Effectiveness of Exercise for Falls in Dementia: A Systematic Review

This report presents a systematic review and meta-analysis that investigates the effectiveness of exercise programs in reducing falls among older people with dementia living in the community. The study reviewed articles published between January 2000 and February 2014 from six electronic databases, focusing on randomized controlled trials and quasi-experimental trials. The findings suggest that exercise programs can potentially assist in preventing falls, with the exercise group showing a lower mean number of falls and a reduced risk of being a faller compared to the control group. The review highlights the need for further research with larger sample sizes, standardized measurement outcomes, and longer follow-up periods to inform evidence-based recommendations. The study emphasizes the importance of exercise as a falls prevention strategy for older adults with dementia, despite the limitations of the current research. The review included studies using PLWD in a residential care or hospital setting, four had a majority of participants living in the community, nine had a mix of older people living in both community and residential care, and three were unclear in describing the setting.

© 2015 Burton et al. This work is published by Dove Medical Press Limited, and licensed under Creative Commons
License. The full terms of the License are available at http://creativecommons.org/licenses/by-nc/3.0/. Non-commer
permission from Dove Medical Press Limited, provided the work is properly attributed. Permissions beyond the scope of the Licen
how to request permission may be found at: http://www.dovepress.com/permissions.php
Clinical Interventions in Aging 2015:10 421–434
Clinical Interventions in Aging Dovepress
submit your manuscript | www.dovepress.com
Dovepress
421
R e v I e w
open access to scientific and medical research
Open Access Full Text Article
http://dx.doi.org/10.2147/CIA.S71691
effectiveness of exercise programs to reduce
falls in older people with dementia living in th
community: a systematic review and meta-an
elissa Burton1,2
vinicius Cavalheri1
Richard Adams3
Colleen Oakley Browne4
Petra Bovery-Spencer4
Audra M Fenton3
Bruce w Campbell5
Keith D Hill1,6
1School of Physiotherapy and exercise
Science, Curtin University, Perth, wA,
Australia;2Research Department,
Silver Chain, Perth, wA, Australia;
3Community Services, west Gippsland
Healthcare Group, warragul, vIC,
Australia;4Falls Prevention for
People Living with Dementia Project,
Central west Gippsland Primary
Care Partnership, Moe, vIC, Australia;
5Allied Health, Latrobe Regional
Hospital, Traralgon, vIC, Australia;
6Preventive and Public Health
Division, National Ageing Research
Institute, Melbourne, vIC, Australia
Objective:The objective of this systematic review and meta-analysis is to evaluate the effec-
tiveness of exercise programs to reduce falls in older people with dementia who are living in
the community.
Method:Peer-reviewed articles (randomized controlled trials [RCTs] and quasi-experimental
trials) published in English between January 2000 and February 2014, retrieved from six elec-
tronic databases – Medline (ProQuest), CINAHL, PubMed, PsycInfo, EMBASE and Scopus –
according to predefined inclusion criteria were included. Where possible, results were pooled
and meta-analysis was conducted.
Results:Four articles (three RCT and one single-group pre- and post-test pilot study) were
included. The study quality of the three RCTs was high; however, measurement outcomes,
interventions, and follow-up time periods differed across studies. On completion of the interven-
tion period, the mean number of falls was lower in the exercise group compared to the control
group (mean difference [MD] [95% confidence interval {CI}] =-1.06 [- 1.67 to - 0.46] falls).
Importantly, the exercise intervention reduced the risk of being a faller by 32% (risk ratio [95%
CI] =0.68 [0.55–0.85]). Only two other outcomes were reported in two or more of the studies
(step test and physiological profile assessment). No between-group differences were observed
in the results of the step test (number of steps) (MD [95% CI] =0.51 [- 1.77 to 2.78]) or the
physiological profile assessment (MD [95% CI] =-0.10 [- 0.62 to 0.42]).
Conclusion:Findings from this review suggest that an exercise program may potentially assist
in preventing falls of older people with dementia living in the community. However, further
research is needed with studies using larger sample sizes, standardized measurement outcomes,
and longer follow-up periods, to inform evidence-based recommendations.
Keywords:cognitive impairment, older people, physical activity, fallers, community
dwelling
Introduction
Dementia is a major health issue predominantly affecting older people. It is estimated
that over 44 million people worldwide are living with dementia, and by 2050 there
may be as many as 135.5 million people diagnosed with dementia.1 The world popula-
tion is aging and as such it is expected that the increase in the number of older people
will correspond with an increase in the number of older people living with dementia
(PLWD). Dementia is a syndrome that impairs brain function and cognition. As the
severity of dementia increases over time, the person with dementia often has increased
difficulties with many important functions, including gait impairments, problems with
postural control, reduced participation in activities such as shopping and driving,
Correspondence: elissa Burton
Curtin University, School of Physiotherapy
and exercise Science, GPO Box U 1987,
Perth, wA 6845, Australia
Tel [phone]
Fax[phone]
email [e-mail]

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
422
Burton et al
and an increase in disability leading to difficulties in eat-
ing, bathing, and dressing.2,3 The impairments in cognition,
gait, and postural control also increase the risk of falls in
people with dementia. Approximately 30% of adults aged
65 years and over living in the community experience one or
more fall each year,4 but up to 50%–80% of PLWD fall in a
12-month period.5,6 There are many identified risk factors for
falls, including intrinsic factors such as postural instability
(gait and balance impairments), medications, neurocardio-
vascular complications, and vision impairment, as well as
extrinsic factors such as the environment (curbs, rugs, or poor
lighting).7 Falls can often lead to a fear of falling or loss of
confidence, which may result in a decline in activity and ulti-
mately a decrease in strength, balance, and mobility, leading
to decreased functional ability and a loss of independence.8,9
Falls are also often a trigger for emergency department or
hospital admission for older people with dementia 10 and/or
admission to residential care.10,11
Balance and mobility impairments in older people
have been shown to be a strong independent risk factor for
falling,12 and have been shown to decline at a significantly
faster rate in PLWD than age-matched older people without
cognitive impairment.13 To combat postural instability and
decreases in function, a large number of studies have been
conducted, investigating the effectiveness of exercise or
physical activity programs to prevent falls for older people
with a history of falling. 4,14 Reviews of these studies have
shown that strength- and balance-focused exercise programs
have been successful in decreasing the rate of falls for older
people living in the community with no cognitive impair-
ment, using both group- and home-based environments for
exercise. Based on these results, exercise or physical activity
programs are viewed as an important part of falls prevention
programs.3,4,15 However, direct translation of falls prevention
programs that have been shown to be effective in reducing
falls in samples with no cognitive impairment (eg, Close
et al16 multifactorial intervention) may not be effective when
implemented with people with cognitive impairment.17
Despite the higher risk of falls and greater rate of decline
of balance and gait in PLWD in the community, there has
been only a small but growing amount of research investigat-
ing the effect of exercise on improving physical performance
and reducing falls in people with dementia. There have been
nine systematic reviews14,18–25 and four general reviews7,17,26,27
investigating the effects of dementia/Alzheimer’s disease
on falls, and eight systematic reviews exploring the effects
of exercise (types of exercise) on people with cognitive
impairment. 3,28–34 Six of the 21 reviews reported earlier
included studies using PLWD in a residential care or hospital
setting, four had a majority of participants living in the
community, nine had a mix of older people living in both
community and residential care, and three were unclear
in describing the setting. Of the four reviews where older
people with dementia living in the community setting were
the majority sample population, one focused on medicine
and falls,21 another recruited participants living in the com-
munity from health care settings (emergency department,
dementia specific service, etc),25 the third looked at falls
risk with no specific emphasis on exercise,19 and the fourth
investigated the relationship between executive function,
falls, and gait abnormalities.32 There is a dearth of research
that specifically explored falls prevention exercise interven-
tions for people with dementia (cognitive impairment) living
in the community.
One systematic review by Hauer et al 35 did investigate
the effectiveness of physical training on motor performance
and fall prevention in cognitively impaired older persons
(search strategy was between 1966 and 2004). Again, of the
eleven randomized controlled trials (RCTs) included in this
review, nine were from a residential/long-term care/hospital
setting and one was of people living in the community, and
in the other the setting was unknown. Physical training was
not defined and appeared to represent strength and balance
programs designed by physiotherapists to be conducted in
either an individual or group setting.35 Physical activity
programs such as Tai Chi, which have been shown to be
effective in reducing falls, were not included.
In summary, to date the reviews published in this area are
limited when exploring those that have specifically focused
on older PLWD in the community with exercise and/or physi-
cal activity as the intervention, and with falls as the outcome
of interest. This systematic review seeks to address this gap.
The purpose of this review is to evaluate the available evi-
dence on the effectiveness of exercise or physical activity
programs to reduce falls in older people with dementia who
are living in the community.
Method
eligibility criteria
The review is limited to studies meeting the following eli-
gibility criteria:
• aged 60 years and over (at least 50% of the sample size);
• living in the community;
• PLWD or cognitively impaired. Dementia had to have
been identified by diagnosis by a doctor/specialist, or a
validated test, such as the Mini Mental State Examination

