Abstract
Fine particulate matter (PM(2.5)) is a well-established health hazard, yet population-level causal evidence on the long-term effects of its chemical constituents and their interactions with environmental and socioeconomic factors remains scarce. This study leveraged quasi-experimental variation in PM(2.5) exposure across Guangdong province, China, during 2007-2018 to evaluate its causal impact on emergency department (ED) visits. We applied a Difference-in-Differences (DID) causal inference framework to obtain counterfactual estimates of long-term exposure effects and complemented this with generalized Weighted Quantile Sum (gWQS) regression to treat PM(2.5) as a complex mixture, quantify joint effects, and identify toxic components. The results showed that each interquartile increase in long-term PM(2.5) exposure was associated with a 10.2% rise in ED visits, with nitrate (weight = 0.299) and sulfate (0.294) contributing the most strongly, while organic matter exerted greater effects in less-developed regions. Temperature variation further modified these effects, with a 1 °C increase in average summer temperature associated with a 3.3% increase and a decrease in winter temperature linked to a 0.54% increase in constituent-related ED visits. Socioeconomic stratification revealed heterogeneous toxicity profiles across regions. These findings provide robust causal evidence on constituent-specific risks of PM(2.5), highlight the utility of integrating causal and mixture methods for complex exposures, and support targeted emission control and climate-adaptive strategies to protect vulnerable populations.