rs1076560 has been associated previously with cocaine, alcohol and in the present study, opioid dependence (Moyer et al., 2011, Sasabe et al., 2007). There is considerable overlap between these addictions and it is rare to find individuals solely dependent on one substance. The prevalence of alcohol use disorders was estimated to be 89% amongst cocaine dependent individuals and 74% in opioid dependent individuals (Stinson et al., 2005). Various studies of polydrug use have reported that 50% of intravenous cocaine users also use heroin on a regular basis (Lauzon et al., 1994) and that 92% of heroin users were also taking cocaine (Hasin et al., 1988). It is possible that the original finding by Moyer et al reflects an indirect association with opioid or alcohol dependence, explaining our failure to detect an association with cocaine dependence in the present study. In our sample, a proportion of heroin addicts report the abuse of cocaine. This makes our positive association with opioid addiction and the lack of association with cocaine addiction difficult to interpret. Considering that the OR in cocaine addiction was similar
