Journal of Rheumatic Diseases

Download original image
Fig. 2. Three core assumptions for Mendelian randomization. (A) Instrumental variable (IV) assumption 1: The genetic variant should be associated with the exposure. (B) IV assumption 2: The genetic variant should not associate with confounder. (C) IV assumption 3: The genetic variant should influence the outcome only through the exposure.
J Rheum Dis 2020;27:241~246 https://doi.org/10.4078/jrd.2020.27.4.241
© J Rheum Dis