Abstract
Several literatures have examined the risk of chronic respiratory diseases in association with short-term ambient PM(2.5) exposure in China. However, little evidence has examined the chronic impacts of PM(2.5) exposure on morbidity of chronic respiratory diseases in cohorts from high pollution countries. Our study aims to investigate the associations. Based on a retrospective cohort among adults in northern China, a Cox regression model with time-varying PM(2.5) exposure and a concentration-response (C-R) curve model were performed to access the relationships between incidence of chronic respiratory diseases and long-term PM(2.5) exposure during a mean follow-up time of 9.8 years. Individual annual average PM(2.5) estimates were obtained from a satellite-based model with high resolution. The incident date of a chronic respiratory disease was identified according to self-reported physician diagnosis time and/or intake of medication for treatment. Among 38,047 urban subjects analyzed in all-cause chronic respiratory disease cohort, 482 developed new cases. In CB (38,369), asthma (38,783), and COPD (38,921) cohorts, the onsets were 276, 89, and 14, respectively. After multivariable adjustment, hazard ratio and 95% confidence interval for morbidity of all-cause chronic respiratory disease, CB, asthma, and COPD were 1.15 (1.01, 1.31), 1.20 (1.00, 1.42), 0.76 (0.55, 1.04), and 0.66 (0.29, 1.47) with each 10 μg/m(3) increment in PM(2.5), respectively. Stronger effect estimates were suggested in alcohol drinkers across stratified analyses. Additionally, the shape of C-R curve showed an increasing linear relationship before 75.00 μg/m(3) concentrations of PM(2.5) for new-onset all-cause chronic respiratory disease, and leveled off at higher levels. These findings indicated that long-term exposure to high-level PM(2.5) increased the risks of incident chronic respiratory diseases in China. Further evidence of C-R curves is warranted to clarify the associations of adverse chronic respiratory outcomes involving air pollution.