For the second goal, we analyzed the data in two steps. First, we examined simple bivariate correlations between the trauma measures from the CTQ (total score as well as individual subtype scores), potential mediator variables, and the alcohol-related outcomes of alcohol dependence severity (ADS) score (for the AD subjects) and AUDIT score (for the CON subjects). Variables that exhibited significant bivariate associations with the ADS score or the AUDIT score were then carried forward into a multiple mediation analysis to more fully examine relationships between predictors and outcomes. Based on previous theoretical and empirical research, we used the following variables as potential mediators between childhood trauma exposure and alcohol-related outcomes: lifetime drinking history variables such as age of first alcohol use and history of heavy alcohol use, anxiety and depression symptoms during withdrawal (AD subjects only), and personality factors.
