MR analyses in which genetic instruments for more than one exposure are used. Multivariable MR can be used to estimate mediating effects in two-step MR, to adjust for possible pleiotropy (e.g., to isolate specific lipid effects) bias due to horizontal pleiotropy of a specific effect, or to adjust for potential confounding.
All of the genetic instruments should fulfil the IV assumptions. In addition, each instrument should be associated with more than one of the exposures included in the analyses; each instrument does not need to be associated with all exposures but each one does need to be associated with more than one exposure. When using multivariable MR to adjust for confounding or horizontal pleiotropy, it is important to make sure that an exposure on the causal path between the primary exposure of interest is not being adjusted for.
- Sanderson E, Davey Smith G, Windmeijer F, Bowden J. An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings. International Journal of Epidemiology 2018.
- Burgess S, Thompson SG. Multivariable Mendelian randomization: the use of pleiotropic genetic variants to estimate causal effects. American journal of epidemiology 2015;181:251-60.
- Xu L, Borges MC, Hemani G, Lawlor DA. The role of glycaemic and lipid risk factors in mediating the effect of BMI on coronary heart disease: a two-step, two-sample Mendelian randomisation study. Diabetologia 2017;60:2210-2220.
Other terms in 'Definition of MR and study designs':
- Bidirectional MR
- Binary exposure MR
- Factorial MR
- Instrumental variable (IV)
- Mendelian randomization (MR)
- One-sample MR
- Two-sample MR
- Two-sample MR with individual participant data (IPD)
- Two-sample MR with summary (aggregate) data
- Two-step/Mediation MR