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
423
exercise to reduce falls for community dwelling people with dementia
(MMSE), the Clinical Dementia Rating Scale, or the
National Institute of Neurological and Communicative
Disorders and Stroke, Alzheimer’s Disease and Related
Disorders Association (NINCDS-ADRDA) Alzheimer’s
criteria;
• an exercise or physical activity program (intervention)
targeting a reduction in falls (and/or) risk of falls;
• outcome measures, which include number of falls, rate
of falls, or number of fallers, or time to first fall. Other
outcomes of interest were fear of having a fall, functional,
physical performance (eg, balance or mobility), or cogni-
tive benefits, or adherence to exercise/physical activity
intervention;
• study design: RCTs and quasi-experimental trials.
Information sources
Studies were identified by searching six databases (Medline
[ProQuest], CINAHL, PubMed, PsycInfo, EMBASE, and
Scopus, from January 2000 to February 2014). The search
strategy commenced from 2000, given a detailed review by
Hauer et al35 which searched across residential care, hospital,
and community settings, and did not identify any relevant
papers prior to 2003 in the community setting. In addition,
reference lists of the identified papers were scanned. Only
papers in English were included, no unpublished data,
books, conference proceedings, theses, or poster abstracts
were included.
Search strategy
The search was conducted using a mix of keywords to be
identified in the abstract and/or title of the paper or MESH
terms. The search strategy undertaken in Medline is presented
in Table 1. Each search was limited to English language
and the time period of January 1, 2000, to February 2014.
Language and syntax were adapted to individual databases:
for example, PubMed allowed title/abstract searches but not
all databases allowed this, so in these cases only the abstract
was searched.
Study selection
The study selection was conducted in three stages: stage
one involved the first author (EB) initially screening the
titles and scanning abstracts against the inclusion criteria to
identify relevant articles. This was followed (stage two) by
a full screening of the abstracts by EB. Stage three included
screening of the full articles by two of the authors (EB and
KH) to identify whether they met the eligibility criteria. Any
disagreements regarding potential inclusion were resolved by
discussion between EB and KH to achieve consensus, after
referring to the eligibility criteria and protocol.
The PRISMA checklist was used to ensure that the results
were reported systematically.36
Data collection process
Each study included in this review was evaluated, and the
following data were extracted: study design, purpose, inter-
vention, sample size, sex proportion, age of participants,
dropout rate of participants, MMSE or rating of dementia
score, number of fallers, time to first fall, fear of having a
fall, measures of balance, mobility, and function, intervention
effect, and length of follow-up.
Study quality
Methodological quality was assessed using the Cochrane
Collaboration’s risk of bias tool by two independent research-
ers (EB and KH).37 A third independent researcher completed
the risk of bias tool for one of the included studies because
of a conflict of interest for KH. 38 The categories assessed
were sequence generation, allocation concealment, blinding
of participants and staff, blinding of outcome assessment,
incomplete outcome data, selective outcome reporting, and
other sources of bias.37 Risk of bias was assessed to be “low
risk”, “unclear risk”, or “high risk” of bias.37
Data analysis
The studies are described according to their characteristics,
interventions utilized, outcome measures, adherence to
exercise interventions, study quality, and effectiveness of
the intervention programs.
Table 1 Search strategy (according to Medline terminology)
1 cognitive* impair* ti,ab.
2 cognition disorders/[MeSH]
3 dementia/[MeSH]
4 1 or 2 or 3
5 physical activity ti,ab.
6 physical active* ti,ab.
7 physical exerc* ti,ab.
8 exercise* ti,ab.
9 5 or 6 or 7 or 8
10 community* ti,ab.
11 home ti,ab.
12 10 or 11
13 fall* ti,ab.
14 accidental fall/[MeSH]
15 fall* prevent* ti,ab.
16 13 or 14 or 15
17 4 and 9 and 12 and 16

