Simultaneously accounting for winner's curse and sample structure in Mendelian randomization: bivariate rerandomized inverse variance weighted estimator

TL;DR

This paper introduces the BRIVW estimator, which improves Mendelian randomization analysis by simultaneously correcting for winner's curse and sample structure, leading to more accurate causal effect estimates.

Contribution

The paper develops the bivariate RIVW (BRIVW) estimator that models joint SNP associations and adjusts for sample structure, addressing limitations of previous methods.

Findings

BRIVW provides unbiased causal estimates in presence of sample structure.

Simulation studies show BRIVW outperforms existing methods.

Real data analysis confirms improved accuracy of BRIVW.

Abstract

The recently developed rerandomized inverse variance weighted (RIVW) estimator provides a simple and efficient framework to break the winner's curse in two-sample Mendelian randomization (MR). However, this method has ignored the possible presence of sample structure (e.g., residual population stratification and sample overlap), a common confounding factor in MR studies. Sample structure can not only distort SNP-exposure and SNP-outcome association estimates but also induce correlation between them, leading exposure-side instrument selection to propagate bias to the outcome side. To address this challenge, we propose the bivariate RIVW (BRIVW) estimator that can simultaneously account for the winner's curse and sample structure. The BRIVW estimator extends the RIVW framework by modeling the joint distribution of SNP-exposure and SNP-outcome associations, first adjusting their covariance matrix via linkage disequilibrium score regression to account for sample structure, and then applying randomized instrument selection and Rao-Blackwellization to obtain unbiased post-selection association estimates as well as their covariance matrix. Under appropriate conditions, we show that the BRIVW estimator is consistent and asymptotically normal. Extensive simulations and real data analyses demonstrate that the BRIVW estimator provides more accurate causal effect estimates than existing methods.