Ensuring Mendelian Randomization Rigor Amid Rapid Growth

Mendelian randomization (MR) has become a widely used approach in observational epidemiology, employing genetic variants as instrumental variables to infer causal relationships. However, as discussed in a recent editorial in eGastroenterology, the rapid increase in MR studies—from 3,545 in 2023 to over 6,600 in 2024—has led to concerns about declining study quality and interpretive validity. Many new MR publications rely on simplified two-sample designs and summary-level data without sufficient attention to critical MR assumptions, raising the risk of misleading conclusions and diminishing the field’s credibility.

MR relies on three core assumptions: relevance (the genetic instrument must correlate with the exposure), independence (no confounding between the instrument and outcome), and exclusion restriction (the instrument affects the outcome only through the exposure). The authors from Charite - Universitatsmedizin Berlin point out that independence and exclusion restriction are rarely testable and often violated, especially when using summary data or unvetted instruments. Common pitfalls include using exposures with weak or unclear genetic bases, inadequate instrument selection based solely on statistical significance, and overuse of two-sample MR without proper confounding or pleiotropy checks.

To improve MR research, the team recommends several best practices: ensuring exposures are genetically relevant, combining biological plausibility with statistical strength in instrument selection, using multiple statistical models to account for assumption violations, and validating findings with external evidence or complementary analyses. They also stress the importance of cautious interpretation, suggesting results should be described as “genetically predicted associations” rather than definitive causality. For readers, editors, and reviewers, the editorial advocates for strict adherence to methodological guidelines, careful scrutiny of instrument selection, and discouragement of redundant or poorly justified studies.

Ultimately, the editorial argues that MR’s value lies in methodological rigor and cross-validation, not in the sheer number of publications. By prioritizing quality over quantity, the field can maintain its scientific integrity and continue to offer meaningful insights into causal relationships in medicine.