Avoiding dynastic, assortative mating, and population stratification biases in Mendelian randomization through within-family analyses

Estimates from Mendelian randomization studies of unrelated individuals can be biased due to uncontrolled confounding from familial effects. Here we describe methods for within-family Mendelian randomization analyses and use simulation studies to show that family-based analyses can reduce such biase... View More