Active later life tracks with longer, sharper years

Rates of cognitive impairment and dementia are rising, so the search for modifiable factors that extend healthy years is pressing. Plenty of research already links physical activity to a later onset of cognitive impairment, but that is a different question from whether being active once you are already old still changes the trajectory. Cognitive ageing is not a single event: people move between being unimpaired, mildly impaired and severely impaired, and they die at different rates from each of those states. Knowing where activity acts along that path matters for how it is worth framing to older adults.

The researchers used a coordinated analysis approach, applying the same modelling strategy independently to 14 longitudinal studies from North America and Europe with a combined sample of 52,039 people, whose mean baseline ages across studies ranged from 69.9 to 81.73 years. Multistate survival models estimated how physical activity related to transitions between three cognitive status categories — nonimpaired, mildly impaired and severely impaired — and to death. Multinomial regression models estimated life expectancy according to American physical activity recommendations. Results from each study were then pooled by meta-analysis, with models controlling for baseline age, sex, education and chronic conditions.

More physical activity was significantly associated with a reduced risk of transitioning from nonimpaired to mildly impaired cognitive functioning, a reduced risk of death, and substantially longer estimated life expectancy. There was also evidence of a protective role after impairment had begun: lower risk of moving from mild to severe impairment, and a greater likelihood of moving back from severe to mild impairment. However, the authors flag considerable between-study heterogeneity for those post-onset transitions, which makes that part of the picture less robust than the earlier-stage findings.

The result extends an established literature on activity and delayed onset of impairment into the question of progression, suggesting the relationship is not confined to preventing the first step. Because the underlying data are observational cohorts rather than trials, reverse causation remains a live concern: early cognitive changes can themselves reduce how much people move. The variability between studies also leaves open how much of the post-onset benefit is real and how much reflects differing populations and activity measures. A randomised test of activity in people who already have mild impairment would be the natural next step.

This is a reason to think that physical activity remains relevant to cognitive health late in life, not just in midlife as prevention. The strongest part of the evidence concerns staying unimpaired for longer and living longer, so those are the findings to weight most heavily. The suggestion that activity relates to slower progression, and even to moving back from severe to mild impairment, is intriguing but inconsistent across the studies pooled here. None of it establishes cause, and the analysis says nothing about which type or amount of activity would produce a given result.