In addition, our analysis which included a sex*genotype interaction term revealed suggestive evidence for interaction for nearly all of the SNPs in CHRNB3 and CHRNA10. These data are presented in Supplementary Table 1, with columns for the SNP beta and p-values, and interaction term beta and p-values, in the full model. The initial association appears to be driven by the males, since an analysis of only males yielded similar (slightly stronger) evidence for association, while there was no evidence for an association in the female subgroup.
