JNCI | Articles 541
Advance Access publication on March 14, 2012.
© The Author(s) 2012. Published by Oxford University Press.
This is an Open Access article distributed under the terms of the Creative Commons Attribution
Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0), which permits unrestricted
non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
After the US Surgeon General’s first report on Smoking and
Health was issued in 1964, several initiatives, collectively known
as “tobacco control,” included restrictions on smoking in public
places, large increases in cigarette excise taxes, reduced access to
cigarettes, and increased public awareness of the hazards of smoking.
These smoking regulations have been cited as the principal con-
tributors to the observed decline in US adult tobacco use from
1975 to 2003 (1,2) and to subsequent declines in smoking-related
mortality (3). In 2004, the Surgeon General (4) estimated that
during 1965–1999, approximately 3 million lung cancer deaths
(2.2 million among men and 0.8 million among women) in the
United States were attributable to smoking. Using a straightforward
demographic projection, Thun and Jemal (5) estimated that reduc-
tions in tobacco smoking averted approximately 146 000 lung cancer
deaths among US men during 1991–2003.
In this article, we analyzed the direct influence that changes in
smoking behaviors that began in the mid-1950s had on lung cancer
mortality rates among men and women aged 30-84 years in the
United States during 1975–2000. We also estimated the total
number of lung cancer deaths averted among men and women
during the same period as a direct result of changes in smoking
behavior. Finally, we estimated the numbers of avoidable deaths,
that is, the number of lung cancer deaths that could have been
averted had smoking been completely eliminated as of 1965.
A consortium of six universities and research centers (Erasmus
Medical Center [Erasmus MC], Rotterdam, the Netherlands; Fred
Hutchinson Cancer Research Center [FHCRC], Seattle, WA;
Pacific Institute for Research and Evaluation [PIRE], Calverton,
MD; Rice University and M.D. Anderson Cancer Center [Rice-
MDA], Houston, TX; Massachusetts General Hospital and Harvard
Medical School [MGH-HMS], Cambridge, MA; and Yale University,
New Haven, CT) developed independent models to estimate the
impact of tobacco control policies on lung cancer mortality.
Impact of Reduced Tobacco Smoking on Lung Cancer Mortality
in the United States During 1975–2000
Suresh H. Moolgavkar, Theodore R. Holford, David T. Levy, Chung Yin Kong, Millenia Foy, Lauren Clarke, Jihyoun Jeon,
William D. Hazelton, Rafael Meza, Frank Schultz, William McCarthy, Robert Boer, Olga Gorlova, G. Scott Gazelle, Marek Kimmel,
Pamela M. McMahon, Harry J. de Koning, Eric J. Feuer
Manuscript received March 25, 2011; revised February 1, 2012; accepted February 3, 2012.
Correspondence to: Suresh H. Moolgavkar, MD, PhD, Program in Biostatistics and Biomathematics, Fred Hutchinson Cancer Research Center, 1100
Fairview Ave North, Seattle, WA 98109 (e-mail: firstname.lastname@example.org).
Background Considerable effort has been expended on tobacco control strategies in the United States since the mid-1950s.
However, we have little quantitative information on how changes in smoking behaviors have impacted lung
cancer mortality. We quantified the cumulative impact of changes in smoking behaviors that started in the mid-
1950s on lung cancer mortality in the United States over the period 1975–2000.
Methods A consortium of six groups of investigators used common inputs consisting of simulated cohort-wise smoking
histories for the birth cohorts of 1890 through 1970 and independent models to estimate the number of US lung
cancer deaths averted during 1975–2000 as a result of changes in smoking behavior that began in the mid-1950s.
We also estimated the number of deaths that could have been averted had tobacco control been completely
effective in eliminating smoking after the Surgeon General’s first report on Smoking and Health in 1964.
Results Approximately 795 851 US lung cancer deaths were averted during the period 1975–2000: 552 574 among men
and 243 277 among women. In the year 2000 alone, approximately 70 218 lung cancer deaths were averted: 44 135
among men and 26 083 among women. However, these numbers are estimated to represent approximately 32%
of lung cancer deaths that could have potentially been averted during the period 1975–2000, 38% of the lung
cancer deaths that could have been averted in 1991–2000, and 44% of lung cancer deaths that could have been
averted in 2000.
Conclusions Our results reflect the cumulative impact of changes in smoking behavior since the 1950s. Despite a large
impact of changing smoking behaviors on lung cancer deaths, lung cancer remains a major public health problem.
