Abstract
BACKGROUND: Despite an overall decline in tobacco use in the United States, secular trends of smoking and nicotine dependence with co-occurring substance use are not well characterized. METHODS: We examined self-reported tobacco and other substance use in 22,245 participants age 21-59 in the United States from six waves of the National Health and Nutrition Examination Survey (NHANES). Using Joinpoint regression, we assessed secular trends of smoking and nicotine dependence as a function of co-occurring use of alcohol, prescription opioids, marijuana/hashish, cocaine/heroin/methamphetamine, or other injection drug use. Multivariable logistic regressions were fitted to identify the potential risk factors. RESULTS: During 2005-2016, the prevalence of current smoking decreased (without co-occurring substance use: 17.0 %-12.7 %; with co-occurring use of one substance: 35.3 % to 24.6 %; with co-occurring use of two or more substances: 53.8 %-42.2 %), and moderate-to-severe nicotine dependence decreased as well (8.0 %-4.2 %, 16.0 %-8.8 %, and 23.9 %-15.7 %, respectively). Smoking and nicotine dependence were more likely in those with co-occurring use of one substance (current smoking: odds ratio [OR] = 2.22, 95 % confidence interval [CI] = 2.01-2.45); nicotine dependence: OR = 1.88, 95 % CI = 1.63-2.17) and in those with co-occurring use of two or more substances (current smoking: OR = 5.25, 95 % CI = 4.63-5.95; nicotine dependence: OR = 3.24, 95 % CI = 2.72-3.87). CONCLUSIONS: Co-occurring substance use was associated with smaller reductions in tobacco use, over time, and with increased odds of nicotine dependence. This suggests that co-occurring substance users should be regarded as a tobacco-related disparity group and prioritized for tobacco control interventions.