Introduction

In this chapter, we illustrate the conceptual framework and motivation for Mendelian randomization, explaining how Mendelian randomization offers opportunities to address some of the problems of conventional epidemiology. We describe the specific characteristics of genetic data which give rise to the Mendelian randomization approach.

Conclusions

Mendelian randomization has the potential to be a useful tool in a range of scientific contexts to investigate claims of causal relationships. It must be applied with care, as its causal claims come at the price of assumptions which are not empirically testable. Its methods must be refined, as often data on multiple genetic variants taken from several study populations are required to achieve meaningful findings. But, when properly used, it gives an insight into the underlying causal relationships between variables which few other approaches can rival.