Pollutant-sex specific differences in respiratory hospitalization and mortality risk attributable to short-term exposure to ambient air pollution

Abstract

Background Many studies have reported associations of individual pollutants with respiratory hospitalization and mortality based on different populations, which makes it difficult to directly compare adverse health effects among multiple air pollutants. Objectives The study goal is to compare acute respiratory-related hospitalization and mortality associated with short-term exposure to three ambient air pollutants and analyze differences in health risks by season, age and sex. Methods Hourly measurements of air pollutants (ozone, NO2, PM2.5) and temperature were collected from ground-monitors for 24 cities along with daily hospitalization (1996–2012) and mortality (1984–2012) data. National associations between air pollutant and health outcome were estimated for season (warm, cold vs. year-round), age (base ≥ 1, seniors > 65), and sex (females ≥ 1 and males ≥ 1) using Bayesian hierarchical models. Results Overall, the three air pollutants were significantly associated with acute respiratory health outcomes at different lag-days. For respiratory hospitalization, the increased risks in percent changes with 95% posterior intervals for a 10-unit increase in each pollutant were: ozone (lag1, 0.7% (0.4, 0.9)), NO2 (lag0, 0.7% (0.1, 1.4)), and PM2.5 (lag1, 1.3% (0.7, 1.9)). For respiratory mortality: ozone (lag2, 1.2% (0.4, 1.9)), NO2 (lag1, 2.1% (0.6, 3.5)), and PM2.5 (lag1, 0.6% (−1.0, 2.2)). While some differences in risk were observed by season and age group, sex-specific differences were more pronounced. Compared with males, females had a higher respiratory mortality risk (1.8% (0.6, 2.9) vs 0.5% (−0.3, 1.3)) from ozone, a higher respiratory hospitalization risk (0.9% (0.0, 1.8) vs 0.6% (−0.3, 1.4)) but lower mortality risk (1.4% (−1.0, 3.7) vs 2.2% (0.4, 4.0)) from NO2, and a lower hospitalization risk (0.7% (−0.2, 1.7) vs 1.8% (1.0, 2.6)) from PM2.5. Conclusion This study reports significant health effects of short-term exposure to three ambient air pollutants on respiratory hospitalization (ozone≈NO2 < PM2.5 per-10 unit; ozone>NO2 ≈ PM2.5 per-IQR) and mortality (ozone≈NO2 > PM2.5) in Canada. Pollutant-sex-specific differences were found, but inconclusive due to limited biological and physiological explanations. Further studies are warranted to understand the pollutant-sex specific differences.