Adjusting for medication status in genome-wide association studies
5 affiliations
- 1MRC Integrative Epidemiology Unit, Population Health Sciences, Bristol Medical School, University of Bristol, Bristol, United Kingdom
- 2MRC/UVRI and LSHTM Uganda Research Unit, Entebbe, Uganda
- 3The Department of Non-communicable Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, United Kingdom
- 4Medical Research Council Population Health Research Unit, Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, Oxford, United Kingdom
- 5NIHR Bristol Biomedical Research Centre, University Hospitals Bristol and Weston NHS Foundation Trust and University of Bristol, Bristol, United Kingdom
Abstract
When conducting genome-wide association studies, addressing medication status that is relevant to the trait of interest can prevent biases that could distort estimates of the true genetic effect. Here, we introduce the genetic empirical medication reduction adjustment (GEMRA) method which uses a heuristic search for an empirical adjustment to be applied to phenotypic values of participants reporting medication use. Through simulations we show that the direct genetic effect estimates in the GEMRA approach exhibited less bias and greater statistical power than common strategies such as including medication as a covariate, restricting the sample to unmedicated users or including all samples without adjustment. We then applied the GEMRA approach to estimate statin medication adjustment for analysis of low-density lipoprotein (LDL) cholesterol levels and antihypertensive medication adjustment for analyses of systolic (SBP) and diastolic (DBP) blood pressure using multi-population data from UK Biobank and the Uganda Genome Resource. We found that a relative rather than an absolute adjustment better modelled the effect of medication on LDL cholesterol, SBP and DBP, with an effect of 40%, 10% and 10% reduction appearing to be consistent across populations, respectively. We also show that spurious differences in medication effects across populations can arise when the effect scale is misspecified. Author summary: When studying how genetics influence health outcomes, it is important to account for medications that can affect the trait being studied. Otherwise, medication use may make genetic effects appear stronger or weaker than they truly are. Here, we developed a method, GEMRA, to account for the effects of medication in genetic studies. Using simulated data, we found that GEMRA was better able to detect true genetic associations than several commonly used approaches. We then used GEMRA to account for the effects of statins on LDL cholesterol and blood pressure medications on systolic and diastolic blood pressure. Using data from UK Biobank and the Ugandan Genome Resource, we found that medication effects were better represented as a percentage reduction rather than as the same fixed change for everyone. Our findings also showed that choosing the wrong way to represent medication effects (i.e., as a percentage reduction or fixed value) can create misleading differences between populations.
Related
- Yield and Yield Components in Soybean Using Genome-Wide Association Studies, GWAS: Current Status, Challenges, and Breeding Prospects
- Multi-ancestry genome- and transcriptome-wide association studies expand the susceptibility landscape for breast cancer overall and across subtypes
- Adjusting for Medication Use in GWAS and Its Impact on Mendelian Randomisation Analyses: An Example of Systolic Blood Pressure in UK Biobank
- A review of self-supervised imaging phenotypes for genome-wide association studies
- Bayesian adaptive experimental design for efficient microbial genome-wide association studies
BibTeX
@article{chong2026adjusting,
title = {{Adjusting for medication status in genome-wide association studies}},
author = {Amanda H. W. Chong and Christopher Kintu and Yoonsu Cho and Segun Fatumo and Jason Torres and George Davey Smith and Tom R. Gaunt and Gibran Hemani},
journal = {PLOS Genetics},
year = {2026},
month = nov,
doi = {10.1371/journal.pgen.1012366},
url = {https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1012366},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Adjusting for medication status in genome-wide association studies
AU - Amanda H. W. Chong
AU - Christopher Kintu
AU - Yoonsu Cho
AU - Segun Fatumo
AU - Jason Torres
AU - George Davey Smith
AU - Tom R. Gaunt
AU - Gibran Hemani
PY - 2026
DA - 2026/11/02
JO - PLOS Genetics
DO - 10.1371/journal.pgen.1012366
UR - https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1012366
AB - When conducting genome-wide association studies, addressing medication status that is relevant to the trait of interest can prevent biases that could distort estimates of the true genetic effect. Here, we introduce the genetic empirical medication reduction adjustment (GEMRA) method which uses a heuristic search for an empirical adjustment to be applied to phenotypic values of participants reporting medication use. Through simulations we show that the direct genetic effect estimates in the GEMRA approach exhibited less bias and greater statistical power than common strategies such as including medication as a covariate, restricting the sample to unmedicated users or including all samples without adjustment. We then applied the GEMRA approach to estimate statin medication adjustment for analysis of low-density lipoprotein (LDL) cholesterol levels and antihypertensive medication adjustment for analyses of systolic (SBP) and diastolic (DBP) blood pressure using multi-population data from UK Biobank and the Uganda Genome Resource. We found that a relative rather than an absolute adjustment better modelled the effect of medication on LDL cholesterol, SBP and DBP, with an effect of 40%, 10% and 10% reduction appearing to be consistent across populations, respectively. We also show that spurious differences in medication effects across populations can arise when the effect scale is misspecified. Author summary: When studying how genetics influence health outcomes, it is important to account for medications that can affect the trait being studied. Otherwise, medication use may make genetic effects appear stronger or weaker than they truly are. Here, we developed a method, GEMRA, to account for the effects of medication in genetic studies. Using simulated data, we found that GEMRA was better able to detect true genetic associations than several commonly used approaches. We then used GEMRA to account for the effects of statins on LDL cholesterol and blood pressure medications on systolic and diastolic blood pressure. Using data from UK Biobank and the Ugandan Genome Resource, we found that medication effects were better represented as a percentage reduction rather than as the same fixed change for everyone. Our findings also showed that choosing the wrong way to represent medication effects (i.e., as a percentage reduction or fixed value) can create misleading differences between populations.
SN - 1553-7404
ER -