Fig. 1

Schematic representation of the three assumptions of Mendelian randomization and current study design. Assumption 1: the IVs should be closely associated with exposures; Assumption 2: the IVs selected are not associated with potential confounders; Assumption 3: the IVs should affect results dependently through exposure, but not the direct correlation. This bi-directional MR analysis was performed in two steps: sarcopenia was studied as exposure while osteoarthritis was studied as outcome in the first step, whereas the second step was reversed. IVs instrumental variable, SP sarcopenia, OA osteoarthritis