Continued efforts at tobacco control are critical to further reduce the burden of this disease.
J Natl Cancer Inst 2012;104:541–548
542 Articles | JNCI Vol. 104, Issue 7 | April 4, 2012
Although the models shared common inputs, each group developed
its own model, based on mathematical descriptions of lung carci-
nogenesis as it relates to smoking behaviors. The models explicitly
consider factors associated with the risk of smoking, including the
number of cigarettes smoked per day, the age of initiation, and the
number of years quit.
All models shared the same overall structure (Figure 1). The
central component of each model was a dose–response module that
provided a quantitative description of the age-specific lung cancer
mortality among never smokers and age-specific lung cancer
mortality among continuing smokers and former smokers by detailed
history of smoking. This module was used to predict age-specific
lung cancer mortality rates associated with three specific smoking
scenarios. With the exception of the MGH-HMS group, which
used a set of logistic regression models and tumor progression
functions (6,7), the other groups used multistage models (8–11) for
the underlying dose–response module. Multistage models, based
on mathematical formalisms representing the biological paradigm of
initiation, promotion, and progression, recognize that carcinogenesis
includes accumulation of mutations and clonal expansion of partially
altered cells on the pathway to malignancy (8,12–17). These models
may be used to explore biological hypotheses regarding the mech-
anism of tobacco-induced lung cancer. They have generally shown
clonal expansion (promotion) of partially altered (initiated) cells
by cigarette smoke to be the dominant mechanism and have
confirmed the disproportionate importance of smoking duration on
lung cancer risk (10,11,18–21). Both the multistage models and the
probabilistic model of MGH-HMS are capable of accommodating
detailed individual-level smoking histories, including temporal
factors, such as age at start, age at cessation, and temporal changes
in the level of smoking. The parameters of these models were
estimated as described in the supplementary material (available
online) by fitting the model to specific epidemiological cohort data
(Erasmus MC, FHCRC, PIRE, Yale), case–control data (Rice-MDA),
or registry data (MGH-HMS). The Erasmus, FHCRC, and Yale
model parameter estimates were obtained by fits of the Two-Stage
Clonal Expansion (TSCE) model to the Nurses Health Study
(NHS) and Health Professionals Follow-up Study (HPFS) cohort
data (11). The model parameters for the PIRE model were derived
by fitting the TSCE model to the Cancer Prevention Study II
(CPS II) data (10). The parameters for the Rice-MDA model were
derived from fitting the TSCE model to data from a case–control
study conducted to evaluate the interaction of smoking and genetics
on lung cancer risk (21). The parameters for the MGH-HMS model
were estimated by fitting a nested set of logistic regression models
for cancer development and spread along with a tumor growth
function to lung cancer incidence in the Surveillance, Epidemiology,
and End Results (SEER) registry during the period 1990–2000 (6,7).
Three specific smoking scenarios, the common inputs for the
models, were simulated using a smoking history generator as briefly
described below. Each involved a detailed description of smoking
behaviors by sex and birth cohort starting with the birth cohort of
1890 and ending with the birth cohort of 1970. The actual tobacco
control (ATC) scenario is a quantitative description of the actual
smoking behaviors of men and women in the United States. The
no tobacco control (NTC) scenario is a quantitative description of
the predicted smoking behaviors of men and women in the United
States under the assumption that tobacco control efforts starting
mid-century had never been implemented. Initiation rates in the
NTC scenario are from models fitted to survey data after age and
cohort effects had stabilized. For men, these are cohorts born after
1904. The history for women is more complex because age effects
stabilized after 1919 but cohort effects continued to increase;
therefore, a linear cohort trend for 1930–1955 was used in the
model for log rate, and the estimated parameter was held constant
after 1955. For smoking cessation rates, we used data for individuals
born in 1900–1904 because they would have had little knowledge
of the health effects of smoking for most of their lives while at the
same time being sufficiently well represented in surveys to provide
accurate estimates of these rates. The complete tobacco control
(CTC) scenario is a quantitative description of the predicted
smoking behaviors of men and women in the United States under
the assumption that all smoking ceased abruptly in 1965, that is, all
CONTEXT AND CAVEATS
The proportion of smokers among the US population has gradually
declined since the mid-1950s, as the dangers of tobacco use became
apparent and tobacco control laws were enacted. However, there
are few estimates of how many lung cancer deaths were spared by
the decline in cigarette smoking.
