Abstract
BACKGROUND: Previous studies have thoroughly evaluated the association between long-term exposure to PM2.5 and mortality, but the effect of socioeconomic status on the association remains controversial. METHODS: We utilized the data of all-cause mortality per year, concentration of PM2.5, socioeconomic and demographic characteristics for 345 cities in China from 2000-2019. We applied the Poisson generalized estimating equations (Poisson GEE) model to explore the association between PM2.5 and mortality in cities with different quartiles of demographic and socioeconomic characteristics. RESULTS: Overall, every 10 µg/m ³ increase in PM2.5, the RR was 1.082 (95% CI:1.058-1.106) for all-cause mortality, with residents of the Northeast and Southwest regions having a higher risk of death. Cities with the highest quartiles of gender ratio, age, dependency rate, and migration rate had relative risks of 1.110 (95% CI: 1.020-1.208), 1.075 (95% CI: 1.018-1.135), 1.101 (95% CI: 1.068-1.135), and 1.108 (95% CI: 1.053-1.165), respectively. Stratified analyses showed that the exposure-mortality associations were strongest in cities with both the lowest and highest socioeconomic status (SES) quartiles. CONCLUSION: The association between PM2.5 and mortality varied by socioeconomic status, demonstrating a U-shaped pattern with the highest mortality risk observed among residents in cities with both the lowest and highest socioeconomic levels. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12889-025-24521-2.