To ensure the direction of association between G j and X 1 was the same for all j variants, the absolute value of the genetic associations with X 1( |βX1j|) were used to generate βYj (Equation (14)). It was assumed that βXkj(for all k) and βYj had the same reference allele and the genetic variants were uncorrelated. The multivariable IVW, univariable MR‐Egger, and multivariable MR‐Egger methods were applied to the simulated datasets. The weights for the multivariable IVW and multivariable MR‐Egger are given by Equation 15, while Equation 16 contains the weights for univariable MR‐Egger: (15)se(βYj)−2=(ϵj2+σα′2)−1, (16)se(βYj)−2=(ϵj2+σα′2+θ22σ22+θ32σ32)−1.
