Seasonal Analyses of Air Pollution and Mortality in 100 US Cities

TLDR

Short‑term air‑pollution effects on daily mortality are usually assumed constant, yet they may vary seasonally because of shifts in pollution sources and meteorology. The study developed Bayesian semiparametric hierarchical models to estimate time‑varying pollution–mortality effects in multisite time‑series analyses. The models were applied to the National Morbidity and Mortality Air Pollution Study data from 100 U.S. cities spanning 1987–2000.

Abstract

Time series models relating short-term changes in air pollution levels to daily mortality counts typically assume that the effects of air pollution on the log relative rate of mortality do not vary with time. However, these short-term effects might plausibly vary by season. Changes in the sources of air pollution and meteorology can result in changes in characteristics of the air pollution mixture across seasons. The authors developed Bayesian semiparametric hierarchical models for estimating time-varying effects of pollution on mortality in multisite time series studies. The methods were applied to the database of the National Morbidity and Mortality Air Pollution Study, which includes data for 100 US cities, for the period 1987–2000. At the national level, a 10-μg/m3 increase in particulate matter less than 10 μm in aerodynamic diameter at a 1-day lag was associated with 0.15% (95% posterior interval (PI): −0.08, 0.39), 0.14% (95% PI: −0.14, 0.42), 0.36% (95% PI: 0.11, 0.61), and 0.14% (95% PI: −0.06, 0.34) increases in mortality for winter, spring, summer, and fall, respectively. An analysis by geographic region found a strong seasonal pattern in the Northeast (with a peak in summer) and little seasonal variation in the southern regions of the country. These results provide useful information for understanding particle toxicity and guiding future analyses of particle constituent data.