Background: Stimulating activities are associated with a decreased risk of dementia. However, the extent to which this reflects a protective effect of activity or non-participation resulting from dementia is debated. We investigated the association of stimulating leisure-time activity in late adulthood with the risk of dementia across up to two decades' follow-up.

Methods: We used data from five prospective cohort studies from Finland and Sweden. Mental, social, outdoor, consumptive and physical leisure-time activities were self-reported. Incident dementia was ascertained from clinical diagnoses or healthcare and death registers. Cox regression was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).

Results: Of the 33 263 dementia-free individuals aged ≥50 years at baseline, 1408 had dementia during a mean follow-up of 7.0 years. Active participation in mental (HR: 0.52, 95% CI: 0.41 to 0.65), social (HR: 0.56 95% CI: 0.46 to 0.72), outdoor (HR: 0.70, 95% CI: 0.58 to 0.85), consumptive (HR: 0.67, 95% CI: 0.53 to 0.94) and physical (HR: 0.62, 95% CI: 0.51 to 0.75) activity, as well as variety (HR: 0.54, 95% CI: 0.43 to 0.68) and the overall frequency of activity (HR: 0.41, 95% CI: 0.34 to 0.49) were associated with a reduced risk of dementia in <10 years' follow-up. In ≥10 years' follow-up all associations attenuated toward the null.

Conclusion: Stimulating leisure-time activities are associated with a reduced risk of dementia in short-term but not long-term follow-up. These findings may reflect a reduction in leisure-time activity following preclinical dementia or dilution of the association over time.