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
424
Burton et al
Three continuous outcomes (mean falls, step test, and
physiological profile assessment [PPA]) and one dichoto-
mous outcome (faller status [ie, faller versus non-faller]) were
included in the quantitative analyses. The mean difference
(MD) and 95% confidence intervals (CI) were calculated
for continuous outcomes, whereas risk ratio (RR) and 95%
CI were calculated for dichotomous outcomes. The Review
Manager (RevMan) version 5.2 was used to conduct the anal-
yses and generate the forest plots, and a fixed-effect model
was applied. Heterogeneity was assessed by the I 2 statistic
and by visual inspection of the forest plots. For continuous
outcomes, the results of homogeneous studies were subjected
to meta-analysis using the inverse variance DerSimonian and
Laird method.39 For faller status (ie, dichotomous outcome),
the results of homogeneous studies were meta-analyzed using
the Mantel–Haenszel’s fixed effects model.40 Two-sided
value of P , 0.05 was the statistically significant level set
for all analyses.
In instances where data provided in the published papers
were insufficient for the meta-analysis, the corresponding
authors for the RCT papers were contacted and asked for
the total number of falls pre and post intervention, the mean
number of falls per group at post intervention, the standard
deviation, the number of fallers per group (post intervention),
and the number of participants for both groups at pre and
post data collection.
Results
Study selection
The search strategy yielded 2,279 articles from six databases.
Duplicate articles within each database were removed,
leaving 446 articles. The 446 articles were then combined
into an excel spreadsheet, with duplicates again removed,
resulting in 286 remaining articles. Articles were then
screened on the basis of title, with 179 articles excluded (rea-
sons for exclusion are reported in Figure 1). The 107 articles
were then checked, and 79 articles were excluded based
on the abstracts. The full manuscripts of the 28 remaining
articles were then examined in detail, and 24 were found not
to meet the inclusion criteria. A total of four articles were
left for inclusion in the review. Three of these were RCTs
and were included in the meta-analyses.
Study characteristics
Three of the four articles included in the review were
RCTs.38,41,42 The fourth was a single-group pre- and post-test
pilot study.43 As presented in Table 2, sample sizes across the
studies ranged from 2242 to 210 participants.41 Three-hundred
and thirty-six participants completed pre-testing in the four
studies, but only 243 completed post-testing (72.3%). The
largest dropout rate was found for Mackintosh and Shep-
pard’s43 study, with half of the participants not continuing
across the 6-month study period. In contrast, Wesson et al42
only had one participant from the intervention group drop
out prior to completing the 12-week intervention (95.2%
completed the follow-up assessment).
All studies included MMSE scores of participants, with an
average score (and standard deviation) of 18.9 (5.5) across all
studies. There was little difference in MMSE scores between
the intervention groups (20.4 [5.1]) and controls (20.6 [5.0])
for the three RCTs.38,41,42 All studies reported the mean age
and standard deviation; the average age of the participants
was 79.8 (5.8) years.
Interventions
Interventions ranged between 3 months and 12 months, and
participants randomized to the intervention group were rec-
ommended to complete the exercises once a week, 43 twice
a week,41 three times a week, 42 or five times a week. 38 The
exercise interventions took place in a group (or individual
assistance where required) at a facility41,43 or at home.38,41,42
Pitkälä et al’s41 study comprised three groups: group-based
exercise, home-based exercise, and usual care (control). Two
studies provided multifactorial programs, which included an
exercise component.42,43 Other interventions included in the
two studies were foot health, medication management, vision
assessments, walking aids and footwear issues,43 and home
hazard reduction.42 The exercise programs predominantly
concentrated on strength, balance, and mobility, and these
programs were established and supervised by physiothera-
pists, occupational therapists, or physiotherapy students who
were trained and supervised by physiotherapists. In two of
the studies,38,42 carers were actively involved in monitoring
and encouraging participation between therapist visits for
the assigned exercise program.
Participants in Mackintosh and Sheppard’s43 study under-
took lower limb strength exercises (hip abductor, knee exten-
sor, and ankle dorsiflexion, bilaterally) using velcro ankle
weights; balance exercises while standing, and walking was
based on time or distance. The falls risk assessment assisted
the physiotherapist to develop individual falls and injury
prevention plans for each participant. Pitkälä et al41 had two
exercise groups (home-based and group-based) with different
programs. The home exercise group was given individually
tailored exercise programs provided by a physiotherapist with
specialist dementia training and the exercises addressed the

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
425
exercise to reduce falls for community dwelling people with dementia
Figure 1 Study selection flowchart.
Literature search
Databases:
Medline (N=255)
CINAHL (N=44)
PubMed (N=181)
PsycINFO (N=18)
EMBASE (N=119)
Scopus (N=1,662)
Total articles (N=2,279)
Duplicates removed
Databases:
Medline (N=64)
CINAHL (N=23)
PubMed (N=54)
PsycINFO (N=8)
EMBASE (N=47)
Scopus (N=250)
Total articles (N=446)
All articles combined into one
spread sheet and duplicates removed
(N=286)
Accepted
Articles screened on the basis
of title included (N=107)
Excluded
Reasons for exclusion:
Age: under 60 years (N=2)
Not dementia specific or majority (over 50%) (N=58)
No exercise intervention (N=31)
No falls outcome measure (N=21)
Not in English (N=11)
Not community dwelling (N=56)
Total articles excluded (N=179)
Accepted
Articles screened on the basis
of abstract included (N=28)
Excluded
Reasons for exclusion:
Not specific to eligibility criteria (N=1)
Not dementia specific/cognitively impaired (N=39)
No exercise intervention (N=24)
No falls outcome measure (N=5)
Not a peer reviewed article (N=1)
Not in English (N=1)
Not community dwelling (N=8)
Total articles excluded (N=79)
Accepted
Articles screened on the basis
of full text included (N=4)
Excluded
Reasons for exclusion:
Not dementia specific/cognitively impaired (N=7)
No exercise intervention (N=1)
Not peer review article (N=4)
Review: relevant articles already included (N=3)
Study protocol with no results (N=2)
Dwelling status unclear/proportion community dwelling
too low (N=2)
Residential care (N=2)
Not an intervention study (N=1)
No falls outcome measure (N=2)
Total articles excluded (N=24)

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
426
Burton et al
Table 2 Summary of included articles
Reference Study
design
Study purpose Intervention Sample size;
% female; age
(years) (SD);
drop out
MMSE score
(SD) or rating
of dementia
Number
of falls
Intervention
effect
Follow up
Mackintosh
and Sheppard43
Australia
Single-group
pre and post
test
To assess the
feasibility and
effectiveness of
preventing falls for
older people with
dementia from an
Italian background,
living in the
community
HLDR program, individualized falls and
injury management plan given to each
participant, including strength, balance and
mobility exercises, foot health, medication
management, vision assessment, footwear
issues, walking aids. They attended HLDR
once a week for 6 months
64 completed
pretest; 32.8%
(21/64) female;
79.6 (7.4) years;
32 completed
post-test
MMSe score 13.0
(7.4)
17 participants
had fallen in the
12 months prior,
and 12 participants
fell in the 6-month
study period
No significant
differences between
baseline and post-test
for number of fallers,
balance, cognitive
function (MMSe) or
aerobic capacity. 17
people at baseline
were classed as fallers
(previous 12 months)
but only 12 people
(38%) had falls over the
6-month study period
6 months
Pitkälä et al41
Finland
Randomized
controlled
trial
To investigate the
effects of intense
and long term
exercise on physical
functioning and
mobility of people
with Alzheimer’s
Disease living in the
community
Three-arm trial: 1) group-based exercise
(Ge: 4 hour visits, twice a week with
approx. 1 hour training each session); 2)
tailored home-based exercise (He: 1 hour
training), both twice a week for a year;
and 3) a control group (CG) receiving
usual community care
210 participants
randomized
into three equal
groups; 61.4%
average female;
78.03 (5.3) years;
49 participants
dropped out prior
to 12-month
follow-up (23.3%)
MMSe score
18 (6.3); 67.1%
suffered moderate
or severe AD
according to the
Clinical Dementia
Rating (CDR),
and 96% were
receiving AD
medication
No baseline falls
data provided. At
12 months there
were 355 falls in
total
He and Ge had
significantly fewer falls
over the 12 months;
all groups deteriorated
in functioning; CG
deterioration was
significantly faster than
the He or Ge groups
at 6 and 12 months
3, 6, and
12 months
Suttanon
et al38
Australia
Randomized
controlled
trial
To evaluate the
feasibility and safety
of a home-based
exercise program
for people with
AD, and provide
initial evidence
of intervention
effectiveness in
improving balance
and mobility and
reducing falls risk
Individualized home-based exercise
program supervised by a physiotherapist.
Balance and strength home exercise
group versus education/support
(controls). Included exercises and a
graduated walking program based on the
Otago exercise program, participants
completed exercises 5 times a week. Both
intervention and control group had six
home visits and five phone calls over the
6-month study
40 participants:
19 to exercise
group and 21
to control
(education)
group; 62.5%
female; 81.9
(5.72) years.
11/19 exercise
group completed
post testing,
three of the
control group
did not complete
follow-up
MMSe score
21.28 (4.58);
Intervention
group: 20.89
(4.74), control
group: 21.67
(4.43)
Falls rate of
exercise group
declined by 33%
for the 6-month
intervention,
control group
increased by~89%.
Similar pattern
for proportion
of fallers, though
neither of these
between-group
changes was
significantly
different
Significant improvement
by the intervention
group compared to
the control group for
functional reach, and
the Falls Risk for Older
People – Community
version. Trends
towards improvement
in the exercise group
for step test, modified
Clinical Test of Sensory
Interaction of Balance
and the Timed Up and
Go test
6 months

