A Source-Oriented Evaluation of the Combined Effects of Fine Particles and CopollutantsEPA Grant Number: R827997
Title: A Source-Oriented Evaluation of the Combined Effects of Fine Particles and Copollutants
Investigators: Ito, Kazuhiko , Thurston, George D.
Current Investigators: Ito, Kazuhiko , Thurston, George D. , Xue, Nan , Lall, Ramona , DeLeon, Samantha
Institution: New York University
Current Institution: New York University School of Medicine , Albert Einstein College of Medicine
EPA Project Officer: Chung, Serena
Project Period: February 18, 2000 through February 17, 2004 (Extended to February 17, 2006)
Project Amount: $478,522
RFA: Airborne Particulate Matter Health Effects (1999) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Air , Human Health , Particulate Matter
The objective of this project is to apply novel approaches to estimate the combined effects of size-specific particulate matter (PM) air pollution (i.e., PM2.5 and PM10-2.5) and their copollutants in major U.S. cities where source types, levels of PM copollutants, and weather patterns vary considerably. PM is a chemically non-specific pollutant, and may originate or be derived from different emission source types. Thus, its toxicity should vary depending on its chemical composition and perhaps on the presence of any gaseous co-pollutants. In time-series analyses of the acute effects of PM, the prevailing approach to dealing with gaseous copollutants is to treat them as confounders, and to include them simultaneously in regression models. Such an approach can not only lead to misleading conclusions in "identifying" the causal pollutant (e.g., when pollutants are correlated and have varying extents of exposure error), but also cannot address the likely combined effects of PM and gaseous copollutants. The regulatory implications of these limitations in the prevailing regression approach could be serious, because the wrong sources may be identified for regulatory control. The expected reduction in risk (i.e., benefits) may also not be optimized, depending on the extent to which the PM and the gaseous copollutants share the same source types. To address these issues in this project, the health outcomes will be regressed on source-oriented factors, as identified from factor analysis of PM and gaseous pollutants, rather than upon individual pollutants. This will allow the identification of any especially toxic pollutant sources, which is expected to be more fruitful than trying to tease apart individual pollutant that are correlated with each other. The null hypothesis to be tested is that the PM effect size estimate is constant for all source types, regardless of composition or the presence of gaseous copollutants.
The first stage of the data analysis will characterize common daily fluctuations among PM2.5 (and chemical speciation data where available), and gaseous pollutants (O3, NO2, SO2, and CO) using factor analysis in order to identify and characterize pollution types (e.g., automobile, secondary aerosols, etc.). Secondly, Poisson models will be used to regress daily health outcomes on the identified factors (factor scores) that include PM2.5 in order to estimate the combined effects of PM2.5 and gaseous copollutants, adjusting for seasonal cycles, day-of-week effects, and weather, as well as to model interaction between factors. This factor analysis/Poisson regression approach has considerable advantages over the past regression models that have attempted simultaneous inclusion of correlated variables, as it can provide combined effects of a group of pollutants that may come from the same source, or be generated under common conditions. Since this is a relatively new approach to be used in air pollution epidemiology, the conventional approach will also be applied to the same data in parallel with the new approach, so that the results can be compared and contrasted. This project will take advantage of the growing PM2.5 network database that is currently being established nation-wide. Since up to one-fourth of the population-oriented monitors (900-1000) in heavily populated areas will collect PM2.5 every day, this database will provide sufficient sample size to conduct daily mortality and elderly hospital admission health effects analyses, as well as detailed source type characterization of PM-gaseous pollutant mixtures. The data analyses will be first conducted on the existing PM data and mortality (1985-1998) and will be extended to up to 2001 for mortality and 2002 for the elderly hospital admission data. The study area will include major urban cities to be chosen from CA, IL, MI, NY, PA, TX, and WA, based on the initial analysis to examine the completeness of the PM data.
This study will provide the mortality and morbidity risk estimates for combined PM and gaseous pollutants by emission source class, and will provide a link between source types and estimated health impacts. Factor analysis of PM2.5, PM10-2.5, gaseous pollutants, and weather variables are expected to yield pollution and weather components that are readily distinguishable from each other. The PM contributions will be apportioned into these different pollution categories (factors). These source factors will then be tested for their varying extents of associations with mortality and morbidity by inter-comparing their health effects RR's. Understanding the respective importance of various pollution sources to health effects is an essential step towards a more comprehensive air pollution regulatory policy. These results will therefore provide valuable input towards achieving more cost-effective regulatory policy. Moreover, comparisons with the conventional time-series modeling approach will allow us to understand and make better use of the available results from past and on-going short-term health effects studies. Overall, these results will help reduce the uncertainty associated with existing estimates of the short-term mortality and morbidity effects of PM and copollutants, and will thereby improve air pollution risk assessment.