Association between Long-Term Exposure to Fine Particulate Matter Air Pollution and Mortality in a South Korean National Cohort

Abstract

Previous cohort studies in Europe and North America have reported the relationship between long-term exposure to fine particulate matter air pollution and human health. However, there have been few studies in other areas because of the lack of nationwide PM2.5 regulatory air quality monitoring data. This study examined the association between long-term exposure to PM2.5 and mortality in a South Korean national cohort. We included 275,337 of 20-65 years at baseline, who had survived and underwent health screening for 2002-2006 with follow-up data through 2013 from the National Health Insurance Service-National Sample Cohort. Individual exposures to PM2.5 were calculated as 5-year averages of district-specific annual average predictions of PM2.5 at annually-updated residential addresses for 2002-2006. PM2.5 predictions in 250 districts were estimated by using predicted district-specific annual means of PM10 multiplied by annual ratios of PM2.5 to PM10. We used Cox proportional hazards models to estimate hazard ratios (HRs) and their 95% confidence intervals (CIs) of non-accidental (The International Classification of Diseases, 10th revision: A00–R99), cardiovascular disease (I00–I99), and respiratory disease (J00–J99) mortality per 10 micrograms per cubic meter increase in PM2.5. In the models, we adjusted for individual covariates such as income, type of health insurance, smoking status and district-level demographic and socioeconomic variables. We did not find associations between PM2.5 and all mortality with statistically insignificant HRs. HR for respiratory mortality was higher than for cardiovascular mortality (HR: 1.48 [95% CI: 0.87, 2.51]; 0.97 [0.76, 1.25] for respiratory and cardiovascular disease mortality, respectively). Our study do not provide evidence for associations between long-term PM2.5 exposure and mortality. Future studies should investigate our findings of no associations by including additional confounders and/or subgroup analysis.