LIFESTYLE AND ENVIRONMENTAL RISK FACTORS

Tobacco The causal relationship between tobacco smoking
and lung cancer was established by epidemiologic studies
conducted in the 1950s and 1960s. The complexity of tobacco
smoke, with over 3000 different chemicals, has made
it difficult to identify the contribution of more than 50 putative
carcinogenic agents. The carcinogens in tobacco smoke
include the polynuclear aromatic hydrocarbons (PAHs),
N- nitrosamines, aromatic amines, other organic (e.g., benzene,
acrylonitrile) and inorganic (e.g., arsenic, acetaldehyde) compounds,
and polonium-210. The composition of the smoke
depends on the ambient conditions of smoking, the blend of
tobacco leaf, filtration, additives, and paper wrapping. Tobacco
smoke produced by the tobaccos in pipes and cigars is both
harsher and more alkaline than that produced by cigarettes.
Most of the compounds in tobacco are produced in an oxygendeficient,
hydrogen-rich environment, arising from pyrolysis
and distillation, in the region immediately behind the burning
tip of the cigarette. The nicotine concentration is addictive and
toxic but not carcinogenic. 23,24
As mainstream cigarette smoke emerges from the cigarette,
it has approximately 10 9 to 10 10 particles per mL. The aerodynamic
diameters of the particles, ranging in size from 0.1 to
1.0 m, determine the sites of deposition in the airways and
alveolar regions of the lung. The fraction of smoke retained
varies markedly with the pattern of inhalation. The chemical
analysis of tobacco smoke is separated into particulate or “tar,”
and gaseous phases. Filter tips of cellulose acetate remove volatile
nitrosamines and phenols selectively. The neutral fraction
of the particulate phase contains potentially important tumor
initiators such as the PAHs. 25
In 1964, the first surgeon general’s report on smoking
summarized existing evidence and declared cigarette smoking
to be the major cause of lung cancer among American men. 26
In the ensuing 30 years, epidemiologic studies have established
that there were increasing risks in women and underscored the
relationships with onset, duration, intensity, and cessation of
smoking. Prospective studies demonstrated a rising trend in lung
cancer death rates with increasing average amounts smoked per
day in current smokers. The initial emphasis of epidemiologic
studies of lung cancer and smoking was on men who, in almost
all countries, began smoking earlier, consumed greater quantities
of tobacco, and exhibited higher RRs than women. 27–29
In the past 20 years, the prevalence of cigarette smoking in
many countries, including the United States, has increased significantly
among women; concomitantly, changes in smoking practices
have been accompanied by increasing relative and attributable
risks for lung cancer. 30–32 In a follow-up study of approximately
600,000 women conducted in 1980s by the American Cancer
Society, the RR of dying of lung cancer in current smokers was
12.7; for those who smoked 30 or more cigarettes per day, the
RR was increased 22.3 times compared with the never-smoker.
