Abstract
BACKGROUND: In recent decades, smoking has become an increasingly non-normative behavior. Because deviant behaviors are associated with greater clinical and genetic risks, current-generation smokers may have greater concentrations of psychiatric comorbidity than previous generations. We examined this question empirically by testing whether associations between measures of smoking, psychiatric diagnoses, and risk-associated personality traits, increased across seven birth-cohorts of the 20th century. METHOD: 4326 subjects from a cross-sectional NIMH control sample were categorized into one of seven groups based on birth (born before 1930, and 1930s-80s) and one of three smoking levels (lifetime dependent smoker, never dependent smoker, never smoker). Smoking and ND were assessed using the Fagerstrom Test for Nicotine Dependence; psychiatric diagnoses (drug and alcohol dependence, major depression, and generalized anxiety disorder) using the Composite International Diagnostic Interview-Short Form, and personality traits (neuroticism and extraversion) with the Eysenck Personality Questionnaire. RESULT: Lifetime prevalence of smoking decreased across the seven cohorts. Associations between smoking and drug dependence, generalized anxiety, and neuroticism, as well as total psychiatric comorbidity, were greater in more recent cohorts [smoking-by-cohort interaction: p<0.01], with greatest increases contributed by nicotine-dependent smokers. Smoking was also independently associated with alcohol dependence and depression, but these associations did not significantly vary across cohorts. CONCLUSIONS: More recent generations included fewer persons who smoked, but their smoking was associated with greater psychiatric morbidity. Failure to account for systematic variation in comorbidity across smoking cohorts may lead to unwanted heterogeneity in clinical, and possibly genetic, studies of nicotine dependence.