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
427
exercise to reduce falls for community dwelling people with dementia
wesson et al42
Australia
Randomized
controlled
trial
To explore the
design and feasibility
of a novel approach
to fall prevention
for people with mild
dementia living in the
community
Strength and balance training exercises
and home hazard reduction. Six
occupational therapy home visits (and
three phone calls), five home visits from
a physiotherapist. The physio-prescribed
and progressed the exercises. exercise
intervention consisted of a maximum of
six exercises from the weBB program,
including sit to stand, calf raises, step ups
onto a block, stands with diminishing base
of support (eyes open or closed), step-
overs, foot taps onto block, lateral side
steps and sideways walking. Participants
were asked to complete the exercises
three times per week. Control group
received usual care
22 participants
randomized into
two equal groups;
40.95% female;
79.8 (4.6) years; all
controls and ten
of the intervention
group completed
follow-up
MMSe score:
23.5 (3.7)
Falls during previous
12 months: 16 falls
in total. Intervention
group had fewer
falls (n=5) than
the control group
(n=11) during the
intervention
Risk of falling and
number of falls
was lower in the
intervention group;
however neither
was significant. No
significant differences
in physiological
outcome measures
between groups
due to small sample
size and incomplete
data primarily in the
intervention group at
follow-up
3 months
(12 weeks)
Notes: Data from: Mackintosh S, Sheppard L. A pilot falls-prevention programme for older people with dementia from a predominantly Italian background. Hong Kong Phys J. 2005;23:20–26.43 Pitkälä KH, Pöysti MM, Laakkonen ML, et
al. effects of the Finnish Alzheimer’s disease exercise trial (FINALeX). JAMA Intern Med. 2013;173(10):894–901. Copyright © 2013 American Medical Association. All rights reserved.41 Suttanon P, Hill KD, Said CM, et al. Feasibility, safety
and preliminary evidence of the effectiveness of a home-based exercise programme for older people with Alzheimer’s disease: a pilot randomized controlled trial. Clin Rehabil. 2013;27(5):427–438.38 wesson J, Clemson L, Brodaty H, et al.
A feasibility study and pilot randomised trial of a tailored prevention program to reduce falls in older people with mild dementia. BMC Geriatr. 2013;13:89. Published by BioMed Central.42
Abbreviations: HLDR, healthy lifestyle dementia respite; SD, standard deviation; MMSe, mini mental state examination; AD, Alzheimer’s disease; He, home exercise; Ge, group exercise; CG, control group.
participant’s needs and problems regarding daily functioning
and mobility.41 Exercises included climbing stairs, balance and
transfer training, walking, dual tasking, and outdoor activities.
Equipment included exercise bikes, ankle/hand weights, balls,
canes, and balance pillows. The group exercisers undertook
predetermined exercises consisting of endurance (exercise
bikes), balance (walking in a line, training while bouncing
a ball, climbing a ladder), and strength training (specialized
weights in the gym). All exercises increased in intensity and
dosage over time and were recorded. The group ratio was ten
participants to two physiotherapists.
Suttanon et al 38 based their home exercise regimen on
the Otago exercise program, which has been shown to
reduce falls for older people living in the community.44 The
exercise program included strengthening exercises, standing
balance exercises, and a walking program, and meets the
physical activity guidelines (30 minutes at least five times
a week) recommended by the World Health Organization
and the Australian Government.45,46 Ankle weights were
used to progress the strengthening exercises over time,
and each participant also received a booklet with illustra-
tions and instructions of the exercises. A physiotherapist
conducted six home visits over the 6-month intervention
(four in the first two months) 47 to prescribe individualized
exercises based on assessment findings and to ensure that
the carer and the person with dementia understood the
exercises and safety issues, and subsequent visits were used
to monitor, modify exercises if required, and motivate the
participant. The carer supervised the exercise program on
a day-to-day basis.
Wesson et al42 utilized strength and balance exercises from
the Weight-Bearing Exercise for Better Balance (WEBB)
program.48 Up to six exercises were individually prescribed
by a physiotherapist for each participant based on their physi-
cal performance assessment. Depending on the participant’s
cognitive level, they were fully supervised, or given a written
booklet and visual cues around the home with supervision
from a carer, or completed the exercises unsupervised.42
Exercises included sit to stand, calf raises, step-ups onto a
block, stances with reduced base of support (eyes open and
closed), step-overs, foot taps onto a block, lateral sidesteps,
and sideways walking. 42 The number and intensity of the
exercises were increased as the participant progressed.
Outcome measures
There were a wide range of measurements used across the
four studies. Falls were reported as number of falls, number
of people falling, and rate of falls. Falls risk was assessed

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
428
Burton et al
using the Falls Risk for Older People – Community version
(FROP-Com) questionnaire, and the PPA. Table 3 presents
the falls and physical measurement tools utilized by the four
studies, and also shows whether a significant difference
between the groups was found in any study (see asterisk).
Three of the four studies reported the number of falls, 41–43
and two the rate of falling.38,41 Apart from these and the
PPA38,42 and Hill Step Test, 38,42 no other (physical or falls
efficacy) measurements (or tools) were used in more than
one study.
Dropouts and adherence to exercise
interventions
Dropout rates varied between 4.5% (n=1)42 and 50% (n=32).43
Each study reported reasons for clients withdrawing, and
these included Mackintosh and Sheppard:43 admitted to resi-
dential care (n =11), died (n =7), did not complete program
(n=7), variety of reasons (n=7); Pitkälä et al:34 died (n=17),
admitted to residential care (n =18), impaired health (n =6),
spouse/caregiver died (n=1), declined to continue (n=7);
Suttanon et al:38 refused to continue (n=5), admitted into
residential care (n =4), hospitalized (n =1), died (n =1); and
Wesson et al:42 hospitalized (n=1).
Adherence to the exercise interventions varied between
not being reported43 and the three RCTs providing the per-
centage of participant involvement in the recommended dos-
age across the study periods. Over 80% (n=9) of the eleven
participants who completed the 6-month home exercise
intervention by Suttanon et al 36 completed the home-based
exercise program 5 days a week for 6 months. Two partici-
pants in this study were considered to have low adherence
(21.6% and 65.2%), which was due to an unwillingness to
participate and the presence of another health condition
preventing involvement.38
Pitkälä et al41 compared group and home-based exercise
programs to a control group and measured adherence as
participants completing at least 50% of the sessions. Almost
93% of home exercisers completed a minimum of half of
the sessions compared to 78.6% of the group exercisers.
The median number of sessions attended for each group
was as follows: home exercise: 81 (range 7–89), and group
exercise 75 (range 7–89). 41 The exercise group in Wesson
et al’s42 study participated an average of 2.8 times per week
(range 0–7) over 3 months, with almost three-quarters (n=8)
of the participants completing an average of 3.3 exercises
per session (range 2–5 exercises). Reasons for not participat-
ing were as follows: had not exercised in a long time prior
to the study, could not exercise without supervision, short
Table 3 Falls and physical measures of included studies
Study Number
of falls
People
falling
Falls
rate
FROP-
COM
BBS 6MW FIM SPPB FR Sit to
stand
TUG PPA LOS Hill step
test
Tandem FES-I ICONFES
Mackintosh et al41   
Pitkälä et al39    
Suttanon et al36        
wesson et al40       
Note:  denotes use by the study,  denotes statistical significance.
Abbreviations: FROP-COM, falls risk for older people – community version; BBS, Berg Balance Scale; 6Mw, 6-minute walk; FIM, functional independence measure; SPPB, short physical performance battery;
TUG, timed up and go; PPA, physiological profile assessment; LOS, limits of stability (movement velocity); Tandem, near tandem test of standing balance with eyes closed; FES-I, Falls Efficacy Scale – Interna
Iconographical Falls Efficacy Scale.

