Smoking and Tobacco Control Monograph No. 7 
Then, if you think about trying to develop approaches that might be 
used in epidemiological studies, based around questionnaires to try to 
develop tools that would provide a better measure of dose, which I think 
is what you are calling for, it becomes very difficult. 
You know, using some of our nested approaches, one might begin to use 
biomarkers within studies, within cohort studies, probably particularly, to 
sort this out. But 1 think you are pointing to a significant limitation of 
approaching this question in large population studies. 
DR. BENOWITZ: The biggest effect was clearly in the lung cancer data, and 
the lung cancers occurred as a result of cigarettes smoked a long time ago. 
Is there any evidence that there is any difference in risk if you looked at 
modern or filtered cigarettes? 
DR. SAMET: Let me see if 1 can rephrase the question. Are you asking, has 
there been an attempt to assess whether some estimate of tar dose, or tar 
received, is a better predictor of lung cancer risk than simply proportion of 
filter use? 
DR. BENOWITZ: Yes. What 1 am wondering is, is there any relevance to 
the data when people were mostly smoking nonfiltered cigarettes to today's 
cigarette market, where they are filtered? Can the whole thing be done just 
by adding a filter? 
DR. SAMET: Probably the right answer to the question is: 1 do not know. 
But if we think we could begin to use the information from studies of 
smokers of old nonfiltered products, through smokers of newer products, 
to try and define some kind of an exposure-response relationship, then 
I suppose it could be done. But 1 think that, if we were to do that, it would 
be subject to a great deal of uncertainty. 
DR. HARRIS: 1 noticed that one of the studies omitted from your review 
was the second American Cancer Study, CPS-11, which followed people from 
1982 to 1986. 1 am wondering if anyone knows whether that study will be 
analyzed in terms of the yield or type of cigarette and health outcomes. 
DR. SAMET: There has already been a paper describing the demographics of 
tobacco use in that study and predictors of tar yield by various demographic 
predictors. 1 would anticipate seeing such an analysis eventually. 
DR. WOOSLEY: We have already heard this morning how the marketing and 
the promotion of the low-tar and -nicotine cigarettes have been toward the 
more highly educated portion of the population. We have already seen how 
they responded to that by switching. We have already heard how they have 
expressed greater concern for their overall health. 
I have a serious concern. Do you feel the data have adequately addressed 
the possibility that you are looking at a subset of the population who have 
done something else to modify their health risks and, therefore, have looked 
at a selected population with decreased negative outcomes because of these 
88 
