The genetic correlations between depression and other traits/disorders reported in the previous section may arise from genes with pleiotropic effects and biological influences across both traits. Alternatively, there may be a causal effect of depression on other traits (e.g. depression influencing triglyceride level) or from other traits causally influencing depression (e.g. triglyceride level leading to depression). To determine whether causal relationships exist between depression and the 41 genetically correlated traits in Supplementary Table 3, we used a bi-directional, two-sample Mendelian randomization (MR) approach using MR-Base v0.4.9 11 in R and an inverse-variance weighted (IVW) regression analysis. Where there was also evidence of variant heterogeneity additional sensitivity tests were conducted using an MR Egger test and a weighted median test (Supplementary Table 4).