Six groups of investigators built independent models based on
cohort, case–control, or registry data and calibrated to mortality or
other data to estimate the number of lung cancer deaths averted in
1975–2000. The data were stratified by sex and birth decade (1890–
1970), and the prevalence of smoking and lung cancer deaths under
three scenarios were considered: actual tobacco control (ATC),
based on historical changes in smoking rates; no tobacco control
(NTC), based on predicted smoking rates if tobacco control had not
been enacted; and complete tobacco control (CTC), which considers
what might have happened if all smoking ceased in 1965.
In the United States in 1975–2000, there were 2 067 775 lung cancer
deaths among men and 1 051 978 lung cancer deaths among women.
The models predict that over 550 000 lung cancer deaths among men
and over 240 000 lung cancer deaths among women were averted
by tobacco control efforts. If all smokers had quit in response to the
Surgeon General’s first report in 1964, over 1.6 million lung cancer
deaths might have been averted among men, and over 880 000
Tobacco control efforts do appear to have reduced smoking behav-
ior and lung cancer deaths.
There is some variation in the estimates among models, depending
on data source and calibration. Data were not available for other
exposures related to lung cancer incidence and non-cigarette forms
of tobacco use were not considered. The relative contributions of
decreased smoking initiation and increased smoking cessation
were not addressed.
From the Editors
JNCI | Articles 543
during the period 1975–2000 for individuals born between 1890
and 1970. As an example, we show one output of the smoking
history generator, the proportion of current smokers in the three
smoking scenarios (Figure 2). For the ATC scenario, the initial
increase in smoking prevalence for each cohort results primarily
from initiation and the subsequent decline reflects both smoking
cessation and increased mortality among smokers. Deaths from
causes other than lung cancer were also simulated by the level of
smoking; each individual history was terminated at 84 years or at
the age of death from a cause other than lung cancer if the death
occurred before 84 years. Details of the construction of the
smoking history generator can be found on the CISNET website
(http://cisnet.cancer.gov and supplementary material, available
online). Using the outputs of the smoking history generator, each
group estimated the number of lung cancer deaths during the pe-
riod 1975–2000 while adjusting for other-cause mortality in each
The models yielded a range of results for the numbers of lung
cancer deaths among the three smoking scenarios, but the esti-
mates of the fraction of lung cancer deaths averted were reasonably
consistent across models. For purposes of illustration, we chose
one of the models calibrated against the US population data (the
Yale model) as an exemplar model to present our results (Figure 3).
Based on this model, we could estimate the actual age-adjusted
rates and the actual number of lung cancer deaths among men and
women aged 30-84 years in the United States during the period
1975–2000 and the numbers that would have been expected assuming
the NTC and CTC scenarios. During the period 1975–2000, there
were 2 067 775 lung cancer deaths among men and 1 051 978 lung
cancer deaths among women in the United States. Assuming NTC
conditions, the Yale model estimated 2 670 897 lung cancer deaths
among men and 1 273 151 lung cancer deaths among women,
whereas assuming CTC conditions, the deaths numbered 958 862
and 438 858, respectively, among men and women. The difference
between the NTC and observed numbers provides an estimate of
the numbers of lung cancer deaths averted (A), which for the Yale
model are 603 122 and 221 173 for men and women, respectively. The
difference between NTC and CTC (B) is an estimate of the total
number of lung cancer deaths that could have been averted if tobacco
control efforts had been immediately and completely successful with
all smoking ending in 1965, which were approximately 1 712 035 and
834 293 for men and women, respectively, in the Yale model. The
averages for these figures across all models are 1 620 686 and 883 356.
We also examined the difference between the NTC and observed
numbers to obtain an estimate of the numbers of lung cancer
deaths averted in all models. Approximately 795 851 US lung cancer
deaths were averted. The high estimate for men was 658 529, the
low estimate was 454 517, and the average was 552 574 (Table 1).
For women, the high estimate was 333 976, the low estimate was
201 788, and the average was 243 277. When we looked at the
difference between NTC and CTC, again, as an estimate of the
total number of lung cancer deaths that could have been averted
if tobacco control efforts had been immediately and completely
successful, we estimated that approximately 2 504 042 lung cancer
smokers quit permanently at that time and there was no initiation
of smoking after 1964. Clearly, the CTC scenario represents the
best imaginable outcome for smoking behavior. Our justification
for using this scenario is that it is transparent and unambiguous.
More details about the construction of these scenarios are on the
Cancer Intervention and Surveillance Modeling Network (CISNET)
web site (http://cisnet.cancer.gov last accessed Feb 19, 2012) and
in a forthcoming group of articles (22).
