Hello there ResearchGate community:

I have recently come across multiple Mendelian Randomization (MR) analyses which address the casual relationship between the same exposure and outcome.

Examining such a finding prompted me to look at the possibility of conducting a systematic review & meta-analysis of these MR analyses.

I personally have no experience conducting a systematic review where data is not based on real patients from primary research articles. MR analyses use genetic data from biobanks and genetic instruments such as SNPs to establish causal relationships. Thus, MR analyses could eliminate confounding factors - to a very great extent - and this helps ensure the reliability of your conclusion.

Meta-analyses of MR analyses are scarce and I didn't manage to find clear and appropriate guidelines on how to approach such a situation.

Would somebody please help narrate any similar experiences of theirs?

- How to ensure no overlap of genetic biobank data occurs in your meta-analysis?

- As MR analyses report ORs/RRs, will the sole purpose of the analysis here serve to pool these numbers?

- How could one assess the quality of such MR analyses?

- What is pleiotropy and how could one account for this while pooling the effect sizes from the MR analyses?

- Any tips or recommended readings would be greatly valued!

Thank you in advance.

Enjoy the rest of your day!