In the next steps, we tested the main and interactive effects of totSLE and drinking onset on PYDD. TotSLE was a significant predictor of PYDD in men (b = 0.08, SE = 0.03, R2 = 0.56%), but not in women (b = 0.01, SE = 0.02, R2 = 0.04%) (Models F2 and M2). In contrast, drinking onset was a significant predictor of PYDD in both sexes. Among women, ONS17 accounted for 4.4% of variance in PYDD (b = 0.28, SE = 0.04; Model F3a); and ONS14 accounted for an additional 1.3% (b = 0.25, SE = 0.07; Model F3b). In men, ONS17 accounted for nearly 3.0% of variance in PYDD (b = 0.42, SE = 0.05; Model M3a); and ONS14 accounted for another 1.0% (b = 0.25, SE = 0.06; Model M3b). The interaction terms of totSLE with drinking onset ages were not significant (women: SLE × ONS17: b = 0.05, SE = 0.05; SLE × ONS14: b = 0.07, SE = 0.08; men: SLE × ONS17: b = 0.001, SE = 0.07; SLE × ONS14: b = 0.08, SE
