Background: Physical activity (PA) interventions may reduce the burden of frailty and can prevent mobility disability for older adults. We explored whether a 2-year PA intervention would improve frailty trajectory, lead to clinically meaningful frailty changes (CMC), or impact major mobility disability (MMD) across baseline frailty levels.

Methods: We analyzed data for 1635 community-dwelling participants who were 70-89 years old (mean baseline age [SD]: 78.9 [5.2] years, 67.2% female) from the Lifestyle Interventions and Independence Study. Participants were randomized to either PA or health education (HE) intervention. A 44-item frailty index (FI) was constructed at baseline and 0.5, 1, 1.5, and 2 years after baseline. CMC was defined as change in FI of ≥0.03. MMD was the inability to complete a 400 m-walk within 15 min without assistance. Mixed-effects models were used to estimate frailty trajectory and CMC. Cox regression models were used to determine whether the effect of PA on the composite of MMD or death differed by baseline FI.

Results: Mean FI (SD) at baseline for both the PA and HE groups was 0.18 (0.10). Two years after baseline, mean FIs were 0.23 (0.12) for PA and 0.24 (0.12) for HE. The MMD rates were 30.1% (246/818) and 35.5% (290/817) for PA and HE, respectively. There was no time-by-intervention interaction for frailty trajectory or for CMC. Regarding the composite MMD and death, there was no FI-by-intervention interaction. Simple association analyses revealed that when baseline FI was centered at 0.15 or higher, the PA intervention was associated with lower risk of MMD or death compared to HE (HR [CI] range for FI ≥ 0.15: 0.65-0.81 [0.43-0.67, 0.90-0.98]).

Conclusion: Participants in both groups showed similar frailty trajectories and CMC. Those who were frailer benefitted more from the PA intervention regarding MMD and death and may be a focus of recruitments for future PA program.