To account for multiple SNPs tested, Bonferroni corrections were performed on association results. The most significant DLG4 SNP (rs507506) remained statistically significant after correction (P=0.0016). Following correction in the Caucasian population, two DLG4 SNPs were significant (rs2521985, P=0.016; rs507506, P=0.03), while a trend for an association was observed for a DRD1 SNP (rs265975, P=0.09). No significant associations remained after correction in the African American population.