In part, some of the variability in findings across studies can also be attributed to differences in how drug dependence phenotypes are scored for analysis. Prior studies have primarily utilized clinically-defined phenotypes (based on the Diagnostic and Statistical Manual of Mental Disorders dependence symptoms (version four; DSM-IV) (15, 16)), such as dependence diagnosis (i.e., 3+ DSM-IV dependence symptoms all occurring in a 12 month period). However, alternative and dimensional summary scores, such as problem usage (i.e., 1+ DSM-IV dependence symptoms) and symptom counts have also been utilized to help overcome low-level levels of diagnosis often observed in community and population-based samples. Epidemiological studies (17, 18) show that individuals who meet the clinical criteria for dependence diagnosis for one substance are at greatly increased risk of using or becoming dependent/addicted to other substances, suggesting a generalized pattern of problematic drug usage. Furthermore, studies suggest that the “common liability” (i.e., each substance has its own set of genetic and environmental liabilities that are shared with other substances) and “alternative forms” (i.e., comorbidity across substances arises because each substance is an alternate manifestation
