Background

Previous studies reported triggering of acute cardiovascular events by short-term increasedPM_2.5 concentrations. From 2007 to 2013, national and New York state air quality policies and economic influences resulted in reduced concentrations of PM_2.5 and other pollutants across the state. We estimated the rate of cardiovascular hospital admissions associated with increased PM_2.5 concentrations in the previous 1-7 days, and evaluated whether they differed before (2005-2007), during (2008-2013), and after these concentration changes (2014-2016).

Methods

Using the Statewide Planning and Research Cooperative System (SPARCS) database, we retained all hospital admissions with a primary diagnosis of nine cardiovascular disease (CVD) subtypes, for residents living within 15 miles of PM_2.5 monitoring sites in Buffalo, Rochester, Albany, Queens, Bronx, and Manhattan from 2005 to 2016 (N = 1,922,918). We used a case-crossover design and conditional logistic regression to estimate the admission rate for total CVD, and nine specific subtypes, associated with increased PM_2.5 concentrations.

Results

Interquartile range (IQR) increases in PM_2.5 on the same and previous 6 days were associated with 0.6%-1.2% increases in CVD admission rate (2005-2016). There were similar patterns for cardiac arrhythmia, ischemic stroke, congestive heart failure, ischemic heart disease (IHD), and myocardial infarction (MI). Ambient PM_2.5 concentrations and annual total CVD admission rates decreased across the period. However, the excess rate of IHD admissions associated with each IQR increase in PM_2.5 in previous 2 days was larger in the after period (2.8%; 95%CI = 1.5%-4.0%) than in the during (0.6%; 95%CI = 0.0%-1.2%) or before periods (0.8%; 95%CI = 0.2%-1.3%), with similar patterns for total CVD and MI, but not other subtypes.

Conclusions

While pollutant concentrations and CVD admission rates decreased after emission changes, the same PM_2.5 mass was associated with a higher rate of ischemic heart disease events. Future work should confirm these findings in another population, and investigate whether specific PM components and/or sources trigger IHD events.