The causal links between long-term exposure to major chemical components of PM2.5 and overall outpatient visits in mainland China: A nationwide study in the difference-in-differences framework

Abstract

Introduction

Although the adverse health effects of PM_2.5 exposure has been well documented, evidence of its adverse effect on overall outpatient visits was still limited. Besides, the adverse health effects of PM_2.5 exposure get complicated due to various components within the particles. So far, little is known about the relationship between PM_2.5 components and overall outpatient visits.

Objectives

This study aims to evaluate the causal relationships between long-term exposure to primary chemical components of PM_2.5 and outpatient visits, while estimating the mixture effect and relative contribution of the components.

Methods

Based on nationwide provincial-level surveillance data of outpatient visits in China and well-validated simulations of PM_2.5 components concentration, we employed the Difference-In-Differences (DID) approach to evaluate the causal relationships between long-term exposure to primary chemical components of PM_2.5 and outpatient visits, and used a Bayesian Weighted Quantile Sum (BWQS) regression to assess the mixture effect of the components.

Results

We found a 20.44% increase in the risk (IR%) of outpatient visits following each InterQuartile Range (IQR) increment in PM_2.5 concentration. Our estimation further suggested a 17.07%, 15.91%, and 14.04% increase in the risk of outpatient visits for organic matter, sulfate, and nitrate, but non-significant increases for other components. However, when considering the inter-components correlation, sulfate and black carbon contributed most (42.3% and 28.1%, respectively) to the overall mixture effect of PM_2.5 which was indicated by a 4.84% increase (95%CI: 1.92%, 7.83%) in the risk of outpatient visits following every unit increase in the overall BWQS index. Additionally, stratified analyses showed a stronger association among aged provinces and provinces with lower education rates.

Conclusion

Our findings would improve understanding of the individual and mixture impact of major chemical components of PM_2.5 and may contribute to more targeted and optimized environmental programs for pollution control.