Empirically derived latent classes of tobacco dependence syndromes observed in recent-onset tobacco smokers: epidemiological evidence from a national probability sample survey

Abstract

This study pursued a line of large-sample epidemiological research on tobacco dependence syndromes that may appear during the first 2 years of tobacco smoking, as clinical features begin to emerge. Focusing on smokers who just recently started using tobacco may provide insights into the transitions that lead from initial smoking toward tobacco dependence. A specific focus was a possible excess risk of tobacco dependence associated with early-onset smoking. Data came from public use files of the 1995-1998 National Household Surveys on Drug Abuse. Analyses were based on responses from 2,993 smokers, that is, those whose age at onset of tobacco smoking was either equal to the age at the time of the interview (n=1,030) or within 1 year of the age at the interview (n=1,963). Seven clinical features were assessed for a measure of the tobacco dependence syndrome, elicited during a standardized assessment. Findings from latent class analysis best support a model with three classes of smokers; features of tobacco dependence are prominent in just two of these classes, which in aggregate constitute 29% of the recent-onset smokers. Earlier-onset tobacco smokers may have a modestly higher probability of expressing dependence features within 2 years of smoking onset, compared with later-onset smokers (i.e., those starting after age 20). Clinical features of tobacco dependence emerge within 1-2 years after the onset of smoking. If the three-class model of tobacco dependence is correct, early-onset smoking may confer modest excess risk of becoming tobacco dependent during the first 2 years after smoking onset.