When air gets heavy: joint PM_2.5 and ambient humidity exposure, frailty, and stroke

PM_2.5 exposure has been associated with stroke, but evidence regarding ambient humidity, joint environmental exposure, and the mediating role of frailty remains limited.

We integrated data from the China Health and Retirement Longitudinal Study (CHARLS; n = 17,197) and a single-center hospital-based sample (n = 782). Annual average PM2.5 and relative humidity in the year before baseline or index date were assigned by residential location and dichotomized using database-specific medians. Joint exposure was categorized into four groups. Incident stroke in CHARLS was analyzed using Cox models, and stroke status in the hospital sample using logistic regression. Counterfactual mediation analysis evaluated frailty index as a mediator. Model-based mediation analyses evaluated frailty as a potential mediator, with the hospital-based analysis treated as exploratory.

In fully adjusted models, high PM_2.5 was associated with higher stroke risk in CHARLS (HR 1.184, 95% CI 1.083-1.295) and higher stroke odds in the hospital sample (OR 1.846, 95% CI 1.247-2.733). High humidity was also associated with stroke in CHARLS (HR 1.112, 95% CI 1.021-1.212) and the hospital sample (OR 1.928, 95% CI 1.301-2.859). Participants exposed to both high PM2.5 and high humidity had the highest estimates in CHARLS (HR 1.361, 95% CI 1.171-1.582) and the hospital sample (OR 2.431, 95% CI 1.390-4.251). Frailty mediated 18.1 and 22.3% of PM_2.5- and humidity-related associations in CHARLS, and 15.8 and 17.0% in the hospital sample.

Higher PM_2.5 and ambient humidity were independently and jointly associated with greater stroke risk. Frailty partly mediated these associations, highlighting a clinically relevant pathway linking environmental exposure to stroke.