Between-group differences were tested using the Mann Whitney U test. The General Linear Model (GLM) test with type III sum of square was used to assess the interaction effect between ELS and drinking, partial eta-squared value were used as effect size index. Group-wise bivariate correlation tests were performed using the Spearman’s rank test, and Bonferroni correction for multiple testing was applied. The statistical power of the human study was computed using the Genetic Power Calculator [33]. The size of the population-based sample had >80 % of statistical power considering α = 0.05 and heritability = 1 %. The interaction effect of FKBP5 rs1360780 genotype and parent-child relationship on AUDIT was estimated by univariate ANOVA GLM test with type III sum of square test, which allows adjusting for the main effect of the two factors of interest and is considered robust to violations of normality [34]. As confirmatory test, binomial regression was performed, adjusted for the main effect of genotype and environment. For illustrative purposes, fit lines were computed using standard linear regression with the least squares method. All analyses were performed using the Statistical Package for the Social Sciences software v. 21.
