We recognize limitations in our study. Maximizing the power available for this analysis comes at the cost of accepting broader biobank phenotyping approaches, which may reduce specificity of findings for the core depression phenotype.11 Nonetheless, strong genetic correlations between the ICD derived MDD with the broader definitions provide confidence in internal consistency, and future studies could look to further refine phenotyping. Although all genetic correlations were significant, there was substantial variance (95% CI= 0.72–1.7) in correlations with the FinnGen sample, probably due to power and heterogeneity in the broad phenotype we used from this sample. Finally, other ancestries remain understudied in relation to Europeans. We hope that the initial results reported here for the MVP African ancestry sample can help advance the field by encouraging additional concerted research in African and other non-European ancestral groups.
