The Effect of Exercise on Falls in People Living with Dementia: A Systematic Review

Background: People living with dementia (PWD) are at a heightened risk for falls. However, the effects of exercise on falls in PWD are unclear. Objective: To conduct a systematic review of randomized controlled trials (RCTs) examining the efficacy of exercise to reduce falls, recurrent falls, and injurious falls relative to usual care among PWD. Methods: We included peer-reviewed RCTs evaluating any exercise mode on falls and related injuries among medically diagnosed PWD aged >= 55years (international prospective register of systematic reviews (PROSPERO) ID:CRD42021254637). We excluded studies that did not solely involve PWD and were not the primary publication examining falls. We searched the Cochrane Dementia and Cognitive Improvement Group's Specialized Register and grey literature on 08/19/2020 and 04/11/2022; topical categories included dementia, exercise, RCTs, and falls. We evaluated the risk of bias (ROB) using the Cochrane ROB Tool-2 and study quality using the Consolidated Standards of Reporting Trials. Results: Twelve studies were included (n = 1,827; age = 81.3 +/- 7.0 years; female = 59.3%; Mini-Mental State Examination = 20.1 +/- 4.3 points; intervention duration = 27.8 +/- 18.5 weeks; adherence = 75.5 +/- 16.2%; attrition = 21.0 +/- 12.4%). Exercise reduced falls in two studies [Incidence Rate Ratio (IRR) range = 0.16 to 0.66; fall rate range: intervention = 1.35-3.76 falls/year, control = 3.07-12.21 falls/year]; all other studies (n = 10) reported null findings. Exercise did not reduce recurrent falls (n = 0/2) or injurious falls (n = 0/5). The RoB assessment ranged from some concerns (n = 9) to high RoB (n = 3); no studies were powered for falls. The quality of reporting was good (78.8 +/- 11.4%). Conclusion: There was insufficient evidence to suggest that exercise reduces falls, recurrent falls, or injurious falls among PWD. Well-designed studies powered for falls are needed.