Case-control association analyses for AD were conducted using PLINK (Purcell, http://pngu.mgh.harvard.edu/∼purcell/plink/) at both single marker and haplotype levels for GAD1 and GAD2 genes. For quantitative alcohol-related traits measured within the cases, we used regression based models implemented in PLINK to calculate effect size and significance level for each marker. To examine the gender specific effects in the association of GAD genes with AD and alcohol-related traits, the aforementioned case-control binary and case-only quantitative association analyses were conducted in the overall sample and in the female and male samples separately. Among 1105 total samples, there were 668 males (370 cases), 433 females (204 cases), and 4 samples with unspecified sex1 which were excluded from gender specific analysis.
