Although these findings provide proof of principle for the approach of carrying forward genes identified in GWAS of other externalizing disorders to test for association with conduct disorder, it should be noted that other attempts to replicate the association between GABRA2 and conduct disorder have produced inconsistent or null results. In a case-control sample of adolescents, GABRA2 SNP rs279871 was nominally associated with conduct disorder (n = 428, p = 0.02); however, the effect was in the opposite direction reported by Dick et al. (2006) (Sakai et al., 2010). Furthermore, a broader family-based association test that included siblings and parents of adolescent patients and controls (n = 1582) found no evidence for association between rs279871 and conduct disorder (p = 0.48) (Sakai et al., 2010). Finally, a series of six GABRA2 SNPs genotyped in COGA adolescent (n = 933) and young adult (n = 1191) offspring were not associated with conduct disorder clinical criteria, but several SNPs were associated with subclinical Achenbach externalizing behavior (Dick et al., 2013).