The models used in CISNET did not incorporate other known
or suspected risk factors, such as environmental tobacco smoke,
radon exposure, diet, and air pollution (23–25), which could have
influenced trends in lung cancer mortality in the United States.
Moreover, the datasets from which the parameters of the individual
dose–response modules were estimated may not be representative
of the US population. For these reasons, the outputs of the models
under the ATC scenario cannot be expected to reproduce the
observed lung cancer rates in the US population. Rather, without
further calibration, these models estimate lung cancer rates in
hypothetical populations with the same smoking behaviors and the
same age structure as the US population in 1975–2000. To com-
pensate for these limitations, some groups (FHCRC, MGH-HMS,
Yale, PIRE) chose to calibrate their models further to describe
actual deaths in the US population under the ATC scenario during
the period 1975–2000. In the models of the FHCRC, MGH-HMS,
and Yale groups, this calibration was achieved by embedding
the dose–response module in an age–period–cohort model. The
exception was the PIRE model, which used only a period calibration.
Other groups (Erasmus MC, Rice-MDA) chose not to perform
this additional calibration. The overall model structure and the
specific models used by each group are described in greater detail
in the supplementary materials (available online).
For each smoking scenario (ATC, NTC, CTC), the smoking
history generator provided, as its output, detailed smoking histories
Figure 1. Process shared by all models. Population and smoking inputs
were used to develop the smoking history generator, which, in turn,
simulates detailed individual-level smoking and other-cause mortality
histories. These individual histories were used by each of the modeling
groups to estimate lung cancer mortality rates in the population.
544 Articles | JNCI Vol. 104, Issue 7 | April 4, 2012
deaths could have been averted in men and women combined.
This estimate represents the average of the estimates for the
models shown in Table 1. In the year 2000 alone, approximately
70 218 lung cancer deaths were averted: 44 135 among men and
26 083 among women. These numbers are estimated to represent
approximately 32% of lung cancer deaths that could have potentially
been averted during the period 1975–2000, 38% of the lung cancer
deaths that could have been averted in 1991–2000, and 44% of
lung cancer deaths that could have been averted in 2000.
The other models calibrated to US mortality yielded similar
estimates of the number of lung cancer deaths among the three
scenarios (data not shown). Counts of the differences in the
number of deaths between scenarios are shown for all models in
Table 1 and in Supplementary Figure 1 (available online). The
ratio of deaths averted to total deaths that could potentially have
been avoided (ie, A/B) is also presented in Table 1. The models
estimate that of all avoidable deaths from smoking-related lung
cancer, between 24% and 32% among women and between 30%
and 37% among men were actually averted as a result of the
changes in smoking behaviors that actually began in the mid-1950s
some years before the first Surgeon General’s Report. For both
sexes combined, approximately 32% (28%–35% across models) of
all avoidable deaths were averted. Table 1 also shows the impact
of tobacco control efforts on lung cancer mortality for the decade
1991–2000 and for the year 2000. In the decade 1991–2000, the
fraction of lung cancer deaths averted in men and women combined
increased to about 38% (34%–43% across models). In the year 2000,
this fraction increased to roughly 44% (39%–50% across models).
The increasing trend in the fraction of lung cancer deaths averted
reflects both changes in smoking behaviors and a continuing decrease
in risk among former smokers.
A consortium of six research groups used data from common
sources to recreate detailed cigarette smoking histories under three
Figure 2. Percentage of current smokers in the US population by sex and birth cohort, assuming three different tobacco control scenarios. This is
one of the outputs that can be generated from the smoking history generator. The output from the actual tobacco control scenario describes the
observed data well (not shown).
JNCI | Articles 545
distinct tobacco control scenarios as inputs for mathematical models
to quantify the impact of changing smoking behavior on lung cancer
mortality rates in the United States during 1975–2000. We used a
comparative modeling approach to address this complex problem;
comparative modeling produces a range of results across models
but, when these are reasonably consistent, enhances their credibility.
During the period 1975–2000, approximately 2 504 042 lung cancer
deaths among men and women combined could have been averted
had tobacco control efforts been completely effective in eliminating
smoking as of 1965; of these, we estimate that approximately 795 851
lung cancer deaths were averted or about one-third of what was
possible. During the period 1991–2000, we estimate that approxi-
mately 345 000 lung cancer deaths among US men and 175 000
deaths among US women were averted due to changes in smoking
behaviors starting in the mid-1950s. These estimates of reduced
lung cancer mortality associated with reduced tobacco use are much
larger than an estimate from demographic projections that 146 000
lung cancer deaths among men were averted in 1991–2003 (5). We
estimate that in the year 2000 alone, approximately 44 000 deaths
were averted among US men and 26 000 deaths among US women.
