Remission was operationalized as absence of symptoms at the time of interview and was conditional on the presence of lifetime AUD as defined above. A narrower phenotype based on lifetime presence of three or more AUD symptoms was also created and analyses were repeated using this stricter definition. Remission was defined so that symptom recency was at least 1 year less than current age, consistent with previous studies requiring a minimum of 6 months remission (Knop et al. 2007; Gilder et al. 2008; Penick et al. 2010). Early remission was defined as ≤12 months with no symptoms (i.e. age at most recent symptom was 1 year less than current age), and sustained remission as >12 months with no symptoms (i.e. age at most recent symptom was at least 2 years less than current age). A three-level variable representing remission status (none, early, sustained) was created to test whether early and sustained remission were statistically distinct categories. This variable was regressed on two dummy variables representing co-twin status on early and sustained remission in a multinomial logistic regression (with no remission
