Association analyses were performed using two models. The first model used the imputed minor allele dosage as the dependent variable and the DSM-IV symptom count for AD and the three other major SD diagnoses (cocaine, opioid and nicotine dependence: CD, OD and ND, respectively) as ordinal predictors (adjusted for sex, age and the first three ancestry PCs). The mutual adjustment of AD for measures of dependence on other substances facilitated the identification of SNPs unique to AD and limited confounding because of comorbid dependencies. All individuals contributed to this analysis, including those meeting no DSM-IV criteria for AD and those meeting all criteria. Subjects who reported having less than one full drink ever, including those who endorsed having a sip of alcohol at a religious ceremony, were excluded. This ordinal trait model has greater power to detect genetic associations than a univariate model based on disease status because of greater information content and improved specificity of the dependence measure. Ordinal trait data were derived for all samples included in the study except the German sample.