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
429
exercise to reduce falls for community dwelling people with dementia
periods of non-adherence during holidays, and brief periods
of illness.42
Quality of studies
The study by Mackintosh and Sheppard 43 had a number of
methodological weaknesses when using the risk bias tool due
to the study being a single-group pre-test, post-test trial rather
than an RCT methodology. The assessed potential bias in each
study is reported in Table 4. Pitkälä et al 41 and Suttanon et
al38 were assessed as having low risk of bias across all of the
domains. Wesson et al42 did differ slightly, as it was deemed
that attrition bias may have occurred, because, although the
authors acknowledged missing data, it may have affected the
direction of the results. The authors’ reasoning for not adjust-
ing for this in the data analysis was the “exploratory nature of
the pilot study”,42 and therefore bias may have occurred. All
three RCTs used intention-to-treat analysis. Overall, the three
RCTs were regarded as high-quality studies.38,41,42
effectiveness of intervention programs
Even though all of the groups deteriorated in function over
the year-long study in Finland, Pitkälä et al41 reported dete-
rioration being significantly slower in the exercise groups
(group-based exercise, home-based exercise) compared to
the control group, for the functional independence measure
(FIM; total and motor scores) at 6 months and 12 months, and
that the home- and group-exercise groups had significantly
fewer falls than the control group over 12 months. Suttanon
et al’s38 was the only other study to show a significant dif-
ference between the groups, and these were for the FROP-
Com (P=0.008), Functional Reach (P=0.002), and Limits of
Stability balance measure (movement velocity, P =0.016).
Mackintosh and Sheppard43 reported a reduction in the
number of people who fell (17 to 2) between pre and post
testing, and Wesson et al42 reported fewer falls in the exercise
group (n=5) compared to the control group (n=11), with only
two and four people having fallen from the intervention and
control groups, respectively.
None of the four studies published the mean scores for the
number of falls participants experienced during the intervention
period. On emailing the corresponding authors from the three
included RCTs for further falls data (mean pre and post number
of falls, standard deviation and sample size per group), all of
them responded.38,41,42 However, because of a significant differ-
ence between groups at baseline for proportion of falls, the study
by Suttanon et al38 was not included in the meta-analyses.
Figure 2 reports the forest plot for the mean number of
falls for the studies by Pitkälä et al 41 and Wesson et al. 42
Table 4 Assessment of risk of bias for included studies
Study Selection bias Performance bias Attrition bias Reporting bias Other bias
Sequence
generation
Allocation
concealment
Blinding of participants
and personnel
Incomplete
outcome data
Selective outcome
reporting
Free of
other bias
Mackintosh and Sheppard41 ● ● ● ● ○ ×
Pitkälä et al39 ○ ○ ○ ○ ○ ○
Suttanon et al36 ○ ○ ○ ○ ○ ○
wesson et al40 ○ ○ ○ ● ○ ○
Note: Bias was scored as low risk (○), unclear (×), or high risk (●).
Figure 2 Forest plot of comparison: exercise versus usual care for mean number of falls.
Abbreviations: SD, standard deviation; CI, confidence interval; IV, inverse variance.
Study or subgroup
Home-based exercise program
Subtotal (95% CI) 74 74 72.7% –1.07 (–1.78, –0.36)
Pitkälä et al, 201341 1.22 2.05 63 2.71 3.26 63 40.5% –1.49 (–2.44, –0.54)
–4 –2
Favors (exercise) Favors (control)
0 2 4
Wesson et al, 201342 0.45 1.03 11 1 1.48 11 32.2% –0.55 (–1.62, 0.52)
Heterogeneity:
χ2=1.66, df=1 (P=0.20); I 2=40%
Heterogeneity:
χ2=1.67, df=2 (P=0.43); I 2=0%
Test for subgroup differences:
χ2=0.00, df=1 (P=0.95); I 2=0%
Test for overall effect: Z=2.96 (P=0.003)
Test for overall effect: Z=3.44 (P=0.0006)
Group-based exercise program
1.68 3.29 60 2.71 3.26 63 27.3% –1.03 (–2.19, 0.13)Pitkälä et al, 201341
60 63 27.3% –1.03 (–2.19, 0.13)Subtotal (95% CI)
Heterogeneity: not applicable
Test for overall effect: Z=1.74 (P=0.08)
134 137 100.0% –1.06 (–1.67, –0.46)Total (95% CI)
Mean SD Total Mean SD Total Weight IV, fixed, 95% CI
Mean difference
IV, fixed, 95% CI
Mean differenceControlExercise

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
430
Burton et al
No overall heterogeneity was found between the two stud-
ies (I2=0%). On completion of the intervention period, mean
falls were statistically lower in the exercise group compared
to the control group (MD [95% CI]=-1.06 [- 1.67 to - 0.46]
falls). Subgroup analysis was conducted, showing that this
benefit was slightly greater for home-based exercise than
group-based exercise; however, the difference between
these subgroups was not statistically significant (
χ2=1.67,
P=0.43).
Figure 3 shows the forest plot for fallers versus non-
fallers. No overall heterogeneity was found between the
studies (I2=0%). On completion of the intervention period,
the exercise intervention reduced the risk of being a faller
by 32% (RR [95% CI] =0.68 [0.55–0.85]). Subgroup
analysis was conducted, showing no difference between
home-based and group-based exercise programs (
χ2=0.19,
P=0.91).
Comparison of two of the studies 38,42 was also possible
for the step test 49 (Figure 4) and PPA 50 (Figure 5). For the
step test, no heterogeneity was found between the two studies
(I2=0%). On completion of the intervention period, there was
no significant difference in the results of the step test (number
of steps) between the intervention and control groups (MD
[95% CI] =0.51 [- 1.77 to 2.78]). For the PPA, substantial
heterogeneity was detected between the two studies that
reported this outcome (I2=59%). However, the direction of
effects of the two studies was not different. On completion of
the intervention period, there was no significant difference in
the results of the PPA between the intervention and control
groups (MD [95% CI] =-0.10 [- 0.62 to 0.42]).
Discussion
The purpose of this review was to evaluate the available
evidence on the effectiveness of exercise or physical activity
Figure 3 Forest plot of comparison: exercise versus usual care for fallers versus non-fallers.
Abbreviations: CI, confidence interval; M-H, Mantel–Haenszel.
Study or subgroup
Events
Exercise Control
Total Events Total Weight
Risk ratio
M-H, fixed, 95% CI
Risk ratio
M-H, fixed, 95% CI
Home-based exercise program
Subtotal (95% CI)
Pitkälä et al, 201341 30
28 60
60
134 137 100.0%
0.05
Favors (exercise) Favors (control)
0.2 1 5 20
0.68 (0.55, 0.85)
43 63
63
47.2%
47.2%
0.68 (0.50, 0.94)
0.68 (0.50, 0.94)
4328
9060
32 47
Wesson et al, 201342 2
63
11
74
63
11
74
48.3%
4.5%
52.8%
0.70 (0.51, 0.95)
0.50 (0.11, 2.19)
0.68 (0.50, 0.92)
43
4
Heterogeneity:
χ2=0.19, df=1 (P=0.66); I 2=0%
Total events
Total events
Heterogeneity:
χ2=0.19, df=2 (P=0.91); I 2=0%
Test for subgroup differences:
χ2=0.00, df=1 (P=0.99); I 2=0%
Test for overall effect: Z=2.46 (P=0.01)
Test for overall effect: Z=3.40 (P=0.0007)
Group-based exercise program
Pitkälä et al, 201341
Subtotal (95% CI)
Heterogeneity: not applicable
Test for overall effect: Z=2.34 (P=0.02)
Total (95% CI)
Total events
Figure 4 Forest plot of comparison: exercise versus usual care for step test.
Abbreviations: SD, standard deviation; CI, confidence interval; IV, inverse variance.
Study or subgroup
Step test
Suttanon et al 2013 38
Wesson et al, 201342
12.28
15
3.03
5.12
11
6
17
11.81
14.2
3.52
7.67
18
8
26
88.5%
11.5%
100.0%
–10
Favors (exercise) Favors (control)
–5 0 5 10
0.47 (–1.95, 2.89)
0.80 (–5.91, 7.51)
0.51 (–1.77, 2.78)
17 26 100.0% 0.51 (–1.77, 2.78)
Subtotal (95% CI)
Heterogeneity:
χ2=0.01, df=1 (P=0.93); I 2=0%
Heterogeneity:
χ2=0.01, df=1 (P=0.93); I 2=0%
Test for overall effect: Z=0.44 (P=0.66)
Test for overall effect: Z=0.44 (P=0.66)
Test for subgroup differences: not applicable
Total (95% CI)
Mean SD Total Mean
ControlExperimental
SD Total Weight
Mean difference
IV, fixed, 95% CI
Mean difference
IV, fixed, 95% CI

