After carrying out post hoc tests separately in the two genders for all interaction effects and the main effect of 5-HTTLPR, all significant effects disappeared in women in the combined sample (Supplementary Table S5). In case of men (Supplementary Table S6), there was a significant interaction effect between 5-HTTLPR and total life events using additive (P=0.0014) and recessive models (P=0.0001), which could be replicated both in the Budapest (additive model: P=0.0151, recessive model: P=0.0001) and Manchester subsamples (additive model: P=0.0201, recessive model: P=0.0322). Furthermore, in men we also found a significant interaction effect of RLE-financial in the combined sample (additive model: P=0.0010, recessive model: P<0.0001), which could be replicated in the two independent subsamples of Manchester (additive: P=0.0418, recessive: P=0.0076) and Budapest (additive: P=0.0013, recessive: P<0.0001). The other results could not be replicated in our study.
