Dose-Response Relationships of Daily Steps With Disability Incidence and All-Cause Mortality Stratified by Age and Physical Frailty.

Objectives: To examine the optimal range of steps for an individual based on the dose-response relationship of the number of steps taken with disability incidence and all-cause mortality stratified by age and physical frailty.

Design: Prospective cohort study.

Setting and participants: 8664 community-dwelling older adults.

Methods: The daily number of steps was measured using an accelerometer. Disability incidence and mortality were prospectively determined over 60 months. Participants were stratified using a nonlinear restricted cubic spline based on age >75 or <75 years and physical frailty, per the revised Japanese version of the Cardiovascular Health Study criteria.

Results: The study cohort's median age was 74 years [interquartile range (IQR) range 71-78), and 54.0% were female. Incidental disability and death were observed in 1373 (15.8%) and 529 (6.1%) participants, respectively. The median steps per day were 5514 (IQR 3878-7616). Daily steps were nonlinearly associated with disability incidence and mortality. The optimal cutoff points for frail and nonfrail participants were, respectvely, 2168 [hazard ratio (HR) 0.74, 95% CI 0.56-0.98] and 7459 (HR 0.86, 95% CI 0.74-0.99) steps for disability incidence and 2593 (HR 0.63, 95% CI 0.40-0.98) and 3282 (HR 0.77, 95% CI 0.61-0.98) steps for all-cause mortality. The optimal cutoff points for participants >75 and <75 years were, respectively, 6066 (HR 0.83, 95% CI 0.72-0.99) and 8573 (HR 0.77, 95% CI 0.59-0.99) steps for disability incidence and 1824 (HR 0.67, 95% CI 0.46-0.98) and 4128 (HR 0.72, 95% CI 0.52-0.99) steps for all-cause mortality.

Conclusions and implications: Participants >75 years and frail participants required lower daily steps for preventing disability incidence and all-cause mortality than those <75 years and nonfrail participants, indicating that lower targets may still provide health-promoting benefits. Thus, the optimal step number should be considered based on individual characteristics, including age and frailty.