To date, four previous case-control GWAS analyses of AUDs have not provided evidence of association that reached genome-wide significance with AD or quantitative traits 9–12. In this study our approach to identifying genetic risk factors for alcohol problems focused primarily on a quantitative measure of alcohol symptom count, rather than an AUD diagnosis. Our SC measure was based on 7 DSM-IV AD criteria and deliberately excluded the 4 criteria associated with DSM-IV alcohol abuse. We crafted this measure to allow the most straightforward comparisons between findings for SC and findings for diagnosis of dependence. Dimensional dependence measures such as SC are more powerful than dichotomized phenotypic measures (i.e. DSM-IV AD) for detecting risk factors, especially in samples containing adolescent subjects. In the present study, we compared chromosomal regions that showed strong associations with both SC and DSM-IV AD and consistently observed a stronger relationship for genetic variants with SC than with DSM-IV AD (Table 1, Figure 4); the effect sizes and directions for both phenotypes were relatively consistent.
