Finally, polychoric twin-pair correlations for the three level latent class nicotine withdrawal solution among adults were the following: MZ: r = 0.52 (95% Confidence Interval (CI): 0.43–0.61), DZ: r = 0.13 (95%CI: −0.003–0.27); tetrachoric twin-pair correlations for the two class solution among adolescents were: MZ: r = 0.48 (95% Confidence Interval (CI): 0.18–0.71), DZ: r = 0.15 (95%CI: −0.26–0.52). In each case, the MZ correlation was more than twice that of the DZ twin-pair correlation, suggest important additive (A) and non-additive (D) genetic influences, with minimal influences of shared family environment. The results of structural equation modeling for the latent class measure of nicotine withdrawal, which tested ADE models against submodels, found the most parsimonious model to be an AE model, in which estimates could be constrained across the cohorts and gender. In this best fitting genetic model of nicotine withdrawal, in both the adolescent and adult smokers, A was estimated at: 49% (95% CI: 40%–57%) and E at 51% (95% CI: 43%–60%).