It is not surprising that the various models used in this article
yielded a range of estimates of the fraction of lung cancer deaths
averted by the tobacco control efforts in the United States. This
range of results represents the uncertainty associated with model
choice. First, some of these models were calibrated against US
mortality data, and, as a consequence, these models describe the
lung cancer mortality trends in the United States very well under
the ATC scenario. Second, although five of the six groups used the
TSCE version of multistage models (8,10,11) as the dose–response
module, the estimated parameters were different because they were
estimated by their fit to different cohorts. In addition to the TSCE
model, the Yale group also used the models developed by Knoke
et al. (9) and Flanders et al. (26) and obtained similar estimates of
the relative effect of tobacco control. It is well known that the risks
of tobacco smoking have changed over time; moreover, they could
be modified by other factors such as diet that are not accounted for
in any of the models. Despite these limitations, the estimated
numbers of deaths averted and deaths that could have been averted
under the assumption of CTC were reasonably consistent across
models (Table 1). The main message of these analyses is clear.
Tobacco control strategies implemented mid-century have averted
hundreds of thousands of lung cancer deaths in the United States
during the period 1975–2000, but these are only approximately
30% of the lung cancer deaths that could have been averted had all
cigarette smoking ended in 1965.
The FHCRC, MGH-HMS, and Yale groups calibrated their
models to US mortality during 1975–2000 using birth cohort and
period effects. These calibrations are necessary to describe lung
cancer mortality rates and trends in the United States and indicate
that the lung cancer mortality experience of the entire population
cannot be adequately described by extrapolating from the SEER
registry in one decade, or from various cohort and case–control
Figure 3. Lung cancer death rates and counts for men and women aged 30-84 years as observed and for modeled tobacco control scenarios.
ATC = Actual Tobacco Control; CTC = Complete Tobacco Control; NTC = No Tobacco Control.
546 Articles | JNCI Vol. 104, Issue 7 | April 4, 2012
studies of smoking and lung cancer (please see the supplementary
material, available online, for the datasets used by each of the
groups for parameter estimation). Particularly among men, US lung
cancer mortality is considerably higher than would be expected from
the cohort studies against which the dose–response modules were
calibrated. In addition, models from cohort studies and available popu-
lation smoking histories cannot adequately describe temporal compo-
nents of trend, that is, the effects of age, period, and birth cohort.
There could be several reasons why the models were poor
at predicting population lung cancer rates without additional
calibrations. First, the datasets used for estimating the parameters
of the dose–response modules were almost certainly not represen-
tative of the US population. Second, the smoking history generator
was based on smoking histories for birth cohorts in the general
population that were inferred from simulations using cross-sectional
histories that often relied on subjects’ recall of events that occurred
several years earlier. Third, potentially important covariates (eg, diet,
air pollution, and radon exposure) and occupational exposures
(including asbestos and ionizing radiation) were not available for the
overall population, and different exposure distributions could con-
tribute to rate discrepancies. Fourth, although the models discussed
assume a consistent effect of exposure on lung cancer mortality,
temporal changes in the manufacture of cigarettes and smoking
behaviors could explain some of the discrepancies in trend, and data
on changes in cigarette manufacturing and composition are not
readily available. Changes in tobacco or cigarette composition, which
were not explicitly addressed in these analyses, could be important
contributors to population trends in lung cancer mortality. However,
one would expect changes in tobacco or cigarette composition to
manifest themselves as period effects, whereas models that used
age–period–cohort calibrations find that trends are dominated by
birth cohort effects. Finally, uncertainty remains with respect to
the models themselves.
In particular, our estimates of the lung cancer rates in the
US population under the CTC scenario appear to be higher than
would have been expected on the basis of recent work on lung
cancer rates among never smokers (27). However, for the reasons
given above (cohorts not representative of the general population,
omission of important covariates), the never-smoker rates reported
by Thun et al. (27) may not reflect the never-smoker rates in the
general population. Some confidence in the lung cancer rates under
the CTC scenario estimated from the models in this article can be
derived from the fact that the dose–response modules describe
lung cancer rates among former smokers well (9,10,11).
One limitation of the calibrations is that the same period and
cohort parameters are applied to current smokers, former smokers,
and never smokers. Factors, such as diet, that could affect trends
in lung cancer rates might be expected to have different effects
among current smokers, former smokers, and never smokers.
