Stephen J.LurieMD, PhD, Senior EditorIndividualAuthor
Copyright 2002 American Medical Association. All Rights Reserved.
Applicable FARS/DFARS Restrictions Apply to Government Use.2002
In Reply: We controlled for educational level,
as well as for smoking and other risk factors, in our pollution mortality
estimates. There was no evidence that the estimated pollution-mortality effects
were due to confounding by education or other observed risk factors. Nevertheless,
we found an apparent effect modification by smoking status and education levels.
We agree with Dr Finkelstein that the education-related effect modification
may be at least partially due to higher exposures due to closer proximity
to pollution sources, poorer housing, less home and work-place air conditioning
and filtering, and possibly other factors. There are, of course, various important
consequences related to this effect modification. For example, because individuals
with less education are somewhat underrepresented in our cohort, we may have
underestimated the average pollution-mortality effect estimates for the overall
Pope III CA, Burnett RT, Thurston GD. Pollution-Related Mortality and Educational Level—Reply. JAMA. 2002;288(7):830. doi:10.1001/jama.288.7.828