Physical activity offsets air-pollution–related cognitive decline: A retrospective cohort study based on CHARLS

Abstract BACKGROUND This study examined whether physical activity (PA) buffers air-pollution-related cognitive decline in middle-aged and older adults, quantified the dose-response relationship, and derived pollution-specific PA recommendations. METHODS Data came from five waves (2011–2020) of the China Health and Retirement Longitudinal Study (CHARLS), including 12,196 adults aged ≥45 years. Ambient pollutants were estimated using a high-resolution satellite-based model. Linear mixed-effects models assessed main and interactive effects of PM2.5, PA, and PA × PM2.5 on cognition, stratified by socioeconomic status (SES) and residential setting. Isotemporal substitution and generalized additive models evaluated risk-benefit trade-offs and non-linearities. PA prescriptions were calculated using (PM2.5−25) × 1.316, with values ≤0 set to 0. RESULTS Higher PM2.5 exposure predicted poorer cognition (β = −0.0146, p < 0.001). PA buffered this effect (interaction β = 0.0344, p = 0.001), consistent across SES and residence. Among PM2.5 constituents, sulfate (β = −0.0136) and black carbon (β = −0.1059) were harmful. Vigorous PA neutralized the BC effect, while light-to-moderate PA offset the sulfate effect. Isotemporal substitution showed that 13.16 min/day of PA offset the cognitive impact of a 10 µg/m³ increase in PM2.5. Region-specific estimates required 10.92 min/day in Beijing and 4.01 in Shanghai, while Guangdong and Fujian required none. CONCLUSIONS Sulfate and BC are key drivers of PM2.5-related cognitive decline. Roughly 13 min of daily PA neutralizes the effect of each 10 µg/m³ PM2.5 rise. Light-to-moderate PA is preferable in sulfate-dominated areas, while vigorous PA is more effective in BC-dominated regions.