Multisite Time-Series Studies Versus Cohort Studies: Methods, Findings, and Policy Implications

Abstract

Time-series and cohort studies of air pollution on human health have advanced greatly our understanding of the effects of air pollution on health since the earliest studies. Availability of large national databases and progress in computational tools and statistical methods have made possible the estimation of national average pollution effects and the exploration of potential sources of heterogeneity in the effects of air pollution across countries or regions. Interpretation of the findings needs to account for several challenges, including confounding and the resolution of seemingly conflicting results from time-series and cohort studies. This article presents an overview of the time-series and cohort studies’ approaches for estimating the relative risk of mortality from particulate air pollution and discusses the statistical issues and challenges inherent in each of these studies. We also discuss policy relevant summaries in air pollution epidemiology, approaches for estimating the impact of particulate matter on mortality from time-series and cohort studies, and research opportunities under the National Medicare Cohort Study (NMCS).