The American Journal of Public Health (July 1998) reports the lung cancer risk from workplace exposure to secondhand smoke to be about the same as the reported risk from home exposure -- about a 19 percent increase in risk. This report is based on a meta-analysis of the 14 studies with so-called worker exposure data. When the analysis is limited to the five studies judged to have the best quality data, the reported increase in lung cancer risk is 43 percent.
The report is a statistical mess, though.
As the authors acknowledge, "It is evident that the meta-analysis was dominated by the large Fontham/Reynolds study, which constituted 63 percent of the [lung cancer] cases and 56 percent of the statistical weight." In the 5-study analysis, the Fontham study is the only one to achieve statistical significance. So, as it always has been, the Fontham study is the key to the house of cards that is the link between secondhand smoke and lung cancer.
But the reliability of the worker exposure data in the Fontham study was previously dismantled in a 1994 report I did for the U.S. Department of Energy titled "Choices in Risk Assessment: The Role of Science Policy in the Environmental Risk Management Process."
On page 191 of "Choices in Risk Assessment," I wrote:
"However, the Fontham study's occupational risk estimates may be overstated because they may be confounded by uncontrolled and unaccounted for nonoccupational exposures to ETS, including exposures in the household and social settings. The Fontham study reported increases in lung cancer risk of 23 percent from household exposures to ETS, 50 percent from social exposures to ETS, and 39 percent from occupational exposure to ETS." The authors acknowledge that these exposures may be concurrent (i.e., study subjects may have lived with smokers, worked with smokers, and socialized with smokers during the same period). No statistical adjustment for concurrent exposures was made in the analysis. Without adjustment, the occupational risk estimate may also reflect and be confounded by potential risks from household or social exposures to ETS.So the Fontham study's worker exposure data is anything and everything but worker exposure data.
In combination with the numerous other shortcomings of the secondhand smoke/lung cancer epidemiology (i.e., weak statistical associations, unverified exposure data, and uncontrolled confounding risk factors), this new report says more about passive thinking than passive smoking.
Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.
Material presented on this home page constitutes opinion of Steven J. Milloy.