Chapter 6 
other factors, and that we really have not seen any influence of the cigarettes 
themselves? 
DR. SAMET: 1 referred to that set of concerns under the rubric of selection 
bias. That is, people may select themselves to products based on either 
their response to what they were smoking or other characteristics that are 
relevant — an argument in epidemiology called confounding. 
I think you are right; these are concerns. 1 think, on the other hand, in 
many of the studies there have been attempts to "adjust," to the extent one 
can, for such differences in the characteristics of those using different types 
of products. As you look across the consistency of the evidence in different 
populations with different approaches to controlling for such factors, and 
different study designs, a consistency emerges, 1 think at least for lung 
cancer, that would suggest some modest reduction of risk for those using 
the lower delivery products. 
Could there be some element of residual bias in there? 1 certainly 
could not exclude it. But when we weigh the evidence in an attempt 
to understand those other factors, the socioeconomic indices and other 
measures in different studies would support that conclusion. 
DR. BENOWITZ: I think that is the most important issue that we have to 
address here today. If labeling something low-tar and -nicotine implies 
improved health compared with higher tar and nicotine, I think those 
confounders have the most impact on that decision. 
DR. HOFFMANN: With regard to Dr. Benowitz' question to me, it is rather 
interesting to see that multiple studies have shown that the increase in 
adenocarcinoma today is much higher than previously, because the nature 
has changed. So, to me, this has something to do with the cigarette. You get 
more adenocarcinoma in the peripheral lung than in former times; it is a 
ratio of 20 to 1 squamous cells, and today you have 1 to 1. So, I think at 
least the type of lung cancer that appears today has something to do with 
the change in cigarettes. 
DR. SAMET: But certainly the histologic distribution of lung cancers has 
changed and 1 agree; we would like to know why. 
DR. DEBETHIZY: Your data about the relationship between nicotine and 
cardiovascular disease are curious to me, because most of the data in the 
literature show that people who smoke low-yielding cigarettes actually 
absorb less nicotine. Could you comment on the fact that you do not see 
any dose-response relationship there? 
DR. SAMET: I am not sure how you would like me to comment. 1 am 
describing the findings of a case-control study that describes how risks of 
nonfatal myocardial infarction varied with the level of nicotine or carbon 
monoxide intake, as estimated by what brand was being smoked at the 
time of the infarct. 
89 
