III). But for rs2869546, using the number of cigarettes smoked produced a slightly stronger association (P = 0.0060, effect = 0.092, 95% CI, 0.026-0.159). As pointed out in the previous studies [Saccone et al., [2007]; Berrettini et al., [2008]; Thorgeirsson et al., [2008]], rs16969968 and rs1051730 were in high LD (r2 = 0.96 in our combined sample), our results were consistent with this observation. From our study, it seems that a large sample is required to detect the small effect at this locus. For rs16969968, carrying one copy of the minor allele, the A allele (Asn), only increased the mean by 0.13-0.17 standard deviation. Of note, none of the markers typed in this study is located in the CHRNB4 gene. Since the markers used in our study were selected to tag the variants in the CHRNA5/CHRNA3/CHRNB4 genes, where extensive high LD was evident in both our samples and others reported [Saccone et al., [2007]; Berrettini et al., [2008]; Bierut et al., [2008]; Grucza et al., [2008]; Schlaepfer et al., [2008]; Wang et al., [2008]; Weiss et al., [2008]], we believe that the association signal observed at rs16969968 and rs1051730 could extend to CHRNB4 gene. In other words, while we did