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
431
exercise to reduce falls for community dwelling people with dementia
programs to reduce falls in older people with dementia
who are living in the community. Data from three RCTs
and one single pre- and post-test study with a total of
243 participants completing post testing were included.
The meta-analyses demonstrated that exercise programs
could potentially reduce falls for older people with dementia
living in the community. Caution should be taken when
interpreting the findings of our systematic review because
of the inconsistencies between the studies, small sample
sizes, as well as differences in outcome measures and the
interventions used.
Overall, the exercise interventions achieved a signifi-
cant reduction in the mean number of falls for home-based
exercise and a 32% reduction in the risk of being a faller
in the intervention group. The three RCTs included in the
review were of high quality, although some of the studies
were underpowered and follow-up limited, which has also
been reported as a problem in falls prevention studies for
older people without cognitive impairment living in the
community.4,51 Limited comparison or effect was found for
the other outcome measures used across the four studies.
However, exercise programs that incorporate strength,
balance and endurance, and progress in intensity over time
have been shown to reduce falls in older people without
cognitive impairment living in the community.15 This review
and the small number of studies that have investigated this
issue for older people living with dementia are a beginning to
building on this evidence base. It appears that similar exercise
interventions are likely to be beneficial in the reduction of
falls and number of people falling, for older people living in
the community with dementia. However, there may need to
be some modifications, such as those utilized by Suttanon et
al38 and Wesson et al42 including engagement of carers, regu-
lar contact by physiotherapist, more detailed exercise history
in order to tailor the exercises to participant preferences, and
a greater choice of exercises, which are challenging enough
but can be implemented safely to accommodate the different
cognitive and social circumstances.
Adherence to each exercise program ranged greatly
between the studies, and two of the RCTs38,42 provided
explicit reasons why exercise adherence was lower for some
participants. They included lack of exercise previously, ill
health, holidays, and the required support needed to exercise.
Suttanon et al52 conducted a detailed qualitative analysis,
interviewing the participants and caregivers as an exten-
sion of the Suttanon et al’s38 included RCT, and found that
preexisting or acute health conditions, a dislike of structured
exercise, and caregiver causes (health and unavailable)
were additional reasons for lower levels of adherence in the
study. These reasons are similar to those of other studies that
explored adherence to exercise for people with Alzheimer’s
disease and their caregivers.53,54 Teri et al54 found health issues
such as dizziness and muscle strain while exercising, health
problems for their caregivers, memory loss, confusion, and
agitated or paranoid behavior were reasons for non-adherence
in their study. When including people with dementia in future
studies, there is likely benefit in addressing some of these
factors in the implementation.
Another important finding with implications for future
research in this area is the wide variety of secondary mea-
sures used in the included studies, limiting the potential
for meaningful pooling of data and meta-analysis on these
measures. An international working group for falls preven-
tion studies has recommended in 2005 the need for improved
definitions and standardization of key measures between
fall prevention studies.43 These included a standard defini-
tion of a fall; daily recording of falls by participants, and a
monthly follow-up phone call or face-to-face visit to resolve
missing data and gain further falls and injury information;
falls data should include number of falls, number of fallers/
non-fallers/frequent fallers, fall rate per person year, and time
to first fall; data on injuries should be collected alongside falls
Study or subgroup
PPA
Suttanon et al 2013 38
Wesson et al, 201342
1.86
1.42
0.26
1.63
11
9
1.81
2.65
1.15
1.83
18
11
88.3%
11.7%
0.05 (–0.50, 0.60)
–1.23 (–2.75, 0.29)
–0.10 (–0.62, 0.42)
–0.10 (–0.62, 0.42)
20 29 100.0%
20 29 100.0%
–4
Favors (exercise) Favors (control)
–2 0 2 4
Subtotal (95% CI)
Heterogeneity:
χ2=2.41, df=1 (P=0.12); I 2=59%
Heterogeneity:
χ2=2.41, df=1 (P=0.12); I 2=59%
Test for overall effect: Z=0.38 (P=0.71)
Test for overall effect: Z=0.38 (P=0.71)
Test for subgroup differences: not applicable
Total (95% CI)
Mean SD Total Mean
ControlExperimental
SD Total Weight
Mean difference
IV, fixed, 95% CI
Mean difference
IV, fixed, 95% CI
Figure 5 Forest plot of comparison: exercise versus usual care for PPA.
Abbreviations: PPA, physiological profile assessment; SD, standard deviation; CI, confidence interval; IV, inverse variance.

Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
432
Burton et al
data, including peripheral fracture rate per person year of
follow-up, number of peripheral fractures, number of people
sustaining fractures, and number of people sustaining mul-
tiple events.51 Other outcome measures suggested included
the modified Falls Efficacy Scale (mFES)55 to measure the
psychological consequences of falling, the Short Form 12
(SF12) version 2, and the European Quality of Life Instrument
(EuroQoL EQ-5D) for Health Related Quality of Life.51 The
ProFaNE team also recommended a 12-month follow-up.
In the four studies included in this systematic review,
few of these recommendations were met. Only one study
completed a 12-month follow-up; 41 no study included
the mFES, EQ-5D, or SF12-v2; and only two studies
included a definition of a fall. 38,42 Recording of falls and
monthly follow-up were undertaken by all four studies
and, as mentioned previously, falls were also measured in
different ways across each study. These differences and
the studies choosing different outcome measures to those
recommended by ProFaNE may have been due to the tar-
get population being people with dementia and therefore
the four studies used tools that were tested to be valid and
reliable specifically for people with dementia. However,
further research is needed to confirm this and additional
recommendations for people with dementia living in the
community who fall.
Two of the studies in this review used a diagnosis-related
sample (Alzheimer’s disease), while the other two studies
included participants with cognitive impairment of any type.
Given the differing etiology, pathology, signs and symptoms,
and progression of disease between the various types of
dementia,56 and also the differing rates of falls,5 there is value
in future studies focusing on a specific diagnosis. However,
given the small number of studies meeting the review crite-
ria, we have included these studies with different sampling
together in our meta-analyses.
Limitations
The strength of this review is that a systematic methodology
was adopted to identify relevant studies and appraise the
methodological quality and effectiveness of the interventions
included and that meta-analyses were performed. However,
there are several limitations to this review. Firstly, although
six major databases were searched and strategies were used
to supplement these, there is always a possibility that rel-
evant studies were not included. The electronic search was
limited to January 2000 to February 2014, and it is possible
that studies published prior to this interval may be missing.
However, Hauer et al35 had similar inclusion criteria with no
limit to setting (ie, community living), and their search went
back to 1966. No further studies were found to meet our cri-
teria, and therefore we are confident that the chances of this
are slight. Because the search only included peer-reviewed
literature, there is also the possibility that publication bias
may have occurred because of unpublished, grey literature, or
organizational (falls prevention) reports not being included.
Language bias may have resulted because articles published
in languages other than English were not included in the
review; as fall prevention appears to be an issue facing many
older people throughout the world, relevant articles written
in languages other than English may have been excluded.
The lack of studies and common outcome measures across
the studies limit the ability to perform further meta-analyses,
and the sub-grouped meta-analyses are preliminary and there
is a need for further adequately powered trials to strengthen
interpretation of the results of this review. The small sample
sizes and shorter follow-up periods for some studies might
also have been a limitation. The quality of the three RCTs,
however, was high, which is a strength of the review,
although caution must be taken when drawing conclusions
because of the limitations discussed earlier.
Conclusion
This systematic review is the first of its kind to conduct
meta-analyses to investigate the effectiveness of exercise
interventions in reducing falls for older people with demen-
tia living in the community. There is promising evidence
that suggests falls prevention interventions, which include
strength, balance, and endurance/mobility training and prog-
ress in intensity over time can assist in the reduction of
falls in this population, although some caution is advised in
interpreting these findings. The program can be undertaken
either at home, in a group at a community center, or in a
gymnasium. Further research is required, particularly RCTs
using common outcome measures, standard follow-up peri-
ods of a minimum of 12 months, and appropriately powered
sample sizes in order to determine the most appropriate falls
reduction interventions for community-dwelling older people
with dementia.
Disclosure
The authors report no conflicts of interest in this work.
References
1. Alzheimer’s Disease International. Dementia Statistics; 2014. Available
from: http://www.alz.co.uk/research/statistics. Accessed March 27, 2014.
2. Australian Institute of Health and Welfare. Australia’s Health 2012.
Australia’s Health Series No. 13. Cat. No. AUS 156. Canberra: AIHW;
2012.