In 1985, cigarette smoking accounted for an estimated 82%
of lung cancer deaths, or 31,600 deaths. 33 The International
Agency for Research on Cancer (IARC) estimated that the smoking-
attributable fraction of lung cancer deaths occurring in the
United States and in England and Wales was 92% in men and
78% in women. 25 In 2001, the U.S. Health Interview Survey estimated
that 46.2 million adults were current smokers and that
44.7 million adults were former smokers. Current smoking prevalence
was highest among persons aged 18 to 24 years (26.9%),
and among those aged 25 to 44 years (25.8%), and lowest among
those aged older than 65 years (10.1%). From 2000 to 2001, current
smoking prevalence, for the first time, was similar in non-
Hispanic white (25.4%) and black (27.7%) men; in contrast, the
prevalence in non-Hispanic black women (17.9%) was less than
that in non-Hispanic white women (22.8%). 32
Lower tar content and the use of filters are factors that
may result in reduced lung cancer risks in those who smoke. In
the earlier American Cancer Society (ACS) Twenty-five–State
Study, men who smoked low-tar ( 22 mg) cigarettes experienced
20% lower risk of dying of lung cancer when compared
with men who continued to smoke high-tar cigarettes. The
excess lung cancer risk for current smokers was directly proportional
to the estimated milligrams of tar consumed daily. In
the more recent ACS Fifty-State Study, Garfinkel and Stellman
concluded that doubling the cigarette tar yield would result in a
40% increase in the RR of dying of lung cancer, independently
of the amount smoked or depth of inhalation. The Federal
Trade Commission estimated that the current average salesweighted
tar content of cigarettes manufactured in the United
States was about 12 to 13 mg of tar per cigarette, compared
with nearly 40 mg in the early 1950s. Lifelong filter cigarette
smokers have experienced 20% to 40% lower risk of lung cancer
than lifelong nonfilter smokers, after adjusting for differences
in the amount smoked. Presumably, larger reductions in
risk have not been observed because of alternations in smoking
behavior in response to low-nicotine yield of manufactured
cigarettes. Namely, it has been shown that in maintaining addiction,
the smoker will inhale larger volumes of mainstream
smoke and at more frequent intervals. 34–37
Although these studies suggested that switching to filtered
or low-tar cigarettes may modestly reduce the risk of lung cancer,
the more significant reduction in risk would be derived from
cessation of smoking. Whereas approximately 25% of smoking
adults in the United States continue to smoke, an additional
40% to 50% have become former smokers. The RR of lung
cancer among ex-smokers decreases significantly after 5 years of
smoking cessation. In the initial 1 to 4 years after quitting smoking,
however, the RR of lung cancer among ex-smokers may
appear to be higher than among current smokers, presumably
because a proportion of individuals may have stopped smoking
because of illness or premonitory symptoms of lung cancer. 38,39
It has been suggested that the risk of lung cancer in former
smokers will approximate but never equal that of lifelong nonsmokers.
The baseline risk of lung cancer in lifelong nonsmokers
increases in relation to age raised to the fourth or fifth power.
In the British Physicians Study, Doll and Peto38 showed that
the incidence of lung cancer in cigarette smokers increased approximately
in proportion to the fourth power of duration of
smoking, and was multiplicative with the previously described
exponential increase with age among never-smokers. The percentage
reduction in risk after quitting depended on the prior
duration and average amount smoked each day, being more
readily demonstrable among lighter smokers and smokers of
lesser duration or those who quit at a younger age. Lung cancer
results from a multistep process in which persistent genetic
lesions accumulate at specific chromosomal loci. Most current
or former smokers, in contrast to never-smokers, exhibit loss of
heterozygosity at multiple allelic sites (e.g., 3p14, 9p21, p16,
p53) in both normal and metaplastic or dysplastic bronchial
epithelium. 40,41
Pipe and cigar smoking have been linked to lung cancer,
particularly squamous cell and small cell carcinomas, but the
estimated RRs, compared with people who never smoked, who
are assigned an RR of 1.0, were considerably lower than the risks
reported among cigarette smokers; the risks among exclusively
pipe or cigar smokers in the United States or Europe have been
estimated to range from 2.0 to 9.0. In countries such as Sweden,
Switzerland, and Holland, where pipe or cigar smoking was nearly
as common as cigarette smoking, the RRs of lung cancer, when
controlling for cumulative exposure levels, were equally high for
all forms of smoked tobacco. Differences in the manner in which
pipes and cigars were smoked in different countries, (i.e., depth
of inhalation or average daily or cumulative amount smoked)
may provide an explanation for differences in the estimated risks.
Cigars are products made of tobacco, wrapped in tobacco leaves
rather than in paper. Cigars smoked in Europe weigh 2 to 8 g and
are similar to American “small cigars.” Cigarillos are smaller than
cigars weighing 1.5 to 3 g and are described as “little cigars” in
the United States. Risks of lung cancer also varied with the type
of tobacco used. Dark tobaccos were associated with greater risk
of lung cancer than light tobaccos, and with formation of higher
levels of 4-aminobiphenyl-hemoglobin adducts. 42–45