Some previous reports that have conducted latent class or cluster modeling of nicotine withdrawal symptoms include the following: Madden et al. (1997) used latent class analysis to examine the psychometric structure of lifetime nicotine withdrawal in a survey of a cohort of adult female twins (age 32–48). Three nicotine withdrawal severity classes emerged with higher severity more strongly associated with more severe smoking behaviors, more psychiatric comorbidity (major depression, alcohol dependence, conduct disorder, and anxiety disorders), and increased neuroticism. Xian et al.(2005), using latent class analysis, found a similar pattern of nicotine withdrawal severity in middle-aged (age 36–55) male twins from the Vietnam Era Twin Registry. Piasecki et al. (1998; 2000) conducted cluster analyses using nicotine withdrawal reports from adults in smoking cessation clinical trials and in unaided quitters and found that specific symptom profiles, particularly those characterized by negative affect, were associated with relapse to smoking across time. Less work has reported on the clustering of nicotine withdrawal symptoms in adolescent smokers, and to our knowledge, no single study has compared the pattern of nicotine withdrawal symptoms across population-based samples of adolescents and adults.
