English
This article is from Economics and Human Biology , volume 13 . Abstract •We model potential biases that may arise in Mendelian randomization analysis.•Genetic variants should robustly associate with exposures in independent samples.•If not, Mendelian randomization can suggest causality despite no true associations.
Sign in with your free account to read, download EPUBs or listen.
Ebook · Mendelian randomization in health research: Using appropriate genetic variants and avoiding biased estimates☆. · Taylor, Amy E., Davies, Neil M., Ware, Jennifer J., VanderWeele, Tyler, Smith, George Davey, Munafo, Marcus R. · Internet Archive