Lung Cancer Mortality Among Smokers and Never-Smokers in the United States.

Abstract

To the Editor: Lung cancer in never-smokers may be increasing in the United States.1 To further explore these temporal trends, we estimated lung cancer mortality from 1992 to 2011, by linking the National Longitudinal Mortality Study (NLMS)2 and Tobacco Use Supplements (TUS). The NLMS is a prospective study of mortality in combined samples of the civilian noninstitutionalized population.2 NLMS is managed by the Census Bureau and consists of responses to Current Population Surveys (CPS) (response rate >90%),3 linked to mortality data from the National Death Index, which is maintained by the National Center for Health Statistics.4 The CPS supplement is a weighted, national probability sample of households that provides demographic, economic, and social information about the US population. NLMS consists of ≈3.8 million records with >550,000 deaths during 1992–2011. The TUS in CPS collects nationally representative information regarding tobacco use from ≈240,000 adults every 3–4 years since 1992.5 Participants with CPS and TUS responses were eligible for analysis. Never-smokers were defined as those having smoked ≤100 cigarettes in their lifetime. Lung cancer deaths were identified using underlying cause as the cause of death. Age-adjusted mortality rates (per 100,000; adjMRLC) and 95% confidence intervals (CIs), based on the 2000 Standard US population, were estimated using SASv9.4 (Cary, NC). Annual percentage change (APC) in adjMRLC was estimated by the best fit model using joinpoint regression (Joinpointv4.6.0.0, Bethesda, MD). We identified ≈4,900 lung cancer deaths (rounded to nearest 100 per Census regulations). A majority of deaths (86.5%) were among smokers and 13.5% among never-smokers. From 1992 to 2011, the adjMRLC was 86.5 (69.2–106.7), with rates higher among males [116.3 (96.1–139.4)] compared with females [65.6 (50.7–83.5)]. Temporal trends show rates trending downward for males [APC = −1.6 (−2.2 to −1.0)] and, conversely, upward for females [APC = 0.8 (0.3–1.2)] since 1992. Further, those who attained a high school diploma only also experienced a positive APC [=0.9 (0.4–1.3)] since 1992. The adjMRLC suggested an increase over time by ≈20% among smokers [1992–2001: 144.7 (122.1–170.3); 2002–2011: 171.3 (146.6–199.0)] and among never-smokers [1992–2001: 17.5 (10.3–27.8); 2002–2011: 20.8 (12.8–31.9)]. The increase in lung cancer mortality in never-smokers was also seen in each sex [males: 1992–2001: 17.4 (10.2–27.7); 2002–2011: 20.1 (12.3–31.0), females: 1992–2001: 17.5 (10.3–27.8); 2002–2011: 21.3 (13.2–32.5)]. This pattern was also evident among never-smokers by educational attainment [<high school (HS): 1992–2001: 32.0 (21.9–45.2); 2002–2011: 36.0 (25.2–49.8), HS: 1992–2001: 16.7 (9.7–26.8); 2002–2011: 24.2 (15.5–35.9), >HS: 1992–2001: 10.4 (5.1–18.9); 2002–2011: 12.8 (6.8–22.0)] and annual income [<25,000: 1992–2001: 26.5 (17.4–38.7); 2002–2011: 32.3 (22.1–45.5), 25,000–39,000: 1992–2001: 13.4 (7.2–22.7); 2002–2011: 16.4 (9.4–26.5), >40,000: 1992–2001: 10.1 (4.9–18.5); 2002–2011: 12.5 (6.6–21.6)]. The NLMS-TUS linkage, a nationally representative weighted sample with oversampling of small area and minority groups, is the largest contemporary study of tobacco use and mortality in the United States. We demonstrated that age-adjusted mortality for lung cancer in never-smokers is increasing among males and females despite decreasing mortality rates for lung cancer overall since 1997. Our observed suggestion that rates among never-smokers are increasing over time is consistent with some of the prior literature.6 Our study’s prospective design allows for estimation of mortality trends, which should correlate with incidence given lung cancer’s poor prognosis. Although our lung cancer mortality estimates are consistent with previous NLMS reports,7 they are higher than other national estimates,8 potentially attributed to the increased representation of rural populations in NLMS. Our results suggest lung cancer in never-smokers is a US public health concern. Given the overlapping 95% CIs of the rates for the time periods studied, additional research is warranted to ensure sufficient statistical power to determine if lung cancer in this group is truly increasing. Potentially increasing lung cancer mortality in never smokers indicates the need to further study environmental exposures, their potential interaction with molecular drivers of lung cancer, and how they may inform potential lung cancer screening guidelines in never smokers. ACKNOWLEDGMENTS This paper is released to inform interested parties of research and to encourage discussion. Any views expressed on statistical, methodological, technical, or operational issues are those of the authors and not necessarily those of the US Census Bureau. All results have been reviewed by the Census Bureau’s Disclosure Review Board to ensure that no confidential information is released. The views expressed in this manuscript are those of the authors and do not necessarily represent the views of the National Heart, Lung, and Blood Institute; the National Institutes of Health; or the U.S. Department of Health and Human Services. H. Dean Hosgood, IIIDepartment of Epidemiology andPopulation HealthAlbert Einstein College of MedicineBronx, NY[email protected] Candace CosgroveMortality Research BranchCenter for Administrative Records Researchand Applications United States Census BureauSuitland, MD Madelyn KlugmanThomas RohanDepartment of Epidemiology andPopulation HealthAlbert Einstein College of MedicineBronx, NY Sean AltekruseEpidemiology Branch, Prevention andPopulation Sciences ProgramDivision of Cardiovascular SciencesNational Heart Lung and Blood InstituteNational Institutes of HealthBethesda, MD