Dense, longitudinal clinical data from large genotyped samples provide new opportunities to disentangle shared genetic risk factors from phenotypic hitchhiking. We hypothesized that pleiotropic genetic risk factors in part explained the phenotypic associations and genetic correlations between MDD, loneliness, and CAD (Fig. 1c). We assessed this hypothesis by testing polygenic scores for MDD, loneliness, and CAD for association with clinical diagnoses of MDD and CAD, before and after adjusting for known disease risk factors and comorbidity patterns, in a large collection of patient electronic health records (EHR) linked to genome-wide genotyping. Polygenic scores aggregate the effects of many genetic variants into a single measure of genetic liability to a trait, and can be calculated for all individuals with genome-wide genetic data if their ancestry is similar to the ancestry of the participants in the external GWAS from which the individual genetic variant effect estimates are derived.
