Descriptive univariate analyses of all the variables were performed and means and standard deviations were calculated. We computed the ratio of serum 3HC to cotinine (the nicotine metabolite ratio) using the cotinine and 3HC measurements from the 24-hour serum samples. There are no published data on adolescent smokers to determine cut off values for faster versus slower nicotine metabolism. Therefore, we chose to consider participants in the top quartile (e.g., participants with a 3HC/Cot ratio of 0.5) as the faster metabolizers, such that participants with ratios <0.5 were considered slower metabolizers and participants with ratios ≥ 0.5 were considered to be faster metabolizers. Using Wilcoxon 2-tailed analyses we compared faster and slower metabolizers with regards to number of cigarettes smoked per day, withdrawal symptoms and addiction. We then used separate bivariate models to compare withdrwal symptoms between each group while adjusting for time since last cigarette smoked, duration of smoking (in months), and number of cigarettes smoked per day. Correlations between baseline cotinine, 3-HC, CO and number of cigarettes smoked per day were analyzed using Spearman’s rho. Finally, comparisons between faster and slower metabolizers by gender and race/ethnicity were made using the Fishers exact test.
