Distinguishing Health Effects Among Different PM2.5 Components

Abstract

Researchers began to consistently find associations with adverse health risks of elevated PM2.5 (particulate matter 2.5 µm or less in diameter), measured as mass, in the 1990s. However, monitoring information was then available for few PM2.5 components. Advisory bodies such as the Clean Air Scientific Advisory Committee (1996) and the National Research Council (NRC) of the U.S. National Academies of Sciences (in 1998 and later) suggested the need to better understand which PM2.5 constituents are most or least harmful, to enable targeted control strategies. But how should researchers best distinguish risks of harm from different PM2.5 constituents in complex mixtures of primary and secondary particles? Methodological issues may play a primary role in this research endeavor. Principles of methodology discussed herein include: (1) need for accurate exposure information in epidemiological studies; (2) using epidemiological studies which compare effects of a wide range of health-relevant pollutants against the same health endpoints in the same study; (3) recognition that a given source can emit several pollutants, and conversely that some common pollutants have multiple sources, but with different co-pollutants, thus creating the possibility that statistical associations with one pollutant could be attributable to co-emitted pollutant(s), perhaps unmeasured; and (4) accounting in some way for atmospheric chemical processes. Further, examining for a relevant pollutant both toxicology studies (in vitro and in vivo) and epidemiological studies (including population based and human panel studies reflecting use of these four principles), might provide the fullest evidentiary pathway to determining differential toxicological properties among PM2.5 components.