There were no significant interactions between parental divorce/separation and perceived stress (HR=0.88, 95% CI: 0.71, 1.09), nor between parental divorce/separation and family history of drinking problems (HR=0.81, 95% CI: 0.47, 1.42). However, there was a significant interaction between divorce/separation and parent drinking (HR=1.10, 95% CI: 1.01, 1.19). Findings indicate that the hazard ratio increased as parent drinking scores increased, with non-significant values at AUDIT values less than 2.5 but significant and increasingly large values with higher AUDIT scores; Figure 3 presents the hazard ratios as a function of different AUDIT values).
