Abstract
To determine the dose-response association between physical activity (PA) and all-cause mortality to identify minimum PA thresholds for significant mortality risk reduction by using a nationally representative sample of Chinese adults. This population-based cohort study included 109,407 Chinese participants from the Kailuan study with linkage to national death records up to December 31, 2022. Baseline data were derived from self-reported surveys among participants aged 53.32 ± 13.55 years. During a median follow-up of 6.99 (IQR 5.99–6.99) years, 4571 participants died from all causes. Performing at least 75 min per week of physical activity contributed to significantly reduced all-cause mortality. Compared with physical inactivity, total physical activity was inversely associated with the risk of mortality in the very low (HR: 1.08; 95% CI: 0.98–1.19), low (HR: 0.89; 95% CI: 0.81–0.98), medium (HR: 0.88; 95% CI: 0.79–0.96) and high (HR: 0.77; 95% CI: 0.69–0.85) quartiles. The dose‒response curve revealed a steady decline in HRs with more minutes of physical activity per week with all-cause mortality. Our study demonstrated an inverse nonlinear dose‒response relationship between physical activity and all-cause mortality in elderly adults, with as little as 75 min of weekly activity showing clinically meaningful risk reduction compared with inactivity. These findings suggest that mortality benefits at lower thresholds for elderly adults. The results provide evidence-based justification for health policies that promote graduated physical activity targets to improve population-level adherence among aging communities. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1038/s41598-025-26356-8.