Two-sample MR in which summary (aggregate) results are used for the genetic IV-risk factor and genetic IV-outcome associations. These are usually summary statistics from GWASs that are publicly available.
Assumptions and key sources of bias are as for two-sample MR. With aggregate data, it may be difficult to ensure the two samples are from the same underlying population. It is difficult to explore whether the genetic IVs are related to confounders of the risk factor-outcome association. It is not possible to detect non-linear effects unless there are aggregate data results for non-linear effects. Where the two samples overlap, the greater the overlap, the more the study properties become like one-sample MR.


References
- Hemani G, Zheng J, Elsworth B et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife 2018;7.
- Lawlor DA. Two-sample Mendelian randomization: opportunities and challenges. . International Journal of Epidemiology 2016;doi:10.1093/ije/dyw127.
Other terms in 'Definition of MR and study designs':
- Bidirectional MR
- Binary exposure MR
- Factorial MR
- Instrumental variable (IV)
- Mendelian randomization (MR)
- Multivariable MR
- One-sample MR
- Two-sample MR
- Two-sample MR with individual participant data (IPD)
- Two-step/Mediation MR