However, different cohort and period effects could not be estimated
Table 1. Realized and potential reductions in lung cancer mortality from changes in smoking behavior among men and women aged
of potential benefit
control, by year(s)
1 064 443
1 757 857
1 680 867
1 597 733
1 329 972
1 645 651
1 712 035
1 620 686
2 564 177
2 543 477
2 451 845
2 394 415
2 524 010
2 546 328
2 504 042
1 296 837
1 363 642
1 357 809
1 552 462
1 246 727
1 378 358
1 365 972
* ATC = Actual Tobacco Control; CTC = Complete Tobacco Control; NTC = No Tobacco Control. The realized benefits of ATC are estimated by the difference
(NTC–ATC); the potential total benefits are estimated by the difference (NTC–CTC); the proportion realized is given by the quotient of realized benefits and total
potential benefits: (NTC–ATC)/(NTC–CTC). The six study groups that produced models are as follows: Erasmus MC = Erasmus Medical Center, Rotterdam, the
Netherlands; FHCRC = Fred Hutchinson Cancer Research Center, Seattle, WA; MGH-HMS = Massachusetts General Hospital and Harvard Medical School,
Cambridge, MA; PIRE = Pacific Institute for Research and Evaluation, Calverton, MD; Rice-MDA = Rice University and M.D. Anderson Cancer Center, Houston,
TX; and Yale = Yale University, New Haven, CT.
JNCI | Articles 547
in these subgroups because of identifiability issues. The FHCRC
group did fit period and cohort effects to never smokers alone and
to current smokers and former smokers separately, but the original
model in which these effects are applied equally to all groups
described the data better as judged by the Akaike Information
Criterion Statistics (28).
Overall, our study shows that changes in smoking behaviors led
to a substantial reduction in the lung cancer mortality that would
have been expected had the smoking trends in the 1950s continued
into the future. Our analysis was conducted through to the year 2000,
the latest year for which we were able to obtain sufficiently detailed
data when this project was initiated. Consistent with trends for
continued gains due to past tobacco control policies, smoking
prevalence continued to fall from 23.2% in 2000 to 20.6% in 2008.
Much of this decrease can be attributed to tobacco control policies,
especially the cigarette price increases in 1998–1999 (29).
There are also other limitations to our study. We did not
quantitatively assess the relative contributions made by changing
patterns of smoking initiation and cessation to decreases in lung
cancer mortality. It is clear, however, that most of the benefits
of tobacco control policies during the period 1975–2000 have
accrued from smoking cessation because changing patterns of
smoking initiation would have impacted only individuals who
were aged 55 or younger in 2000 and thus younger than the age
at which lung cancer mortality begins to increase rapidly. Also,
our numbers are likely to greatly underestimate the overall health
impact of tobacco control efforts because they neither consider
the substantial impact of non-cigarette forms of tobacco use (eg,
cigars and pipes) nor the impact of tobacco smoking behaviors on
diseases other than lung cancer. Smoking-associated diseases other
than lung cancer, such as cardiovascular disease, were outside the
scope of this work.
The results of this article show the dramatic impact of the
reduction in smoking associated with tobacco control efforts in the
second half of the 20th century on lung cancer mortality during the
period 1975–2000. Even though other factors, including genetic
polymorphisms (27), contribute to lung cancer risk, the vast majority
of lung cancer cases could be eliminated by eliminating smoking. Our
results indicate that only approximately 30% of the total lung
cancer deaths that could have been averted had tobacco control been
complete were actually averted. This is because smoking rates took
time to decline after the first Surgeon General’s Report in 1965;
smokers’ risk of lung cancer remains elevated for many years after
smoking cessation and a sizable fraction of the population continued
Clearly, further reductions in smoking rates will be required to
reduce lung cancer incidence and mortality rates substantially. The
recently reported 20% reduction in lung cancer mortality (30)
as a result of early detection using low-dose spiral CT suggests that
screening of high-risk individuals may play a role in reducing
mortality from this disease. Because risk of lung cancer remains
elevated for a long time among smokers who quit, effective screening
techniques may have a role in reducing lung cancer mortality
among ex-smokers. However, continued implementation of evidence-
based tobacco control policies, programs, and services remains
the most promising approach to reducing the burden of lung
1. Cokkinides V, Bandi P, McMahon C, Jemal A, Glynn T, Ward E.
Tobacco control in the United States—recent progress and opportunities.
