Interim analyses of phenotypic data revealed high rates of licit and non-opioid illicit drug dependence in neighborhood controls (41.6%, 28.1%, and 36.0% for nicotine, alcohol, and illicit drug dependence respectively in the final sample), which raised concerns that comparisons with these individuals would be ill-suited to identify genetic variants associated with liability that is shared across classes of drug dependence. These concerns prompted a decision to revise the study design to include a non-dependent control group of unrelated individuals selected for the study’s primary genotypic analyses from a large Australian Twin Registry (ATR) pool of twins and family members [see (Hansell et al. 2009) for a description of this sample]. Inclusion criteria for this group included an adequate DNA supply available and existing IRB approval allowing additional genotyping for substance dependence and related phenotypes. Exclusion criteria were lifetime diagnoses of any DSM-IV illicit drug or alcohol dependence at prior interview. Individuals without a lifetime DSM-IV diagnosis of nicotine dependence were also preferentially selected. The resulting non-dependent control sample (N=1500) had a lifetime prevalence of nicotine dependence (12.5%) below that of the Australian general population.
