Gait variability and fall risk in community-living older adults: A 1-year prospective study

Objective: To test the hypothesis that increased gait variability predicts falls among community-living older adults attending an outpatient clinic. Design: Prospective, cohort study. Setting: Three outpatient geriatric clinics. Participants: Fifty-two community-living, ambulatory men and women aged gre ater than or equal to 70 years. Interventions: Not applicable. Main Outcome Measures: Subjects walked at a normal pace for up to 6 minutes wearing force-sensitive insoles that measured the gait rhythm on a stride- to-stride basis. Afterward, subjects reported fall status on a weekly basis for 1 year. The primary outcomes were the association between measures of the stride-to-stride fluctuations in gait rhythm and (1) subsequent falls d uring a 12-month follow-up period and (2) potential contributing factors. Results: Almost 40% of the subjects reported falling during the 12-month fo llow-up period. Stride time variability was 106 +/- 30 ms in subjects who s ubsequently fell (n = 20) and 49 +/- 4ms in those who did not experience a fall (n = 32) during the 12-month follow-up period (p <.04). Logistic regre ssion also showed that stride time variability predicted falls (p <.05). St ride time variability correlated significantly with multiple factors includ ing strength, balance, gait speed, functional status, and even mental healt h, but these other measures did not discriminate future fallers from nonfal lers. Conclusions: These findings show both the feasibility of obtaining stride-t o-stride measures of gait timing in the ambulatory setting and the potentia l use of gait variability measures in augmenting the prospective evaluation of fall risk in community-living older adults.