LD score regression (Bulik-Sullivan et al. 2015b) showed a genetic overlap between AUDIT and numerous traits (Fig. 3 and Supplementary Table 9). Both alcohol consumption and AUD showed the steepest correlations with AUDIT score (rg = 0.68 for both), while the better powered alcohol consumption traits (N = 70,460; N = 112,117) yielded a significant result (P = 3.40 × 10−3; P = 9.01 × 10−10) the P value for the genetic correlation with AUD (N = 7,280) fell just short of nominal significance (P = 6.41 × 10−2). We detected a significant negative genetic correlation between AUDIT and ADHD (rg = −0.29; P = 1.43 × 10−3). We observed a positive genetic correlation between AUDIT and lifetime tobacco use (rg = 0.42; P = 1.52 × 10−3). Unexpectedly, we identified a positive correlation between AUDIT and years of education (rg = 0.27; P = 3.14 × 10−5), college attainment (rg = 0.26; P = 8.11 × 10−3) and childhood IQ (rg = 0.42; P = 6.26 × 10−3). Also surprisingly, AUDIT was negatively genetically correlated with BMI (rg =
