Tobacco industry publishes disinformationBMJ 2003; 327 doi: https://doi.org/10.1136/bmj.327.7413.502-c (Published 28 August 2003) Cite this as: BMJ 2003;327:502
- Michael J Thun, vice president, epidemiology and surveillance research ()
Editor–The American Cancer Society does not agree with the conclusions of Enstrom and Kabat in their analysis of environmental tobacco smoke in the cancer prevention study I (CPS-I).1 Their study is fatally flawed because of misclassification of exposure. The cancer prevention study was started by the society in 1959 to measure the effects of active smoking, not to collect valid estimates of exposure to environmental tobacco smoke.2 No information was obtained on sources of exposure to environmental tobacco smoke other than the smoking status of the spouse. Tobacco smoke was so pervasive in the United States in the 1950s and 1960s that virtually everyone was exposed, at home, at work, or in other settings. Enstrom and Kabat essentially compare non-smokers, married to a smoking spouse, with non-smokers with other sources of exposure to environmental tobacco smoke. Misclassification of exposure is compounded because no information was collected on the smoking status of the spouse between 1972 and 1999. Non-smokers whose spouses reported smoking at the start of the study are classified as “exposed” even if the spouse quit, died, or the marriage ended during this interval. This problem is not solved by the 1999 resurvey of survivors, since these represent only 2% of the original analytic cohort and 5% of those followed after 1972. Other serious flaws of the Enstrom and Kabat paper are discussed elsewhere.3
This is the second attempt by tobacco industry consultants to publish flawed analyses of environmental tobacco smoke using cohort studies from the American Cancer Society.4 Sadly, the forum in which such studies are influential is not the scientific world–scientists recognise these studies for what they are–but in communities that are considering clean air laws.
Competing interests None declared.