Electronic Thesis/Dissertation


Light and Intermittent Smokers—Sociodemographic, Smoking and Other Behavioral Trends, and Lung Cancer Risk Open Access

Downloadable Content

Download PDF

Background Light and/or intermittent smokers have been the fastest growing segment of smokers in the U.S. during the past 15 years. National survey data indicates the prevalence of nondaily smokers ranges from about 20% to nearly 40%. Researchers continue to use diverse definitions of light and/or intermittent smokers: Chippers, occasional smokers, some-day smokers, and light or very light smokers. It is likely however, that smokers who fall in these categories comprise a heterogeneous mix. Defining their smoking behaviors, mental health characteristics and lung cancer risk is a critical public health priority. MethodsObjectives The overall purpose of this research was to characterize light and/or intermittent smoking by studying the demographic, smoking and other behavioral patterns of these smokers, as well as the association to lung cancer. This dissertation is divided into two overall objectives: 1) using publicly available aggregate data, examine the characteristics of the growing numbers of light, intermittent, and LITS (dual light and intermittent smokers) by describing their demographic distributions, smoking behaviors and mental health conditions in a variety of population-based surveys; and 2) investigate the association of light and/or intermittent smoking to lung cancer risk, including potential covariates. Data Sources The first research objective was studied using three population surveys: the National Health Information Survey, the National Health and Nutrition Examination Survey, and the National Survey on Drug Use and Health. Lung cancer risk among light and intermittent smokers was examined using data from two studies by the National Cancer Institute: the PLCO Trial and the EAGLE study.Analyses Our sample included 77,452 U.S. adults from three pooled contemporary population-based surveys: the 2012 NHIS, 2012 NSDUH, and 2011-2012 NHANES. We classified current smokers into four categories and assessed associations of these smoking groups with smoking behaviors, illicit drug use, and mental health indicators. For the second analysis, the constructs of light and intermittent smoking were measured by confirmatory factor analysis (CFA) modeling using the smoking questions in these surveys. Additionally, a Cox Proportional Hazards model was fit to assess lung cancer risk among light smokers in the PLCO trial using 37,000 U.S. men and 37,000 U.S. women screened for prostate, ovarian, lung and colorectal cancers. Further, an unconditional logistic regression model was fit to measure lung cancer risk among light and intermittent smokers from 2,100 incident lung cancer cases and 2,120 controls in a European population-based case-control study.ResultsSpecific demographic, smoking and behavioral associations among categories of light and/or intermittent smokers were detected, particularly related to nicotine dependence, age of smoking initiation and race/ethnicity. Compared to heavier-daily smokers, LITS were most likely to have mild or no nicotine dependence (odds ratio [OR] = 16.92, 95% confidence interval [CI] 13.10–21.85), to start smoking cigarettes regularly after age 21 (OR = 3.42, 95% CI 2.84–4.12), and to be Hispanic (OR = 5.38, 95% CI 4.38–6.61). Additional significant findings were observed for other categories of smokers.Also, a 2-factor structure from a Confirmatory Factor Analysis model using data from NHANES showed an acceptable fit (CFI = 1.0, TLI = 0.999, RMSEA = 0.015, χ2 = 8.97 and p-value = 0.0619), while a one-factor structure using data from NSDUH showed good fit indices (CFI = 0.995, TLI = 0.985, RMSEA = 0.025, χ2 = 13.86 and p-value = 0.0010).In our last analysis, light smoking was associated with a lower risk for lung cancer compared to heavy smokers (HR = 0.32, 95% CI 0.25–0.42) in the U.S. cohort as well as in the European cohort (OR = 0.29, 95% CI 0.15–0.55). Additional significant associations were observed for other smoking behaviors and depression/anxiety.ConclusionBased on pooled data from three large national surveys, light and/or intermittent smokers differed in smoking, drug use and mental health behaviors from heavier-daily, former and never smokers. Notable differences by level of smoking frequency and intensity were observed for nicotine dependence, age of smoking initiation and race/ethnicity.Our findings also demonstrate that light and intermittent smoking have distinct latent structures. Additionally, incorporating the smoking biomarker data from NHANES provides additional information to separate the constructs of light and intermittent smoking. Our work contributes to a better understanding of the heterogeneity in this group of smokers, to develop more targeted interventions in areas ranging from smoking prevention to cessation, to reduce the burden of mental health conditions and substance abuse.We also demonstrated that light and intermittent smoking are each associated with lung cancer. Complete smoking cessation may be the best tool to eliminate lung cancer risk, as reducing the amount of cigarette consumption disproportionally reduces, but does not eliminate risk.

Author Language Keyword Date created Type of Work Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL