The statistical analyses addressed two goals: 1) to assess the prevalence and severity of the different types of childhood trauma exposure in AD subjects compared to the non-alcohol-dependent CON subjects, and 2) to evaluate associations between childhood trauma exposure types and alcohol-related outcomes in both the AD and CON subjects. To assess prevalence, a dichotomous variable representing presence or absence of each trauma type was established for both AD and CON subjects using a cut-off score indicating at least a low level of severity. The cut-offs for each trauma type were as follows: emotional abuse ≥ 9, physical abuse ≥ 8, sexual abuse ≥ 6, emotional neglect ≥10, physical neglect ≥ 8 (Bernstein and Fink, 1998). Severity was assessed using the total scores obtained for each trauma type from the CTQ. Comparisons between AD and CON subjects were performed using chi-square analyses and one way analyses of variance (ANOVA).
