Close
MRC LHA research highlighted in Diabetes Journal
17 February 2023
A new methodological review on selecting genetic instruments in Mendelian Randomisation studies of HbA1c and diabetes has featured as a highlight in the February 2023 issue of Diabetes.
Q&A with first author Dr. Victoria Garfield
What did you do?
We searched the diabetes Mendelian Randomisation (a method which uses genetic data to try to understand whether associations might be causal in nature) literature for studies that assessed causality between diabetes/HbA1c and any health outcome. We also used data from ~350k individuals from the UK Biobank study to calculate important determinants of how good a genetic instrument is in this context (called the F statistic and R-squared).
Why did you do it?
Selection of genetic instruments in Mendelian Randomisation studies of diabetes/HbA1c is often suboptimal and poorly reported. Thus, we aimed to provide an expert methodological review, as MR is now a popular and valuable method when used carefully and genetic instrument selection is a crucial part of the early study design process.
What did you find?
We found that many more studies choose to genetically instrument diabetes (to understand whether diabetes causes other diseases), as opposed to HbA1c. However, the HbA1c instrument is potentially more powerful and robust in this context, as it performed better in the UK Biobank example we showed.
What is next?
We hope that this guide will aid researchers who wish to conduct an MR study of diabetes/HbA1c and that when it comes to genetic instrument selection, more care is taken.
The work and highlight is published in Diabetes.