between gender and the absence or presence of parental alcoholism were the independent variables of interest, adjusting for previously identified potentially confounding factors. Models were fit separately for maternal and paternal history of alcoholism. Similar to other NESARC analyses (Grant et al, 2009), these began with grouping psychiatric disorders into two categories (Axis I and II, respectively). Only when a significant association was observed between these categories and family history did we pursue analysis of the sub-categories of Axis I (mood, anxiety, and substance use) and Axis II (clusters A. B and C) disorders. For sub-categories that showed significant associations, the relationships with specific constitutive disorders was examined. Because of the limited number of initial analyses (two each for paternal and maternal family history for men and women, respectively) that was each significant at p<0.001, we did not adjust our alpha level to account for the multiple subsequent comparisons. Data were analyzed using SUDAAN software (Research Triangle Institute, 2001) and the NESARC calculated weights.
