Abstract
Background Mendelian randomisation uses genetic variants randomly allocated at conception as instrumental variables for a modifiable exposure of interest. Recent methodological advances allow for mediation analysis to be carried out using Mendelian randomisation. When genetic instruments are available for both an exposure and mediator, both multivariable and two-step Mendelian randomisation may be applied.
Methods We use simulations and an applied example to demonstrate when multivariable Mendelian randomisation and two-step Mendelian randomisation methods are valid and how they relate to traditional phenotypic regression-based approaches to mediation. We demonstrate how Mendelian randomisation methods can relax assumptions required for causal inference in phenotypic mediation, as well as which Mendelian randomisation specific assumptions are required. We illustrate our methods in data from UK Biobank, estimating the role of body mass index mediating the association between education and cardiovascular outcomes.
Results Both multivariable Mendelian randomization and two-step Mendelian randomization are unbiased when estimating the total effect, direct effect, indirect effect and proportion mediated when both confounding, and measurement error are present. Where both the exposure and mediator are continuous, in the presence of a rare or common binary outcome, we found little evidence of bias from non-collapsibility of the odds ratio.
Conclusion Phenotypic mediation methods require strong, often untestable, assumptions. Mendelian randomisation provides an opportunity for improving causal inference in mediation analysis. Although Mendelian randomisation specific assumptions apply, such as no weak instrument bias and no pleiotropic pathways, strong assumptions of no confounding and no measurement error can be relaxed.