You are here:
Examination of Different Exposure Metrics in an Epidemiological Study
Citation:
OZKAYNAK, H. A., V. ISAKOV, S. Sarnat, J. Sarnat, AND J. Mulholland. Examination of Different Exposure Metrics in an Epidemiological Study. EM: AIR AND WASTE MANAGEMENT ASSOCIATION'S MAGAZINE FOR ENVIRONMENTAL MANAGERS. Air & Waste Management Association, Pittsburgh, PA, (Nov):22-27, (2011).
Impact/Purpose:
see description
Description:
Epidemiological studies of air pollution have traditionally relied upon measurements of ambient concentration from central-site monitoring stations as surrogates of population exposures. However, depending on the epidemiological study design, this approach may introduce exposure prediction errors and misclassification of exposures for pollutants that are spatially heterogeneous, such as those associated with traffic emissions (i.e., carbon monoxide, elemental carbon, nitrogen oxides and fine and ultrafine particulate matter). The U.S. EPA in collaboration with Emory University recently started developing and evaluating several tiers of exposure metrics and their influence on epidemiological study findings, for a variety of ambient traffic-related and regional pollutants in the large Atlanta metropolitan area. This study examines the spatial and temporal variations of alternative measured and modeled exposure metrics for multiple pollutants (i.e., particulate matter, elemental carbon, sulfates, nitrogen oxides, carbon monoxide and ozone). Alternative exposure assignment approaches considered various techniques ranging from traditional based on central site ambient measurements, to more sophisticated exposure models such as SHEDS (Stochastic Human Exposure and Dose Simulation) model. Several of these exposure metrics for ambient traffic-related and regional pollutants were then utilized in epidemiological models to estimate acute air pollution health impacts. In this work, we hypothesize that using the more refined exposure estimates will provide greater power in detecting epidemiologic associations of interest, particularly for spatially heterogeneous, traffic-related ambient pollutants. Specifically, we explore the associations between daily ZIP code level traffic-related fine elemental carbon (EC) and regional fine particulate matter with diameter 2.5 microns or less (PM2.5) and daily ZIP code level respiratory and cardiovascular ED visits in Atlanta during 1999-2002. We compare both the exposure and epidemiological results obtained when using ambient monitoring data with those obtained using population exposure estimates produced by the SHEDS model, which utilized as inputs hybrid ambient air quality model predictions for PM2.5 and EC.