CA Cancer J Clin. 2009;59(6):352–365.
2. Reducing Tobacco Use. Rockville, MD: Office of the Surgeon General, U.S.
Department of Health and Human Services; 2000. . http://www.cdc.gov/
tobacco/data_statistics/sgr/2000. Accessed January 18, 2011.
3. Doll R, Peto R, Boreham J, Sutherland I. Mortality in relation to smoking:
50 years’ observations on male British doctors. BMJ. 2004;328(7455):
4. The Health Consequences of Smoking: A Report of the Surgeon General. [Atlanta,
GA]. Washington, DC: Dept of Health and Human Services, Centers for
Disease Control and Prevention, Office of Smoking and Health; 2004.
5. Thun MJ, Jemal A. How much of the decrease in cancer death rates in
the United States is attributable to reductions in tobacco smoking? Tob
6. McMahon PM, Kong CY, Weinstein MC, et al. Adopting helical CT
screening for lung cancer: potential health consequences during a 15-year
period. Cancer. 2008;113(12):3440–3449.
7. McMahon PM, Kong CY, Johnson BE, et al. Estimating long-term effec-
tiveness of lung cancer screening in the Mayo CT screening study. Radiology.
8. Moolgavkar SH, Knudson AG. Mutation and cancer: a model for human
carcinogenesis. J Natl Cancer Inst. 1981;66(6):1037–1052.
9. Knoke JD, Shanks TG, Vaughn JW, Thun MJ, Burns DM. Lung cancer
mortality is related to age in addition to duration and intensity of cigarette
smoking: an analysis of CPS-I data. Cancer Epidemiol Biomarkers Prev.
10. Hazelton WD, Clements MS, Moolgavkar SH. Multistage carcinogenesis
and lung cancer mortality in three cohorts. Cancer Epidemiol Biomarkers
11. Meza R, Hazelton WD, Colditz GA, Moolgavkar SH. Analysis of lung
cancer incidence in the Nurses’ Health and the Health Professionals’
Follow-Up studies using a multistage carcinogenesis model. Cancer Causes
12. Little MP. Are two mutations sufficient to cause cancer? Some general-
izations of the two-mutation model of carcinogenesis of Moolgavkar,
Venzon, and Knudson, and of the multistage model of Armitage and Doll.
13. Kopp-Schneider A. Carcinogenesis models for risk assessment. Stat
Methods Med Res. 1997;6(4):317–340.
14. Heidenreich WF, Wellmann J, Jacob P, Wichmann HE. Mechanistic
modelling in large case-control studies of lung cancer risk from smoking.
Stat Med. 2002;21(20):3055–3070.
15. Luebeck EG, Moolgavkar SH. Multistage carcinogenesis and the incidence
of colorectal cancer. Proc Natl Acad Sci U S A. 2002;99(23):15095–15100.
16. Meza R, Jeon J, Moolgavkar SH, Luebeck EG. Age-specific incidence of
cancer: phases, transitions, and biological implications. Proc Natl Acad Sci
U S A. 2008;105(42):16284–16289.
17. Meza R, Jeon J, Renehan AG, Luebeck EG. Colorectal cancer incidence
trends in the united states and united kingdom: evidence of right- to left-sided
biological gradients with implications for screening. Cancer Res. 2010;70(13):
18. Doll R, Peto R. Cigarette smoking and bronchial carcinoma: dose and
time relationships among regular smokers and lifelong non-smokers.
J Epidemiol Community Health. 1978;32(4):303–313.
19. Rachet B, Siemiatycki J, Abrahamowicz M, Leffondre K. A flexible modeling
approach to estimating the component effects of smoking behavior on lung
cancer. J Clin Epidemiol. 2004;57(10):1076–1085.
20. Schollnberger H, Manuguerra M, Bijwaard H, et al. Analysis of epide-
miological cohort data on smoking effects and lung cancer with a multi-stage
cancer model. Carcinogenesis. 2006;27(7):1432–1444.
21. Deng L, Kimmel M, Foy M, Spitz M, Wei Q, Gorlova O. Estimation of
the effects of smoking and DNA repair capacity on coefficients of a carci-
nogenesis model for lung cancer. Int J Cancer. 2009;124(9):2152–2158.
