Causal effect of PM1 on morbidity of cause-specific respiratory diseases based on a negative control exposure

Abstract

Background: Extensive studies have linked PM_2.5 and PM₁₀ with respiratory diseases (RD). However, few is known about causal association between PM₁ and morbidity of RD. We aimed to assess the causal effects of PM₁ on cause-specific RD.

Methods: Hospital admission data were obtained for RD during 2014 and 2019 in Beijing, China. Negative control exposure and extreme gradient boosting with SHapley Additive exPlanation was used to explore the causality and contribution between PM₁ and RD. Stratified analysis by gender, age, and season was conducted.

Results: A total of 1,183,591 admissions for RD were recorded. Per interquartile range (28 μg/m³) uptick in concentration of PM₁ corresponded to a 3.08% [95% confidence interval (CI): 1.66%-4.52%] increment in morbidity of total RD. And that was 4.47% (95% CI: 2.46%-6.52%) and 0.15% (95% CI: 1.44%-1.78%), for COPD and asthma, respectively. Significantly positive causal associations were observed for PM₁ with total RD and COPD. Females and the elderly had higher effects on total RD, COPD, and asthma only in the warm months (Z = 3.03, P = 0.002; Z = 4.01, P < 0.001; Z = 3.92, P < 0.001; Z = 2.11, P = 0.035; Z = 2.44, P = 0.015). Contribution of PM₁ ranked first, second and second for total RD, COPD, and asthma among air pollutants.

Conclusion: PM₁ was causally associated with increased morbidity of total RD and COPD, but not causally associated with asthma. Females and the elderly were more vulnerable to PM₁-associated effects on RD.

Keywords: Causal effect; Cause-specific respiratory diseases; Morbidity; PM(1).