Abstract
OBJECTIVE: Exercise is an evidence-based strategy for preventing falls. However, its efficacy may vary based on individual characteristics, like gait speed. This study examined whether baseline gait speed modified the effects of home-based exercise on subsequent falls among older adults. METHODS: This is a secondary analysis of a 12-month, randomized controlled trial in community-dwelling adults who were ≥70 years old and who had fallen within the previous 12 months. Participants were randomized to either 12 months of home-based exercise (n = 172) or standard of care (n = 172). This study examined intervention effects on fall rates at 6 and 12 months stratified by baseline gait speed (slow [<0.80 m/s] or normal [≥0.80 m/s]) using negative binomial regressions. Baseline gait speed was investigated as a potential modifier of the intervention effects on mobility and cognitive function using linear mixed modeling. RESULTS: At baseline, 134 participants had slow (exercise = 70; standard of care = 64) and 210 had normal (exercise = 102; standard of care = 108) gait speeds. For participants with slow gait speed, exercise reduced fall rates by 44% at 6 months (incidence rate ratio = 0.56; 95% CI [confidence interval] = 0.33-0.95) but not at 12 months (incidence rate ratio = 0.63; 95% CI = 0.38-1.03) compared with standard of care; for participants with normal gait speed, there was no significant effect of exercise on fall rates at 6 or 12 months. Gait speed modified intervention effects; in the exercise group, participants with slow gait showed significant improvements in the Timed "Up & Go" Test at 6 months (estimated mean difference = -4.05; 95% CI = -6.82 to -1.27) and the Digit Symbol Substitution Test at 12 months (estimated mean difference = 2.51; 95% CI = 0.81-4.21). CONCLUSION: Older adults with slow gait speed had a reduction in subsequent falls in response to exercise at 6 months. Gait speed modified the effects of exercise on mobility and cognition. IMPACT: Older adults with slow gait speed may be a target population for exercise-based fall prevention.