Objectives: To examine the association between poorer performance on concurrent walking and reaction time and recurrent falls.
Design: Cross-sectional analysis.
Setting: Community.
Participants: Three hundred seventy-seven older community-dwelling adults (mean age+/-standard deviation 78+/-3).
Measurements: Reaction times on push-button and visual-spatial decision tasks were assessed while seated and while walking a 20-m course (straight walk) and a 20-m course with a turn at 10 m (turn walk). Walking times were recorded while walking only and while performing a reaction-time response. Dual-task performance was calculated as the percentage change in task times when done in dual-task versus single-task conditions. A history of recurrent falls (> or = 2 vs < or = 1 falls) in the prior 12 months was self-reported. Multivariate logistic regression models were used to predict the standardized odds ratios (ORs) of recurrent falls history. The standardized unit for dual-task performance ORs was interquartile range/2.
Results: On the push-button task during the turn walk, poorer reaction time response (slower) was associated with 28% lower (P=.04) odds of recurrent fall history. On the visual-spatial task, poorer walking-time response (slower) was associated with 34% (P=.02) and 42% (P=.01) higher odds of recurrent falls history on the straight and turn walks, respectively.
Conclusion: These findings suggest that walking more slowly in response to a visual-spatial decision task may identify individuals at risk for multiple falls. Prospective studies are needed to confirm the prognostic value of poor walking responses in a dual-task setting for multiple falls.