Mendelian randomization: prospects, potentials, and limitations

Abstract

In this issue of the International Journal of Epidemiology we reprint a letter to the Lancet by Martijn Katan ¹ and several commentaries ^{2 –7} concerning what has become known as ‘Mendelian randomization’ ^{8 –12}—the use of genotype–disease associations to make inferences about environmentally modifiable causes of disease. Here we will reflect on the prospects, potentials, and limitations of Mendelian randomization.