Multi-City Morbidity Study

Title: A Five-City Time-Series Study of Pollutant Mixtures and Acute Morbidity

Investigators: Stefanie Sarnat, Sc.D. (PI, Emory), Lyndsey Darrow, Ph.D. (Emory), Mitchel Klein, Ph.D. (Emory), Paige Tolbert, Ph.D. (Emory), Andrea Winquist, M.D., Ph.D. (Emory), James Mulholland, Ph.D. (Georgia Tech), and Ted Russell, Ph.D. (Georgia Tech)

Objectives.  Although associations between ambient air pollution and acute cardiorespiratory outcomes have been observed in numerous studies, questions remain about the degree to which these findings are generalizable between locations and whether the observed health effects are due to the individual pollutants measured or to pollutants acting in combination with other pollutants. We will conduct a multi-city time-series study to clarify the impacts of air quality on acute cardiorespiratory morbidity in five US cities (Atlanta, GA; St. Louis, MO-IL; Dallas, TX; Birmingham, AL; and Pittsburgh, PA) using novel mixture characterization (MC) metrics. Our overarching hypothesis is that factors related to air pollution mixtures, seasonality and climate, concentration-response functions, exposure measurement error, and population susceptibility and vulnerability can help explain apparent between-city heterogeneity in short-term associations between air quality measures and cardiorespiratory emergency department visits and hospital admissions.

Approach.  Individual-level morbidity data for each city will be acquired from existing in-house databases and limited new collection under the current proposal, resulting in 6.5 to 17 years of data for each outcome type across the cities. Rich air quality databases will be acquired through several sources, and will include unusually extensive multi-year daily speciated particle data that will contribute to the development of novel mixtures characterizations. The MC metrics of specific interest for this project include: (1) population-weighted averages and spatially-resolved concentrations; (2) single-species source tracers; (3) multiphase pollutant source apportionment outputs; (4) factor analysis outputs; and (5) modeled reactive oxygen species. These pollutant mixture characterizations will be assessed in relation to acute cardiorespiratory outcomes.

Expected Results. Results from this project will advance our understanding of the specific sources, attributes and constituents of the ambient air pollutant mix that impact cardiorespiratory morbidities at two levels of disease severity. By exploring possible explanations for the heterogeneity in observed health associations across cities, we seek to elucidate common underlying relationships. In the process, we will identify susceptible and vulnerable subpopulations, describe the shape of concentration-response curves, assess the impacts of exposure measurement error, and identify seasonal and meteorological differences in observed associations. By examining the impacts of single pollutants in a multi-pollutant context and applying novel characterizations of pollutant mixtures, results of this project will help inform the development of multi-pollutant management approaches for the protection of human health.