In 1986, Katan1 described a novel method to reliably estimate the effects of a causal variable without the need to conduct a traditional controlled trial.2 Now known as mendelian randomization (MR), this approach relies on the random assortment of genes (and noncoding DNA) as described by Mendel’s second law of inheritance. At the kernel of MR is the notion that this shuffling of DNA evenly distributes confounding factors, and this pattern, which remains unaltered through the lifetime of an individual, can be used to simulate the effect of modifiable factors (exposures) on susceptibility to a disease (outcome).3