Fig. 1
From: Investigating causal effects of income on health using two-sample Mendelian randomisation
Illustration of the three instrumental variable assumptions: (1) relevance, (2) independence, and (3) exclusion
Fig. 1 | BMC Global and Public Health