A number of causal relationships were not tested because of sample overlap, which can lead to biased effect size estimates, or where there were an insufficient number of instrumental genetic variables (nvariables < 30). A total of 33 causal effects were tested, of which 24 were for a causal effect of depression on another trait, and nine tests for a causal effect of another trait on depression. Directional horizontal pleiotropy, where a genetic variant has an effect on both traits but via differing biological pathways, can bias the estimates of causal effects between traits. Using the MR Egger intercept test, there was no evidence (P ≥ 0.05) of directional horizontal pleiotropy for any of the 33 causal effects examined.
