There are limitations to this study. For example, we used the number of cigarettes smoked per day as an indirect measure of nicotine dependence without consideration of which cigarette brand each smoker smoked. It is known that there exists a significant variation in nicotine concentration present in each cigarette brand. Therefore, the phenotype of smoking rate used here may represent only a very rough measure of nicotine dependence. Given the objective of this study and the limitation of the data set used in the analysis, we did not, nor were we able to, distinguish individuals in the non-smoking group who had been passively exposed to smoking (i.e., through second-hand smoke) from those who were never exposed to nicotine. As documented earlier, the transformed smoking phenotype still deviated slightly from a normal distribution; however, we do not feel that such remaining kurtosis would have a large effect on the linkage results reported herein, because only model-free methods were used in the analysis and they tend to be more robust to the presence of non-normality in the data. Also, the participants in
