Citation

Goldberg MS, Burnett RT, Brook JR. Regul. Toxicol. Pharmacol. 2008; 51(2): 141-7; discussion 148-50.

Affiliation

Department of Medicine, McGill University, Montreal, Que., Canada H3A 1A1. mark.goldberg@mcgill.ca

Copyright

(Copyright © 2008, Elsevier Publishing)

DOI

10.1016/j.yrtph.2008.03.006

PMID

18462855

Abstract

We discuss Bukowski's hypothesis that unmeasured, individual risk factors, notably stress caused by being in traffic or at work, may confound the association found in many time-series studies between air pollution and cardiovascular outcomes. In replying to these concerns, we describe the methods used to carry out time-series studies and discuss whether it is reasonable to believe that stress can confound the association. Our analysis suggests that these factors are unlikely to cause bias in time-series studies, although it may be important in panel and other longitudinal studies. Specifically, we argue that: (1) persons most likely to be susceptible to the effects of air pollution, notably the very young and the elderly, and those suffering from certain chronic diseases (e.g., congestive heart failure), are unlikely to be highly exposed to traffic and therefore to traffic-related stress and (2) over lengthy periods of time and sub-populations, levels of individual stress are unlikely to act coherently to create an association with ambient air pollution.

Language: en