Physical activity and Alzheimer’s disease risk across genetic susceptibility: a prospective UK Biobank study using accelerometer data

Objective Physical activity (PA) has been associated with reduced Alzheimer’s disease (AD) risk, but whether protective effects vary across genetic risk levels remains unclear. Previous studies were limited by self-reported PA measures and simplified genetic models. In this study, we aimed to examine the association between accelerometer-measured physical activity and the risk of incident AD in a large population-based cohort, and to explore potential interactions between PA and polygenic risk scores for AD. Methods We analyzed 93,578 UK Biobank participants aged 40–70 years with accelerometer data and genome-wide genotyping. PA was measured continuously (milligravity, mg) and dichotomized at the optimal point from maximally selected rank statistics. Genetic risk was assessed using polygenic risk scores (PRS) and APOE ε4 status. Cox models estimated hazard ratios for incident AD across genetic risk strata during median 15.5-year follow-up. Results Among 401 AD cases, high PA reduced risk by 48% (HR 0.517; 95% CI 0300–0.891), while high PRS increased risk nearly twofold (HR 2.423; 95% CI 1.757–3.343). PA’s protective association remained consistent across all PRS and APOE ε4 strata. No significant multiplicative or additive interaction was found between PA and genetic risk (RERI = − 0.566, 95% CI − 4.574–3.441). Dose–response analysis revealed maximum benefit with optimal threshold at 21.7 mg corresponding to light-intensity activity. Conclusion Objectively measured PA substantially reduces AD risk regardless of genetic predisposition. Even light-intensity activity provides meaningful protection, supporting PA as a broadly applicable preventive strategy across all genetic risk levels.