Long-term follow-up of exercise interventions aimed at preventing falls in older people living in the community: a systematic review and meta-analysis.

BACKGROUND: Fall-related injuries are the leading cause of accident-related mortality for older adults, with 30% of those aged 65 years and over falling annually. Exercise is effective in reducing rate and risk of falls in community-dwelling adults; however, there is lack of evidence for the long-term effects of exercise. OBJECTIVES: To assess the long-term effect of exercise interventions on preventing falls in community-dwelling older adults. DATA SOURCES: Searches were undertaken on MEDLINE, EMBASE, AMED, CINAHL, psycINFO, the Physiotherapy Evidence Database (PEDro) and The Cochrane Library from inception to April 2017. STUDY SELECTION: Randomised controlled trials (RCTs), cohort studies or secondary analyses of RCTs with long-term follow-up (>12months) of exercise interventions involving community-dwelling older adults (65 and over) compared to a control group. DATA EXTRACTION/ DATA SYNTHESIS: Pairs of review authors independently extracted data. Review Manager (RevMan 5.1) was used for meta-analysis and data were extracted using rate ratio (RaR) and risk ratio (RR). RESULTS: Twenty-four studies (7818 participants) were included. The overall pooled estimate of the effect of exercise on rate of falling beyond 12-month follow-up was rate ratio (RaR) 0.79 (95% confidence interval (CI) 0.71 to 0.88) and risk of falling was risk ratio (RR) 0.83 (95% CI 0.76 to 0.92) Subgroup analyses revealed that there was no sustained effect on rate or risk of falling beyond two years post intervention. CONCLUSIONS: Falls prevention exercise programmes have sustained long-term effects on the number of people falling and the number of falls for up to two years after an exercise intervention. SYSTEMATIC REVIEW REGISTRATION NUMBER: CRD42017062461.