Lung cancer rates as an index of tobacco smoke exposures: validation against black male∼non-lung cancer death rates, 1969–2000
Introduction
The lung cancer death rate is perhaps the best available index of the cumulative hazards of tobacco smoking (smoking) according to the World Health Organization [1], tobacco industry [2], [3] cancer society, and government-supported researchers [4], [5], [6], [7]. Yet that index or exposure load measure has not been directly validated. Nor has the index been directly calibrated against observed exposure/disease associations in geographically representative populations. Nor have the smoking-attributable fractions (SAFs) of cancer death rates been directly estimated. Instead, estimates of smoking-attributable cancer or all-cause death numbers or percents have most often been indirectly calibrated using results from a nearly 20-year-old cohort study of disproportionately healthy, educated, white, volunteer Americans [4], [5], [6], [7]. Such death SAFs may be misleadingly low due to selection and misclassification biases [8], and ignoring the marked prematurity of many deaths from smoking. Studies of smoking-attributable age-adjusted death rates in representative populations with continuous exposure monitoring are needed.
To best calibrate lung cancer death rates as an index of exposure load, one should compare highly representative populations that are nearly identical with the exception of marked smoke exposure disparities. In particular, those populations should have very little medical-care-related mortality disparity. A time series study in a highly underserved population undergoing a large tobacco/lung cancer epidemic may meet all of the above criteria—comparability of the more and less exposed populations (in this case, a single population over time), marked exposure disparities, and minimization of medical care effects.
Both direct validation and calibration of the smoke exposure index and quantitative assessment of the cause of the 1969–1990 cancer death epidemic in United States (U.S.) black men are greatly needed. So this study performs the first such direct index validation, index calibration, and epidemic assessment. Black men were studied due to their relative deficits of medical care and immigration, and their disproportionately large and recent cancer death epidemic [9]. Over 30 years of annual cancer death rates were studied since it is unlikely that annual changes in cumulative hazards from other carcinogens would consistently shadow annual smoke exposure changes throughout so long a period.
Specifically, this study assesses the steep 1969–1990 71% and 34% rises in the black male lung and ∼non-lung (all-sites excepting lung and stomach) cancer death rates, respectively. Then, it assesses the steep 1990–2000 19% and 11% drops in the black male lung and ∼non-lung cancer death rates, respectively (http://www.seer.cancer.gov/canques). Those results are used to help estimate upper- and lower-bound SAFs of all-sites cancer death rates in U.S. black men in 1990 at the peak of the cancer death epidemic, and more recently in 2000.
Section snippets
Methods
To increase the power of the study, just black men are included. Their cancer death rate increases were more recent and steep than whites or women, and less likely to be affected by medical care since black men are particularly underserved [10]. This is supported by past findings that 80–82% of tobacco-related cancers in California's lowest socioeconomic groups are still initially diagnosed at late stages as late as 1996–2000 [11]. Also, men did not benefit from recent likely treatment-related
Results
U.S. black males' ∼non-lung cancer annual death rates are graphed versus year in Fig. 1 and versus the annual smoke exposure load (lung cancer rate) in Fig. 2. The ∼non-lung cancer rates were nearly perfectly correlated (R2 = 0.97+) with their smoke exposure loads, as measured by their lung cancer death rates (Fig. 2). The near-perfect correlations held both during smoke exposure loads' steep increases from 1969 to 1990 [R2 = 0.98; P < 0.0001; y = 1.20x + 101; 95% confidence interval (CI) for
Discussion
During two decades of steep rises, and a subsequent decade of steep falls, U.S. black male smoke exposure loads and ∼non-lung cancer death rates have moved in near-perfect lockstep up and starting down. The smoke exposure load/non-lung cancer death rate associations are strong, dose-response, biologically plausible, and reversible. The associations have been consistent year-by year for over 30 years, with the modest explicable exception of the years 1988–1992 as the introduction of
References (24)
Cancer, cigarette smoking and premature death in Europe: a review including the Recommendations of European Cancer Experts Consensus Meeting, Helsinki, October 1996
Lung Cancer
(1997)- et al.
The changing cigarette
Prev. Med.
(1997) The world health report 2002
(2002)- et al.
The fifty-year decline of cancer in America
J. Clin. Oncol.
(2001) - et al.
Impact of the American anti-smoking campaign on lung cancer mortality
Int. J. Cancer
(2002) - et al.
Mortality from tobacco in developed countries: 1950–2000
(2003) - et al.
Mortality from smoking worldwide
Br. Med. Bull.
(1996) - et al.
Lung cancer trends in young adults: an early indicator of progress in tobacco control
Cancer Causes Control
(2003) - et al.
Measuring the accumulated hazards of smoking: global and regional estimates for 2000
Tob. Control
(2003) - et al.
Benefits of smoking cessation for longevity
Am. J. Public Health
(2002)