Mendelian randomization
DEMendelsche Randomisierung
Mendelian randomization (MR) is a clever way to test cause without a trial. It uses inherited genetic variants (usually SNPs tied to an exposure in a GWAS) as natural stand-ins for that exposure. The trick: your genes are dealt out essentially at random at conception, like a built-in randomized experiment. MR rests on three assumptions. The genetic instrument must be strongly tied to the exposure (relevance). It must affect the outcome only through that exposure (exclusion). And it must be independent of confounders (independence). The big pitfalls are violations: pleiotropy, population stratification, or weak instruments. Sensitivity checks can help. Tools like MR-Egger, weighted-median, and CAUSE spot and partly correct for horizontal pleiotropy. In longevity research, MR is widely used. It tests whether biomarkers like LDL-C, CRP, IGF-1, or BMI causally affect lifespan. And it does so without waiting decades for a randomized trial.
Last reviewed:
This definition is educational and is not medical advice, a diagnosis, or treatment. Talk to a doctor about any health decisions. Read our full medical disclaimer