CYP2A6 haplotypes predictive of slower metabolism were associated with an increased risk of nicotine dependence in both the primary COGA and replication COGEND samples of young adult daily smokers (Table 2, Figure 2). In multivariate models adjusting for age, sex, and study site, the continuous CYP2A6 metabolism metric had a significant effect in COGA (p =0.03) and COGEND (p=0.02), where a slow predicted metabolism was associated with an increased risk of nicotine dependence defined by an FTND score ≥4 (Table 2). Secondary analyses showed that slow metabolizers (defined by a metric of ≤0.85) had a 53% increased odds (OR=1.53, 95% CI 1.11–2.11, p=0.009) of developing nicotine dependence as compared to normal metabolizers (metric>0.85) in meta-analyses of COGA and COGEND studies (Table S2). Figure 2 illustrates this association by showing that a larger proportion of slow metabolizers in both COGA and COGEND developed nicotine dependence as compared to normal metabolizers.
