Among Health and Retirement Study participants aged 65 and older who had been persistently inactive and were free of mobility, ADL and IADL difficulty, those who began regular moderate or vigorous activity developed mobility difficulty less often at about two years: 32.75% versus 39.73%, a 6.98-percentage-point reduction (95% CI, -13.11 to -0.88). By four and six years the gaps shrank and the confidence intervals crossed zero, and ADL and IADL difficulty showed no clear differences. The analysis used sequential target trial emulation with stabilized inverse probability weighting and treated death as a competing event.
The design is the study's strength: emulating a trial and restricting to previously inactive, disability-free adults avoids the prevalent-user and reverse-causation problems that inflate many exercise-and-aging associations. The direction fits randomized evidence such as the LIFE trial, where structured activity reduced major mobility disability in sedentary older adults. Still, this remains observational. Weighting adjusts only for measured confounders, and self-reported activity is prone to error; people who start exercising may simply be healthier or more motivated in unmeasured ways. The fading effect at later waves is ambiguous, since it could reflect lapsed adherence, noninitiators starting later, or limited power. The wide intervals mean the true benefit could be modest. This is a preprint that has not been peer reviewed, so figures may shift. Editorially, it is confirmatory: a reasonable signal that starting late is worthwhile, not a quantified prescription.