Frailty and Physical Activity: A Compositional Isotemporal Substitution Analysis from the All of Us Research Program.

Objectives: The objective of this study was to quantify the impact of shifting time away from sedentary behavior to physical activity on frailty in older adults. Methods: Participants from the All of Us Research Program with Fitbit data were included in the analysis. Fitbit data was used to measure time spent in sleep, sedentary behavior, light-intensity physical activity, and moderate-to-vigorous physical activity. Frailty was assessed using a 33-item frailty index. A compositional isotemporal substitution model was created to assess the estimated effects of substituting 30 minutes of sedentary behavior with an equal amount of time in physical activity on frailty. Results: Reductions in frailty index were seen by shifting 30 minutes from sedentary behavior to light activity (-0.003 [95% CI: -0.004, -0.002]) and moderate-to-vigorous activity (-0.016 [-0.017, -0.014]). Lower odds of frailty were also seen from shifting time from sedentary behavior to light activity or moderate-to-vigorous activity. Associations between physical activity and frailty were generally more pronounced in those participants who were frail or pre-frail than in those who were robust. Conclusions: Shifting time from sedentary behavior to physical activity is associated with lower odds of frailty, and shifting time to higher intensity activities may have a greater benefit to frailty.