Nested within this epidemiological sample is an intensive assessment of a subsample of 1,035 families (n= 2,070 individual twins), largely selected at random (72.3%, 748 families). The remainder of the subsample (27.7%, 287 families) was enriched with families where the participants have a family history of alcoholism based on baseline data. Details about the sub-sample have been described previously (Rose et al. 2001). We drew upon this subsample for the present analyses. In this subsample, both twins at age 14 and their biological parents were interviewed using the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA;(Bucholz et al. 1994)). The final sample included 1,852 interviewed twin individuals (51% male) and both of their parents when data was available (89%). Of these participants, 1,347 had follow-up SSAGA interviews completed at an average age of 22 years. The analytic study design and corresponding sample sizes are detailed in supplemental figure 1. Briefly, this subsample consisted of 34.7% monozygotic (MZ) twin pairs (50.5% female) and 65.3% dizygotic (DZ) twin pairs (47.9% female).
