Critical periods for long-term exposure to PM10 associated with incident cardiovascular diseases

Abstract

TPS 671: Cardiometabolic effects of long-term air pollution exposure, Exhibition Hall, Ground floor, August 26, 2019, 3:00 PM - 4:30 PM While increasing numbers of cohort studies have examined the association between long-term exposure to ambient particles and incident cardiovascular diseases (CVDs), there were little knowledge about an exposure period that critically affects the association. We aimed to compare four exposure periods to identify a critical period of exposure to particulate matter diameter less than or equal to 10 μm (PM10) associated with incident CVDs in a population-representative national cohort in South Korea. We used the National Health Insurance Service-National Sample Cohort (NHIS-NSC) that provides medical records, risk factors based on health screening, and annually-updated address information for one million people from 2002 to 2015 in South Korea. Our study population was restricted to 196,167 subjects who were aged 30-84 in 2007, underwent health screening for 2005-2007, and had not been diagnosed with any circulatory diseases for 2002-2006. Total CVDs consisted of ischemic heart disease, myocardial infarction, stroke, and heart failure, and the incidence was defined as the first diagnosis for 2007-2015. Using annual-average concentrations of PM10 at subjects’ district-specific addresses predicted by a previously-developed exposure prediction model, we calculated a long-term concentration of each subject for four periods: three periods for 1, 3, and 5 years prior to each CVD incidence and the fixed 5-year period for 2002-2006. We applied time-dependent and time-fixed Cox regressions to three time-varying exposures and one time-fixed exposure, respectively, and estimated adjusted hazard ratios (HR) of incident CVDs per a 10 μg/m3 increase in PM10. PM10 exposed for all three person-specific periods was marginally associated with incident CVD, while the 5-year exposure gave the highest HR (HR=1.04, 95% confidence interval=1.00-1.10). However, we did not find an association when early exposure for the fixed 5 years was used. This study suggests that recent exposure to PM is more responsible for incident CVD than early exposure.