Glossary · Methodology
Mendelian randomization
In short
Mendelian randomization uses randomly inherited gene variants as a natural experiment to test whether a risk factor really is a cause. Someone born with, say, slightly higher LDL has in effect been assigned to a group by lot.
The idea follows from Mendel’s laws: which gene variant a person inherits is decided at conception and does not depend on income, diet or smoking. If a variant that changes only the risk factor is also associated with the disease, that points to a causal relationship.
The method has assumptions that cannot be fully tested — above all, that the gene variant acts only through this one risk factor. It also reflects a lifelong, usually small shift, not the effect of a product over a few years.
Genes draw the lots before lifestyle has a say.
Where this comes up in the fact checks
- ApoB test – In the Mendelian randomization study with 654,783 people, ApoB explained heart risk better than LDL cholesterol or triglycerides.
- Cutting down on alcohol – In the China Kadoorie Biobank with 512,715 adults, the U-curve for stroke risk disappeared with this method.
Sources
Related terms: Observational study · Cohort study · LDL and ApoB
Information only, not medical advice. The BK-Score is a subjective assessment according to disclosed rules – not a scientific evaluation. Scoring rules