Each participant’s genetic risk for alcohol use disorders (AUDs) was indexed by his/her parents’ and co-twin’s alcohol dependence criterion counts (at age 22). The contribution of each measure (parents’ AD criterion count, and co-twin’s AD criterion count) to the total AUD risk was based on a modified ridit score (Kendler et al. 2011). When data on both parents were available, the criterion count from the most severely affected parent (the father in 70% of families with data from both parents) was used in risk score calculation. In an effort to use the information gained from the twin study, scores from MZ co-twins were weighted twice as strongly as scores from DZ co-twins or parents. As this method equates the genetic relationship of parents and offspring and siblings, an additive model is assumed. AUD-GRs were computed on 1,312 participants assessed.
