Domain-specific physical activity levels and risk of dementia: a systematic review and meta-analysis.
Level 3 - non-randomized controlled study
Systematic review and meta-analysis of observational cohort and case-control studies
PubMed 42612650 · doi:10.1016/S2468-2667(26)00142-8
What was done
A systematic review and multilevel random-effects meta-analysis was conducted using six databases (through May 2, 2026, combined with an earlier review through October 2021) for cohort and case-control studies. Eligible studies included adults aged 20 years or older with baseline cognitive assessment, physical activity measures, and at least 1 year of follow-up. The authors examined associations between total and domain-specific physical activity (leisure-time, occupational, household, and commuting) and risk of all-cause dementia, Alzheimer's disease, and vascular dementia across 74 studies (17 new, 57 from prior review) comprising 4,227,297 participants (52.8% female, median age 67.3 years, median follow-up 10.0 years).
What was found
High overall physical activity was associated with reduced risk of all-cause dementia (146 effect sizes; RR 0.80, 95% CI 0.75-0.86; I² = 97.4%), Alzheimer's disease (59 effect sizes; RR 0.79, 95% CI 0.70-0.90; I² = 91.3%), and vascular dementia (21 effect sizes; RR 0.71, 95% CI 0.58-0.87; I² = 77.5%). By domain, all-cause dementia risk was lower with leisure-time (20 studies, 33 effect sizes; RR 0.76, 95% CI 0.65-0.88; I² = 96.6%) and household activity (1 study; RR 0.85, 95% CI 0.74-0.98), but higher with occupational (5 studies; RR 1.20, 95% CI 1.01-1.42; I² = 85.6%) and commuting activity (2 studies; RR 1.10, 95% CI 1.03-1.18; I² = 0.0%).
Why it matters
This meta-analysis highlights a domain-dependent physical activity paradox for dementia risk, indicating that protective effects are concentrated in leisure and household domains while occupational and commuting activity may carry increased risk.
Limits
Certainty of evidence was rated very low to low across domains. Heterogeneity was high for total, leisure, and occupational activity (I² > 85%). Only 54% of studies were moderate or high quality, 93% were conducted in high-income countries, and non-leisure domains were supported by very few studies (k=5 occupational, k=2 commuting, k=1 household).
Cited by
- supports Physical activity and exercise significantly reduce the risk and assist in the prevention of dementia and Alzheimer's disease.