Longitudinal Associations Between Falls and Risk of Gait Decline: Results From the Central Control of Mobility and Aging Study

Objective: To examine whether falls are associated with longitudinal changes in different gait domains and onset of clinical gait abnormalities. Design: Longitudinal study. Setting: General community. Participants: Ambulatory older adults free of dementia (N=428; mean age, 77.8±6.4 years). Interventions: Not applicable. Main Outcome Measures: Gait was assessed with a computerized walkway. Pace, rhythm, and variability (outcome measures) were derived from individual gait measures, using principal component analysis. Clinical gait abnormalities (neurologic, nonneurologic, mixed) were visually assessed by clinicians. Linear mixed-effects models were used to examine the associations between falls (the exposure variable coded as none, single, and multiple) and changes in gait domains. Multinomial logistic regression was used to examine associations between falls and the onset of clinical gait abnormalities. Models were adjusted for sex, education, age, body mass index, number of comorbidities, gait speed at the first follow-up, and time between the last fall and the first follow-up gait assessment. Results: Pace declined while rhythm and variability increased at a faster rate (P<.05) among 32 participants with multiple falls in the first year of follow-up compared with 299 participants with no falls. Risk for clinical gait abnormalities between those with no falls, a single fall, or multiple falls was not different. Conclusions: Multiple falls predict future gait decline in multiple domains in aging. Interventions to prevent gait decline after multiple falls should be investigated.