We did an exploratory analysis in which individuals who smoked CPD<25 (N=130) or CPD≥25 (N=120), the defined “breakpoint” (whereby specificity improves from 65% (CPD=20) to 90% (CPD≥25)) for DSM-IV nicotine dependence in Europeans (Berrettini et al., 2008), were categorized as nicotine dependent ‘controls’ or ‘cases’, respectively. Even with very limited power, a nominal trend toward statistical significance was observed for two markers: rs3841324:GGGCGGGGCCAGAGGGAAATAG>- (X2=2.782, p=0.0953) and rs569207:C>T (X2=3.713, p=0.054). One haplotype (f=0.177) showed a nominal trend toward statistical significance (X2=2.954, p=0.0856) (Figure 2). No other single marker or haplotype showed a trend (p<0.1) toward statistically significant association with the CPD phenotype.
