Fig. 1: Study workflow.

Note: 1, “relevance” assumption, the IVs are robustly associated with the exposure; 2, “exclusion restriction” assumption, the IVs are not related to confounding factors; 3, “exclusion restriction” assumption, the IVs do not affect the outcome via a pathway other than exposure. LDSC Linkage Disequilibrium Score Regression, GWAS genome-wide association study, PLACO pleiotropic analysis under the composite null hypothesis, MAGMA multi-marker analysis of GenoMic annotation.