🤖 AI Summary
Classical Mendelian randomization (MR) using summary-level data is susceptible to the “winner’s curse” and horizontal pleiotropy, leading to biased causal effect estimates. To address these limitations, we propose the first unified robust inference framework that simultaneously corrects for the winner’s curse and identifies invalid instruments—without assuming a parametric distribution for pleiotropic effects or requiring perfect instrument selection. Our method integrates adaptive instrument selection, an improved inverse-variance weighted estimator, asymptotic normality theory, and Monte Carlo validation. We establish theoretical consistency and asymptotic normality of the estimator, with analytically tractable variance estimation. Extensive simulations and real-data analyses—including LDL cholesterol and coronary heart disease—demonstrate substantial improvements over existing MR methods in statistical efficiency, robustness to pleiotropy, and control of false positives.
📝 Abstract
In the past decade, the increased availability of genome-wide association studies summary data has popularized Mendelian Randomization (MR) for conducting causal inference. MR analyses, incorporating genetic variants as instrumental variables, are known for their robustness against reverse causation bias and unmeasured confounders. Nevertheless, classical MR analyses utilizing summary data may still produce biased causal effect estimates due to the winner's curse and pleiotropic issues. To address these two issues and establish valid causal conclusions, we propose a unified robust Mendelian Randomization framework with summary data, which systematically removes the winner's curse and screens out invalid genetic instruments with pleiotropic effects. Different from existing robust MR literature, our framework delivers valid statistical inference on the causal effect neither requiring the genetic pleiotropy effects to follow any parametric distribution nor relying on perfect instrument screening property. Under appropriate conditions, we show that our proposed estimator converges to a normal distribution and its variance can be well estimated. We demonstrate the performance of our proposed estimator through Monte Carlo simulations and two case studies. The codes implementing the procedures are available at https://github.com/ChongWuLab/CARE/.
