Although particulate matter (PM), nitrogen dioxide (NO 2 ) and carbon monoxide (CO) typically exist as part of a complex air pollution mixture, the evidence linking these pollutants to health effects is evaluated separately in the scientific and policy reviews of the National Ambient Air Quality Standards (NAAQS). The objective of this analysis was to use meta-regression methods to model effect estimates for several individual yet correlated NAAQS pollutants in an effort to identify factors that explain differences in the effect sizes across studies and across pollutants. We expected that our consideration of the evidence for several correlated pollutants in parallel could lead to insights regarding exposure to the pollutant mixture. We focused on studies of hospital admissions for congestive heart failure (CHF) and ischemic heart disease (IHD), which have played an important role in the evaluation of the scientific evidence communicated in the PM, NO 2 , and CO Integrated Science Assessments (ISAs). Of the studies evaluated,
OPEN ACCESSAtmosphere 2011, 2 689 11 CHF studies and 21 IHD studies met our inclusion requirements. The size of the risk estimates was explained by factors related to the pollution mixture, study methods, and monitoring network characteristics. Our findings suggest that additional analyses focusing on understanding differences in effect sizes across geographic areas with different pollution mixtures and monitor network designs may improve our understanding of the independent and combined effects of correlated pollutants.