match those we controlled for in BioVU, (i.e., waist girth instead of BMI, hypertension medication use and systolic blood pressure instead of hypertension diagnosis), we used them preferentially since they are better measures of the underlying risk factors, but were rarely reported (waist girth) or error-prone (systolic blood pressure) in the EHR. In the fully adjusted models for both BioVU and ARIC, we tested for interaction between sex and the polygenic score for MDD or loneliness and report interaction P < 0.05. In addition, we provide the results of sex-stratified analyses, as effect estimates may differ even in the absence of a statistically significant interaction.
