Summary
This methodological paper addresses a critical problem in Mendelian randomization: the conventional inverse-variance weighted method produces biased estimates when some genetic instrumental variables are invalid. The authors present a novel weighted median estimator that remains consistent even when up to 50% of information derives from invalid instruments, offering improved Type 1 error rates in finite samples. The method is demonstrated through simulation and application to lipid–coronary artery disease risk, where it reveals null effects of HDL cholesterol consistent with experimental evidence, and is proposed as a complementary sensitivity analysis alongside MR-Egger regression.
UK applicability
This is a statistical methodology paper with no direct application to UK farming systems or soil health. However, the methods may be relevant to UK researchers conducting Mendelian randomization analyses of dietary or nutritional exposures and health outcomes, improving the robustness of causal inference in nutritional epidemiology.
Key measures
Type 1 error rates, consistency of causal estimates, performance compared to inverse-variance weighted and MR-Egger regression methods
Outcomes reported
The study developed and validated a weighted median estimator for Mendelian randomization that produces consistent causal estimates even when up to 50% of genetic variants are invalid instrumental variables. The method was evaluated through simulation analysis and applied to lipid–coronary artery disease associations.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.