Attention to the misuse of Mendelian randomisation in medical research

作者：Lanlan Chen, Adrien Guillot, Carolin V. Schneider · 发表于：eGastroenterology · 年份：2025 · DOI：10.1136/egastro-2025-100187 · 被引用次数：12 · 研究领域：Genetic Associations and Epidemiology、Liver Disease Diagnosis and Treatment、Genomics and Rare Diseases

Mendelian randomisation (MR) is a widely used method that employs genetic variants as instrumental variables (IVs, also referred to as genetic instruments below and they represent eligible genetic variants in MR) from genome-wide association studies (GWAS) to explore the causative relationship between putative risk factors (ie, exposures) and outcomes.1 According to a Web of Science literature search conducted at the end of 2024, the number of related publications has grown exponentially, reaching 3545 in 2023 and 6607 in 2024 (figure 1). Almost simultaneously, we observed a peak of citations from 2019 to 2021, which reflected the popularity and acknowledgement of papers published in these 3 years. However, a decline in citations has been witnessed from 2022 and here are two possible explanations for it: (1) papers published in 2022 or later only have two or fewer years to be cited, and it can lead to fewer citations absolutely; (2) some papers published from 2022 are of low quality and may even be redundant analyses, thus, they are less likely to be cited. Maybe we can validate which one is correct after 3 years. Meanwhile, this surge in publications has raised concerns about the potential misuse of the method, often without sufficient critical evaluation, leading to questions about its credibility.2 While the increasing use of MR has contributed to answering critical questions in medical research, inappropriate application of the method—for instance, through simplified two-...