Effectiveness of exercise programs to reduce falls in older people with dementia living in the community: a systematic review and meta-analysis

Figures & data

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

Figure 1 Study selection flowchart.

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 Sheppard^Citation43 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 al^Citation41 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 al^Citation38 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
Wesson et al^Citation42 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.

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 al^Citation41	✓				✓	✓
Pitkälä et al^Citation39	✓★		✓★				✓★	✓
Suttanon et al^Citation36			✓	✓★					✓★	✓	✓	✓	✓★	✓
Wesson et al^Citation40	✓	✓										✓		✓	✓	✓	✓

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; FR, functional reach; 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 – International (short form); ICONFES, Iconographical Falls Efficacy Scale.

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 Sheppard^Citation41	●	●	●	●	○	×
Pitkälä et al^Citation39	○	○	○	○	○	○
Suttanon et al^Citation36	○	○	○	○	○	○
Wesson et al^Citation40	○	○	○	●	○	○

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.

Figure 3 Forest plot of comparison: exercise versus usual care for fallers versus non-fallers.

Abbreviations: CI, confidence interval; M-H, Mantel–Haenszel.

Figure 4 Forest plot of comparison: exercise versus usual care for step test.

Abbreviations: SD, standard deviation; CI, confidence interval; IV, inverse variance.

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.

MackintoshSSheppardLA pilot falls-prevention programme for older people with dementia from a predominantly Italian backgroundHong Kong Phys J2005232026

 Google Scholar

PitkäläKHPöystiMMLaakkonenMLEffects of the Finnish Alzheimer’s disease exercise trial (FINALEX)JAMA Intern Med20131731089490123589097

 PubMed Web of Science ®Google Scholar

SuttanonPHillKDSaidCMFeasibility, safety and preliminary evidence of the effectiveness of a home-based exercise programme for older people with Alzheimer’s disease: a pilot randomized controlled trialClin Rehabil201327542743823117349

 PubMed Web of Science ®Google Scholar

WessonJClemsonLBrodatyHA feasibility study and pilot randomised trial of a tailored prevention program to reduce falls in older people with mild dementiaBMC Geriatr2013138924004682

 PubMed Web of Science ®Google Scholar

DerSimonianRLairdNMeta-analysis in clinical trialsControl Clin Trials1986731771883802833

 PubMedGoogle Scholar

LiberatiAAltmanDTetzlaffJThe PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaborationBMJ2009339b270019622552

 PubMed Web of Science ®Google Scholar

HigginsJGreenSCochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]The Cochrane Collaboration2011 Available from: www.cochrane-handbook.org

 Google Scholar