 15, 17], home
low-intensity exercise programme [
14], usual care [5, 16,
19–23] or placebo supplements [18].
Effect of interventions on frailty
The frailty definitions used are listed in Table 1 and most
studies used the Cardiovascular Health Study frailty phenotype
(also known as Fried’s phenotype) (10 studies) [
5, 13, 16–23],
the Tinetti/Gill criteria (2 studies) [
14, 15] or the Japanese
Frailty check list (2 studies) [
24, 25].
The studies that compared exercise to control conditions
showed a significant reduction in the total number of frailty
markers [
13], all individual frailty markers studied [14, 16]
or reduction in prevalence of frailty [
24] (see Table 2 and
Supplementary data, Table E, available at Age and Ageing
online). These findings were significant up to 12 months
following the completion of the study. Exercise plus nutri-
tion compared with control identified a significant reduction
in the prevalence of frailty [
17, 18, 20] and mean
Edmonton Frailty Scale score [
20]. Kwon et al. [19] showed
a significant impact on improving grip strength immediately
after intervention in the exercise arm compared to control
only. Ng et al. showed a significant reduction in the preva-
lence of frailty ranging from 36% to 48% with the highest
reduction in the combination intervention (exercise, nutri-
tion and cognition) followed by exercise, cognitive function
and nutrition at 12 months.
The studies using geriatric assessment had mixed find-
ings and only Cameron et al. showed a significant reduction
in the prevalence of frailty and the total number of frailty
markers [
22]. Mitoku showed non-significant results of
home modifications on frailty prevalence [
25].
Adherence and adverse events
While not all studies reported adherence rates, those that
did report them ranged from 20% to 28% (for the different
components of the intervention) [
17] to 100% [14].
Tarazona et al. [
20] reported that 77% of participants in the
intervention group had 3–6 h. per week exercise and
attendance of 50% was associated with an OR of reduction
in frailty of 4.4 (95% confidence interval 1.2–16.0) and a
number needed to treat of 3.2 (2.0–4.5).
The adverse event reports were very low for the total of
3,632 participants included in the 14 studies ranging from 0
adverse events [
16], [22, 23], to 2 injuries each in the exer-
cise arms of two studies [
5, 14]. Two studies [13, 15]
reported similar adverse events rates in inter vention and
control groups.
International policies on frailty
Diverse international policies have been developed aiming to
prevent or improve frailty [
26–28] (see Supplementary data,
Table F , available at Age and Ageing online). The European
Commission has recently created a key initiative that targets
the prevention of functional decline and frailty (European
Scaling-up Strategy in Active and Healthy Ageing) [
29] that
includes an action group focused on the prevention and early
diagnoses of frailty and functional decline (both physical and
cognitive) [
27]. Other European initiatives are the SPRINTT
Project (SPRINTT) [
28], a large clinical trial with the overall
goal of improving frailty care and prevention; and advocates
for use of the SHARE-FI, an instrument developed to iden-
tify frailty in Primary Care settings [
26].
M. T. E. Puts et al.
386
Downloaded from https://academic.oup.com/ageing/article/46/3/383/2863854 by guest on 20 August 2022
............................... ...................................... .........................................
Table 1. Characteristics of included participants
First author and
publication year
Country
of study
%
women
Frailty definition used in study Inclusion criteria Exclusion criteria Sample size and mean age of
participants
Binder 2002 [
14] USA CG 53%
IG
52%
Frailty based on 3 measures: (i) the PPT,
(ii) activities of daily living and
(iii) measurement of peak oxygen uptake
• ≥2 of 3 frailty criteria • Those who did not meet ≥2 frailty
criteria
• Those unable to exercise
N = 119, mean age 83, SD 4
Cameron 2013 [
22] Australia 68% The CHS criteria • Aged 70 years and over
• ≥3 CHS frailty criteria
• Unable to participate in programme
due to physical/cognitive health
N = 241, mean age 83.3, SD 6
Cesari 2015 [
13] USA 68.9 The CHS criteria • aged 70–85 years
• <10 on SPBB,
• able to walk 400 min in 15 min
• Severe/uncontrolled diabetes/
HTN/cardiac issues
N = 424, mean age 76.8, SD 4.2
Chan 2012 [
17] Taiwan 59% CCSHA_CFS_TV and the CHS frailty
criteria
• Community-dwelling older adults from
65 to 79 years of age in Toufen
Township (N = 6,828).
• CCHSA-CFS-TV score 3–6 (CHS pre-
frail and frail)
• Institutionalised
• Not frail
• Unable to participate in programme
due to physical/cognitive health
N = 117, mean age 71.4, SD 3.7
Gill 2003 [
15] USA 90% Physically frail which as determined by two
tests (rapid gait speed) and single chair
stand.
• Physically frail community-dwelling
people aged 75 years and over in
Southern Connecticut
• NH patients
• enroled in wellness programme
• Unable to participate in programme
due to physical/cognitive health
N = 188, mean age 83, SD 5 years
Gine-Garriga 2010 [
16] Spain 61% Physical frailty Gill and Tinetti definition or
self-reported exhaustion
• Aged 80–90
• Registered in a primary healthcare
centre in the Barcelona
• Unable to participate in programme
due to physical/cognitive health
N = 51, mean age IG 84.1 (SD 3)
and control 83.9 (SD 2.8)
Gustafsson 2012 [
23] Sweden 64% Adjusted CHS criteria • Community-dwelling
• Not receiving care
• Able to participate
• Unable to participate in programme
due to physical/cognitive health
N = 459, mean age control was 86,
preventative home visit mean age
86 and senior meetings mean 85
Kim 2015 [
18] Japan 100% The CHS criteria • Living in the Itabashi Ward of Tokyo,
Japan
• Women aged >73 years
• Unable to participate in programme
due to physical/cognitive health
N = 131, mean age 81, SD 3 years
Kwon 2015 [
19] Japan 100% Lowest 20% handgrip strength and lowest
quartile walking speed (2 of 5 CHS
criteria)
• Living in the Itabashi Ward, Tokyo,
Japan
• Pre-frail women aged ≥70 years
• Unable to participate in programme
due to physical/cognitive health
• Serum albumin >4.5 mg/dl and/or
taking CA/ Vitamin D supplements
N = 89, mean age 76.8 range (70–84)
Li 2010 [
21] Taiwan 48% The CHS criteria and BI • Living in 2 neighbourhoods in Taipei
• Aged ≥65 years
• Frail or pre-frail
• Unable to participate in programme
due to physical/cognitive health
N = 310, mean age IG is 78.4,
SD 8.2 and the CG is 79.3 SD 8.5
Ng 2015 [
5] Singapore 61% The CHS frailty criteria • Pre-frail or frail community-dwelling
older adults aged ≥65 years
• Able to ambulate without personal
assistance
• Unable to participate in programme
due to physical/cognitive health
N = 246, mean age 70.0, SD 4.7
Tarazona-Santabalbina
2016 [
20]
Spain 54% The CHS and Edmonton Frailty Scale • Frail older adult aged ≥70 years • Unable to participate in programme
due to physical/cognitive health
• Family member centenarian in past
2 generations
• No transportation
N = 100, mean age IG 79.7 (SD 3.6)
and control 80.3 (SD 3.7)
Continued
Interventions to prevent or reduce the level of frailty in community-dwelling older adults
387
Downloaded from https://academic.oup.com/ageing/article/46/3/383/2863854 by guest on 20 August 2022