Clinical Interventions in Aging 2015:10 submit your manuscript | www.dovepress.com
Dovepress
Dovepress
433
exercise to reduce falls for community dwelling people with dementia
3. Suttanon P, Hill K, Said C, Dodd K. Can balance exercise programmes
improve balance and related physical performance measures in people
with dementia? A systematic review. Eur Rev Aging Phys Act. 2010;7:
13–25.
4. Gillespie L, Robertson M, Gillespie W, et al. Interventions for prevent-
ing falls in older people living in the community. Cochrane Database
Syst Rev. 2012;9:CD007146.
5. Allan L, Ballard C, Rowa E, Kenny R. Incidence and prediction of
falls in dementia: a prospective study in older people. PLoS One. 2009;
4(5):e5521.
6. Eriksson S, Gustafson Y, Lundin-Olsson L. Risk factors for falls in
people with and without a diagnose of dementia living in residential
care facilities: a prospective study. Arch Gerontol Geriatr. 2008;46:
293–306.
7. Shaw F. Prevention of falls in older people with dementia. J Neural
Transm. 2007;114:1259–1264.
8. Moschny A, Platen P, Klaaßen-Mielke R, Trampisch U, Hinrichs T.
Barriers to physical activity in older adults in Germany: a cross-sectional
study. Int J Behav Nutr Phys Act. 2011;8:121.
9. Rasinaho M, Hirvensalo M, Leinonen R, Lintunen T, Rantanen T.
Motives for and barriers to physical activity among older adults with
mobility limitations. J Aging Phys Act. 2006;15:90–102.
10. Rowe M, Fehrenbach N. Injuries sustained by community-dwelling
individuals with dementia. Clin Nurs Res. 2004;13(2):98–110.
11. Lord S. Predictors of nursing home placement and mortality of residents
in intermediate care. Age Ageing. 1994;23(6):499–504.
12. Deandrea S, Lucenteforte E, Bravi F, Foschi R, La Vecchia C, Negri E.
Risk factors for falls in community-dwelling older people: a systematic
review and meta-analysis. Epidemiology. 2010;21(5):658–668.
13. Suttanon P, Hill K, Said C, Dodd K. A longitudinal study of change in
falls risk and balance and mobility in healthy older people and people with
Alzheimer Disease. Am J Phys Med Rehabil. 2013;92(8):676–685.
14. Guo J, Tsai Y, Liao J, Tu H, Huang C. Interventions to reduce the number
of falls among older adults with/without cognitive impairment: an explor-
atory meta-analysis. Int J Geriatr Psychiatry. 2014;29(7):661–669.
15. Sherrington C, Tiedemann A, Fairhall N, Close J, Lord S. Exercise to
prevent falls in older adults: an updated meta-analysis and best practice
recommendations. N S W Public Health Bull. 2011;22(3–4):78–82.
16. Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C. Preven-
tion of falls in the elderly trial (PROFET): a randomised controlled trial.
Lancet. 1999;353:93–97.
17. Shaw F. Falls in older people with dementia. Geriatr Aging. 2003;6:
37–40.
18. Bunn F, Dickinson A, Simpson C, et al. Preventing falls among older
people with mental health problems: a systematic review. BMC Nurs.
2014;13(4):1–15.
19. Härlein J, Dassen T, Halfens R, Heinze C. Fall risk factors in older
people with dementia or cognitive impairment: a systematic review.
J Adv Nurs. 2009;65(5):922–933.
20. Jensen L, Padilla R. Effectiveness of interventions to prevent falls in
people with Alzheimer’s disease and related dementias. Am J Occup
Ther. 2011;65(5):532–540.
21. Kim D, Brown R, Ding E, Kiel D, Berry S. Dementia medications and
risk of falls, syncope, and related adverse events meta-analysis of ran-
domized controlled trials. J Am Geriatr Soc. 2011;59(6):1019–1031.
22. Meyer C, Hill S, Dow B, Synnot A, Hill K. Translating falls preven-
tion knowledge to community-dwelling older PLWD: a mixed-method
systematic review. Gerontologist. 2013. doi: 10.1093/geron/gnt127.
23. Muir S, Gopaul K, Odasso M. The role of cognitive impairment in fall
risk among older adults: a systematic review and meta-analysis. Age
Ageing. 2012;41:299–308.
24. Oliver D, Connelly JB, Victor CR, et al. Strategies to prevent falls and
fractures in hospitals and care homes and effect of cognitive impairment:
systematic review and meta-analyses. BMJ. 2007;334(7584):82.
25. Winter H, Watt K, Peel N. Falls prevention interventions for community-
dwelling older persons with cognitive impairment: a systematic review.
Int Psychogeriatr. 2013;25(2):215–227.
26. Perkins C. Dementia and falling. N Z Fam Physician. 2008;35(1):44–46.
27. Taylor M, Delbaere K, Close J, Lord S. Managing falls in older patients
with cognitive impairment. Aging Health. 2012;8(6):573–588.
28. Bowes A, Dawson A, Jepson R, McCabe L. Physical activity for people
with dementia: a scoping study. BMC Geriatr. 2013;13:129.
29. Christofoletti G, Oliani M, Gobbi S, Stella F. Effects of motor inter-
vention in elderly patients with dementia: an analysis of randomized
controlled trials. Top Geriatr Rehabil. 2007;23(2):149–154.
30. Forbes D, Thiessen E, Blake C, Forbes S. Exercise programs for
people with dementia. Cochrane Database Syst Rev. 2013;12:
CD006489.
31. Heyn P, Abreu B, Ottenbacher K. The effects of exercise training
on elderly persons with cognitive impairment and dementia: a meta-
analysis. Arch Phys Med Rehabil. 2004;85:1694–1704.
32. Kearney F, Harwood R, Gladman J, Lincoln N, Masud T. The relation-
ship between executive function and falls and gait abnormalities in
older adults: a systematic review. Dement Geriatr Cogn Disord. 2013;
36(1):20–35.
33. Littbrand H, Stenvall M, Rosendahl E. Applicability and effects of
physical exercise on physical and cognitive functions and activities of
daily living among people with dementia: a systematic review. Am J
Phys Med Rehabil. 2011;90:495–518.
34. Pitkälä K, Savikko N, Poysti M, Strandberg T, Laakkonen M. Efficacy
of physical exercise intervention on mobility and physical functioning in
older people with dementia: a systematic review. Exp Gerontol. 2013;48:
85–93.
35. Hauer K, Becker C, Lindemann U, Beyer N. Effectiveness of physical
training on motor performance and fall prevention in cognitively impaired
older persons: a systematic review. Am J Phys Med Rehabil. 2006;
85:847–857.
36. Liberati A, Altman D, Tetzlaff J, et al. The PRISMA statement for
reporting systematic reviews and meta-analyses of studies that evalu-
ate health care interventions: explanation and elaboration. BMJ. 2009;
339:b2700.
37. Higgins J, Altman D, Gøtzsche P, et al. The Cochrane Collabora-
tion’s tool for assessing risk of bias in randomised trials. BMJ. 2011;
343:d5928.
38. Suttanon P, Hill KD, Said CM, et al. Feasibility, safety and preliminary
evidence of the effectiveness of a home-based exercise programme for
older people with Alzheimer’s disease: a pilot randomized controlled
trial. Clin Rehabil. 2013;27(5):427–438.
39. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin
Trials. 1986;7(3):177–188.
40. Higgins J, Green S. Cochrane Handbook for Systematic Reviews of
Interventions Version 5.1.0 [updated March 2011]. The Cochrane Col-
laboration; 2011. Available from: www.cochrane-handbook.org.
41. Pitkälä KH, Pöysti MM, Laakkonen ML, et al. Effects of the Finnish
Alzheimer’s disease exercise trial (FINALEX). JAMA Intern Med. 2013;
173(10):894–901.
42. Wesson J, Clemson L, Brodaty H, et al. A feasibility study and pilot
randomised trial of a tailored prevention program to reduce falls in
older people with mild dementia. BMC Geriatr. 2013;13:89.
43. Mackintosh S, Sheppard L. A pilot falls-prevention programme for
older people with dementia from a predominantly Italian background.
Hong Kong Phys J. 2005;23:20–26.
44. Robertson M, Campbell AJ, Gardner M, Devlin N. Preventing injuries
in older people by preventing falls: a meta-analysis of individual-level
data. J Am Geriatr Soc. 2002;50(5):905–911.
45. World Health Organisation. Global Recommendations on Physical
Activity for Health 65 Years and Above; 2011. Available from: http://
www.who.int/dietphysicalactivity/physical-activity-recommendations-
65years.pdf. Accessed February 18, 2012.
46. Sims J, Hill K, Hunt S, Haralambous B. Physical activity recom-
mendations for older Australians. Australas J Ageing. 2010;29(2):
81–87.
47. Suttanon P. Balance and Falls Risk in Alzheimer’s Disease: Magni-
tude of the Problem and Feasibility and Effects of a Balance Exercise
Program. Bundoora, VIC: School of Physiotherapy, Faculty of Health
Sciences, La Trobe University; 2011.

Clinical Interventions in Aging
Publish your work in this journal
Submit your manuscript here: http://www.dovepress.com/clinical-interventions-in-aging-journal
Clinical Interventions in Aging is an international, peer-reviewed journal
focusing on evidence-based reports on the value or lack thereof of treatments
intended to prevent or delay the onset of maladaptive correlates of aging
in human beings. This journal is indexed on PubMed Central, MedLine,
CAS, Scopus and the Elsevier Bibliographic databases. The manuscript
management system is completely online and includes a very quick and fair
peer-review system, which is all easy to use. Visit http://www.dovepress.
com/testimonials.php to read real quotes from published authors.
Clinical Interventions in Aging 2015:10submit your manuscript | www.dovepress.com
Dovepress
Dovepress
Dovepress
434
Burton et al
48. Vogler C, Sherrington C, Ogle S, Lord S. Reducing risk of falling in
older people discharged from hospital: a randomized controlled trial
comparing seated exercises, weight-bearing exercises, and social visits.
Arch Phys Med Rehabil. 2009;90:1317–1324.
49. Hill K, Bernhardt J, McGann A, Maltese D, Berkovits D. A new test
of dynamic standing balance for stroke patients: reliability, validity,
and comparison with healthy elderly. Physiother Canada. 1996;48:
257–262.
50. Lord S, Menz H, Tiedemann A. A physiological profile approach to falls
risk assessment and prevention. Phys Ther. 2003;83(3):237–252.
51. Lamb S, Jørstad-Stein E, Hauer K, Becker C, on behalf of the Prevention
of Falls Network Europe and Outcome Consensus Group. Develop-
ment of a common outcome data set for fall injury prevention trials:
the Prevention of Falls Network Europe Consensus. J Am Geriatr Soc.
2005;53(9):1618–1622.
52. Suttanon P, Hill K, Said C, Byrne K, Dodd K. Factors influencing
commencement and adherence to a home-based balance exercise
program for reducing risk of falls: perceptions of people with
Alzheimer’s disease and their caregivers. Int Psychogeriatr. 2012;24(7):
1172–1182.
53. Teri L, Gibbons LE, McCurry SM, et al. Exercise plus behavioral man-
agement in patients with Alzheimer disease: a randomized controlled
trial. JAMA. 2003;290(15):2015–2022.
54. Teri L, McCurry S, Buchner D, et al. Exercise and activity level in
Alzheimer’s disease: a potential treatment focus. J Rehabil Res Dev.
1998;35(4):411–419.
55. Hill K, Schwarz J, Kalogeropoulos A, Gibson S. Fear of falling revisited.
Arch Phys Med Rehabil. 1996;77(10):1025–1029.
56. Bhidayasiri R. Atypical dementia: when it is not Alzheimer’s disease.
J Med Assoc Thai. 2007;90(10):2222–2232.