Chapter 6 
pack-years (packs smoked per day multiplied by the number of years smoked) 
is a cumulative exposure measure. An estimate of kilograms of tar deposited 
in the lung is an absorbed dose measure; nanograms of benzo(a)pyrene, for 
example, reaching basal cells might be considered a biologically effective dose 
for carcinogenesis. New markers take dose measures to the molecular level 
(Vineis and Caporaso, 1995). 
To assess the consequences of changes in the cigarette, it is necessary to 
have information on how changes in tar and nicotine yield, as assessed by 
the Federal Trade Commission (FTC) method, affect dose measures, extending 
to the molecular level. Any new approach to testing cigarette yields should 
be designed to be informative both as an exposure measure and as an 
indicator of biologically relevant doses of cigarette smoke components. 
EPIDEMIOLOGIC 
EVIDENCE ON 
THE CHANGING 
CIGARETTE AND 
DISEASE RISKS 
Overview 
Epidemiologic evidence is available on the effect of the changing 
cigarette on all-cause mortality and on three major categories of 
disease caused by cigarette smoking: lung and other cancers, 
nonmalignant respiratory diseases, and cardiovascular disease 
(CVD). The Health Consequences of Smoking: The Changing Cigarette. 
A Report of the Surgeon General (U.S. Department of Health and 
Human Services, 1981) addressed the changing cigarette, covering 
the relevant toxicologic and epidemiologic evidence. This chapter considers 
key epidemiologic publications since that report but does not provide a 
systematic overview of the many studies on the changing cigarette. 
The principal study designs that have been used to address the health 
consequences of the changing cigarette are the ecological study, a descriptive 
approach conducted at the group level, and cohort and case-control studies, 
analytic approaches conducted at the individual level. Cross-sectional studies 
have proven informative in investigating nonmalignant respiratory diseases. 
The ecological approach is exemplified by a comparison of temporal changes 
in rates of smoking-related diseases with patterns of consumption of various 
types of cigarettes. The American Cancer Society (ACS) studies of large 
groups of volunteer participants are cohort studies; the participants were 
enrolled, information about smoking was obtained on enrollment and 
periodically thereafter, the population was followed over time, and mortality 
was ascertained. Some of the earliest evidence on smoking and lung cancer 
was obtained in the classic case-control studies conducted by Doll and Hill 
(1950) and Wynder and Graham (1950). In these studies, the smoking habits 
of patients hospitalized with lung cancer were compared with the smoking 
habits of control patients having another disease. 
Evidence from epidemiologic studies has well-known strengths and 
limitations (Rothman, 1986). Epidemiologic research has had a central role 
in characterizing the consequences of the changing cigarette because it 
supplies direct information on the consequences of varying tar and nicotine 
yield products. Thus, the findings inherently consider compensatory 
changes in inhalation patterns or in numbers of cigarettes smoked and 
provide the evidence needed to answer the question of immediate public 
79 
