Long-Term Effects of Fine Particles on Mortality: Insights from 1984

I review an important study that Professor Evans published early in his career examining the role of cross-sectional mortality studies in air pollution risk assessment. At a time when both risk assessment and particle effects on mortality were controversial, John's thoughtful analysis of the issues and data relevant to assessing long-term mortality risks from airborne particles provides a comprehensive primer that is still relevant today. The paper includes a critical literature review, a meta-analysis of published particle effect estimates, and a reanalysis of landmark cross-sectional mortality data set. EPA criteria documents and related literature had largely discounted the cross-sectional mortality findings on the basis of criticisms about exposure assessment and control for confounding. John's analysis reached a different conclusion, that is, "we are of the opinion that the cross-sectional studies reflect a causal relationship between exposure to airborne particles and premature mortality. From our point of view it is as likely that parameters have been underestimated … as that they are overestimated due to confounding." The paper acknowledged the impossibility of precisely quantifying the long-term mortality effect of particle air pollution, and that there is a need for further research utilizing alternative approaches. These conclusions foreshadow the emergence, a decade later, of the influential particulate matter (PM) mortality findings from the Harvard Six Cities and American Cancer Society cohort studies. I conclude by suggesting that well designed cross-sectional studies could play a role in identifying exposure-response associations in resource-poor settings where there is a paucity of local evidence to support air pollution regulations.