Gait Speed and Survival in Those With Exceptional Longevity: The LifeAfter90 Study

ABSTRACT Background Although gait speed is a well‐established predictor of mortality, it is understudied among ethnoracially diverse oldest‐old cohorts. We examined the association between gait speed and all‐cause mortality in an ethnoracially diverse 90+ population. Methods We examined the association between gait speed and survival over an average follow up of 3.6 ± 2.0 years. Follow‐up time was defined as years from baseline to death, loss to follow‐up, refusal, or study end, June 19, 2025, whichever occurred first. Baseline gait speed (m/s) was measured using the 4‐meter walk test. Cox proportional hazards model evaluated associations between gait speed and mortality, adjusting for age, sex/gender, race/ethnicity, and education. Interactions by sex/gender, age, race/ethnicity, and device were tested. Results Among 502 participants ( M age 92.9 ± 2.4 years; 62.6% women; 20.9% Asian, 22.3% Black, 12.4% Latinx, 8.6% Multiracial/Other, 35.9% White), mean gait speed was 0.55 ± 0.21 m/s and 259 deaths occurred. Each 1 SD (0.21 m/s) increase in gait speed was associated with a 20% lower risk of mortality (HR = 0.80; 95% CI = 0.69–0.91) after adjustments for demographics. In this cohort, this increase corresponds to completing the 4‐meter walk test about 2 s faster. The association was modified by age ( p ‐interaction = 0.01). There were no significant differences by sex/gender, race/ethnicity or device ( p ‐interaction > 0.1). In age‐stratified analyses, faster gait speed predicted lower risk of mortality among participants aged 90–94 years (HR = 0.70; 95% CI = 0.60–0.82), whereas the results were attenuated for the 95+ group (HR = 1.02; 95% CI = 0.77–1.30). Conclusion Faster gait speed was associated with a lower risk of mortality in adults aged 90 and older. These results were consistent across gender, race/ethnicity, and assistive device use, but differed by age group. Routine gait speed assessment may serve as a simple clinical tool to identify oldest‐old adults at higher risk of mortality.