22. The impact of tobacco control efforts on US lung cancer mortality:
1975-2000. In: Moolgavkar SH, Feuer EJ, Levy DT, Marek K, eds. Risk
548 Articles | JNCI Vol. 104, Issue 7 | April 4, 2012 Download full-text
23. Krewski D, Burnett RT, Goldberg MS, et al. Reanalysis of the Harvard
Six-Cities study and the American Cancer Society Study of Particulate Air
Pollution and Mortality, Part II: Sensitivity Analysis. A Special Report of the
Institute’s Particle Epidemiology Reanalysis Project. Cambridge, Massachusetts:
Health Effects Institute; 2000.
24. Darby S, Hill D, Auvinen A, et al. Radon in homes and risk of lung cancer:
collaborative analysis of individual data from 13 European case-control
studies. BMJ. 2005;330(7485):223.
25. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report
of the Surgeon General. Rockville, MD: Office of the Surgeon General, U.S.
Department of Health and Human Services; 2006. http://www.cdc.gov/
tobacco/data_statistics/sgr/2006. Accessed January 18, 2011.
26. Flanders WD, Lally CA, Zhu BP, Henley SJ, Thun MJ. Lung cancer
mortality in relation to age, duration of smoking, and daily cigarette
consumption: results from Cancer Prevention Study II. Cancer Res. 2003;
27. Thun MJ, Hannan LM, Adams-Campbell LL, et al. Lung cancer occur-
rence in never-smokers: an analysis of 13 cohorts and 22 cancer registry
studies. PLoS Med. 2008;5(9):e185.
28. Sakamoto Y, Ishiguro M, Kitagawa G. Akaike Information Criterion
Statistics. Tokyo, Japan: KTK Scientific Publishers; 1986.
29. Levy DT, Nikolayev L, Mumford E. Recent trends in smoking and the
role of public policies: Results from the SimSmoke tobacco control policy
simulation model. Addiction. 2005;100(10):1526–1536.
Accessed January 18, 2011.
The research reported in this article was funded by the National Cancer
Institute’s Cancer Intervention and Surveillance Modeling Network (CISNET;
http://cisnet.cancer.gov/). Specifically, this work was supported by grants
from the National Cancer Institute at the National Institutes of Health
(5U01CA097415-04 to SHM, 2U01CA097432-04 to TRH, 5U01CA097450-
04 to DTL, 5U01CA097416-04 to RB and HJK, 2U01CA097431-04 to MK,
1U01CA152956-01 to PMM HJK, DTL, SHM, TRH). GSG acknowledges
support from the National Cancer Institute at the National Institutes of
Health (5R01CA097337-02) and the American Cancer Society (ACS RSG
2008A060554), PMM from the National Cancer Institute at the National
Institutes of Health (R00CA126147), and CYK from the National Cancer
Institute at the National Institutes of Health (K25CA133141).
The sponsoring agencies had no role in designing and conducting the research
and in the collection and analyses of the data. The sponsoring agencies did not
participate in the writing of the article and had no role in the decision to publish.
Dr R. Boer is affiliated with RAND and is a consultant at Verner
LifeSciences. Dr G. S. Gazelle is a consultant for GE Healthcare. However, the
authors declare no conflicts of interest.
Affiliations of authors: Program in Biostatistics and Biomathematics, Fred
Hutchinson Cancer Research Center, Seattle, WA (SHM, JJ, WDH, RM);
Department of Epidemiology and Public Health, Yale University School of
Medicine, New Haven, CT (TRH); Department of Economics, University of
Baltimore, Baltimore, MD (DTL); Pacific Institute for Research and Evaluation,
Calverton, MD (DTL); Massachusetts General Hospital, Boston, MA (CYK,
GSG, PMM); Harvard Medical School, Boston, MA (CYK, GSG, PMM); The
Brown Foundation Institute of Molecular Medicine, University of Texas Health
Science Center at Houston, Houston, TX (MF); Department of Epidemiology,
UT MD Anderson Cancer Center, Houston, TX (OG); Department of Statistics,
Rice University, Houston, TX (MF, MK); Cornerstone Systems Northwest, Inc,
Lynden, WA (LC); Department of Epidemiology, University of Michigan, Ann
Arbor, MI (RM); Department of Public Health, Erasmus MC, Rotterdam, the
Netherlands (FS, RB, HJdK); Department of Health Services, School of Public
Health and Department of Psychology, University of California, Los Angeles,
CA (WM); Systems Engineering Group, Silesian University of Technology,
Gliwice, Poland (MK); Division of Cancer Control and Population Sciences,
National Cancer Institute, Bethesda, MD (EJF); Lombardi Comprehensive
Cancer Center, Georgetown University, Washington, DC (DTL).