The choice of statistical modeling was based on distribution of the data, and the assumptions of the statistical test. For all analyses, DAT1 genotype groups were carriers of 9 or less VNTR alleles (10/9, 10/<9, 9/9, and <9/<9) which were grouped together as A9, and 10 VNTR allele homozygotes (10/10) were grouped together as A10. For OPRM1, heterozygote (A/G) and homozygote (G/G) G-allele carriers were grouped together as *G and A allele homozygotes (A/A) were grouped together as AA. We tabulated the number of participants in each combined genotype subgroup: DAT1 A9/OPRM1*G, DAT1 A10/OPRM1*G, DAT1 A9/OPRM1 A/A, DAT1 A10/OPRM1 A/A. Chi-square test results didn’t show dependence between the frequency of the polymorphisms of the two genes ( χ12 = 0.065, p=0.799). That is, presence of one genotype did not predict presence of any other genotype.
