Initial ascertainment of alcohol-dependent probands (designated Stage I) was performed by screening consecutive admissions at treatment facilities. Probands were assessed with the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA), a comprehensive diagnostic instrument developed for this study and now widely used [18,19]. Extensive histories of substance use and abuse were gathered along with diagnostic information for multiple Axis I disorders and antisocial personality disorder. To be recruited into the COGA study, probands had to meet both the diagnostic criteria for alcohol dependence (by DSM-III-R criteria [20] and the criteria for definite alcoholism specified by Feighner et al. [21]); thus, the COGA sample is representative of a severely alcohol-dependent population. All first degree relatives of the probands were invited to participate. Children and adolescents in the families were assessed with complementary age-appropriate instruments (C-SSAGA, child and adolescent versions). A set of control families was ascertained to provide normative measures; they were not screened to eliminate those with psychiatric disorders, and are similar to a general population sample. Written informed consent was obtained from all subjects, and the Institutional Review Boards
