Association of the interaction between daily step counts and frailty with disability in older adults

The differences in the association of objectively measured physical activity with disability between frail and non-frail older adults remain unclear. We (1) evaluated the dose-dependent relationship between daily steps and disability in older adults with and without frailty and (2) examined the interaction between steps and frailty status in relation to the risk of disability. This prospective study used data from 4065 adults aged ≥ 65 years from the Kyoto-Kameoka Study, Japan. The mean daily step count obtained using triaxial accelerometers across ≥ 4 days was recorded. Frailty was evaluated using the validated Kihon Checklist. Disability was identified using the long-term care insurance system's nationally unified database, with data collected until November 30, 2016. Overall, 385 disabilities were recorded during a median follow-up period of 3.32 years (12,855 person-years). After adjusting for confounders, an inverse association was observed between daily step count and disability risk. The disability risk plateaued at 5,000-7,000 steps/day in non-frail people, whereas step counts showed an almost linear inverse relationship with disability risk in frail people. Low step counts (< 5,000 steps) in frail people were more strongly associated with disability risk than were high step counts (≥ 5,000 steps) in non-frail people. The additive interaction between steps and frailty was associated with the relative excess risk of disability in frail people with low step counts (p for interaction = 0.015). The relationship between daily steps and disability differs between older adults with and without frailty, and the adverse effects of frailty on disability risk depend on physical activity.