Assessing Multiple Pollutants’ Impacts on Multiple Health Outcomes in New York City

Abstract

Background: Although air pollution is a mixture that includes local and regional emissions sources, local governments currently rely on single-pollutant-based U.S. national ambient air quality standards (NAAQS) to protect health. We developed a framework to assess multi-pollutants’ impacts on health outcomes in New York City (NYC) using time-series analyses. Methods: Single- and multi-pollutant Poisson models for ozone, nitrogen dioxide (NO2), and fine particles (PM2.5) were evaluated for multiple outcomes (morbidity models: 2005-2014; mortality models: 2004-2016): (1) asthma-related emergency department visits (ages 6-18); (2) cardiovascular-related hospitalizations (ages 40+); (3) hospitalizations for chronic lung diseases (ages 20+); and (4) cardiovascular (CVD) and all-natural-cause mortality (all ages). Seasonal models were adjusted for within-season trend, temperature, and day-of-week, and considered up to 3-day lagged effects. Attributable fractions (AFs) were estimated considering single and multiple pollutants when associations were found. Results: All outcomes were associated with one or more pollutants at levels below their current NAAQS. Individual pollutants’ estimates were sometimes sensitive (e.g., attenuation of risk estimates by 50%) to the addition of another pollutant, and total AFs in multi-pollutant models were typically smaller than the sum of those from single pollutant models. For hospitalization outcomes, AFs from NO2-only models were generally closest to those from corresponding multi-pollutant models (e.g., cold-season NO2-only models for lung diseases and CVD AFs (95% CI): 2.2% (-0.6, 5.0) and 3.3% (1.0, 5.6); NO2 + PM2.5: 2.2% (-1.9, 6.3) and 3.2% (-0.1, 6.5). Mortality outcomes were associated with all three pollutants in the warm season, with ozone-only models most indicative of multi-pollutant effects (e.g. all-cause mortality AFs: ozone only: 1.6% (0.1, 3.7); ozone + PM2.5 + NO2: 1.8% (-1.1, 5.1)).Conclusions: Given the substantial regional source contributions to PM2.5 and ozone, reducing local NO2 emissions (i.e., traffic and buildings) may be most impactful in reducing pollution-associated illness exacerbations in NYC.