Though very encouraging, there have also been some inconsistencies. While some studies have suggested that excluding individuals with comorbid major depression and/or cocaine and opioid dependence improves the association (Covault et al., 2004), others have found the association to be strongest in those with alcohol or illicit drug dependence (Agrawal et al., 2006). In addition, there has been a lack of consistency for association on a SNP by SNP level (Covault et al., 2004; Lappalainen et al., 2005) and in the directionality of association by haplotype (Covault et al., 2004; Soyka et al., 2008).
