By Pooja Toshniwal PahariaReviewed by Lexie CornerMay 15 2025A new study in Nature Communications explores how variance quantitative trait loci (vQTLs) and variance polygenic scores (vPGSs) can improve the genetic assessment of blood cell phenotypes. By using vQTLs to construct vPGSs, researchers stratified individuals by genetic variability in specific traits.
This approach allows for a more refined understanding of how genetic variance influences blood-related phenotypes and disease risk, potentially improving personalized risk prediction and treatment strategies.
The study also found that incorporating vPGS into traditional genetic models increased prediction accuracy. Notably, lifestyle factors like alcohol consumption were shown to elevate phenotypic variance, further influencing risk.
Image Credit: shisu_ka/Shutterstock.com
Background: Blood Cell Traits and Genetic Risk
The complete blood count is a routine diagnostic tool used to monitor physiological functions such as oxygen transport and immune response. Previous genome-wide association studies (GWAS) have linked blood cell phenotypes to complex conditions like cardiovascular and autoimmune diseases.
However, the role of trait variance, which influences phenotypic expression and overall fitness, has received less attention.
vQTLs have been studied in other traits such as body mass index and cardiometabolic markers, but their role in blood cell phenotypes remains underexplored. Polygenic risk scores (PGSs) estimate cumulative genetic contributions to disease, even when individual variants have small effects.
Variance PGSs, based on vQTLs, offer an added dimension by quantifying variability in phenotypic expression. Together, vQTLs and vPGSs could provide a more detailed picture of genetic influence on blood traits.
About the Study
In this study, researchers examined variance quantitative trait loci (vQTLs) and variance polygenic scores (vPGSs) to assess genetic variability in blood cell phenotypes.
The team mapped vQTLs for 29 blood cell traits using data from 408,111 European participants in the UK Biobank (UKB), aged 40 to 69. Levene’s test was used to identify genetic loci associated with trait variance. The INTERVAL study, which includes 50,000 healthy adult blood donors, served as the validation cohort. Functional Mapping and Annotation of Genetic Associations (FUMA) was used for vQTL annotation, pathway analysis, and variant selection.
vPGS values were calculated for each phenotype and used to stratify 40,466 individuals in the validation cohort based on genetic variability. For each trait, participants were grouped into the top 5 % (genetically more variable) and bottom 5% (genetically less variable) based on their vPGS.
Researchers then evaluated the relationship between vPGS and conventional polygenic scores (PGS), assessing how these scores interact across all traits. They also tested whether combining vPGS with traditional PGS improved prediction performance, including the use of multi-trait vPGSs.
To estimate genetic parameters, the study applied Linkage Disequilibrium Score Regression (LDSC). Causal relationships were assessed using Mendelian Randomization (MR), supported by MR-PRESSO and weighted median methods. Additional analyses included OmicS-data-based Complex Trait Analysis (OSCA), deviation regression modeling (DRM), and Horizontal Pleiotropy Score (HOPS) to evaluate pleiotropy in trait variance.
Results
Researchers identified 176 vQTLs linked to blood cell phenotypes, 147 of which are newly reported and were not detected in previous additive QTL studies. Basophil count and percentage accounted for the highest number of vQTLs.
These results underscore the distinct role of vQTLs in capturing phenotypic variability. Compared to additive QTLs, the newly identified vQTLs were 1.8 times more effective at detecting associations with extreme trait values.
In the INTERVAL cohort, individuals with high genetic variability showed a 19% improvement in prediction accuracy when vPGS was added to traditional polygenic scores, compared to individuals with low variability. On average, incorporating variance into PGS models improved prediction accuracy for blood cell traits by 10 %.
Among lifestyle variables, alcohol consumption was associated with increased variance in several blood cell phenotypes. Participants with high genetic variability in red blood cell count, mean corpuscular volume (MCV), and neutrophil percentage were more likely to report alcohol intake. This was supported by vQTLs rs191673261 and rs572454376, located near the ALDH2 gene, which plays a key role in alcohol metabolism and was associated with platelet crit (PCT).
The most pleiotropic vQTL—affecting the greatest number of traits—was found in the HBM gene and was linked to variance in mean corpuscular hemoglobin (MCH), MCH concentration (MCHC), MCV, and RBC count. Overall, the most statistically significant pleiotropic locus was LINC02768.
Several loci associated with alcohol intake also overlapped with vQTLs for platelet count, suggesting that alcohol may influence trait variance through genetic pathways.
Download your PDF copy now!
Implications and Future Directions
These findings suggest that variance in genetic expression, captured through vQTLs and vPGSs, can improve risk stratification for blood-related conditions. The study shows that individuals with higher genetic variability respond differently to genetic risk factors, and that lifestyle factors like alcohol intake may amplify these effects.
Integrating variance-based genetic models with existing risk assessments could improve the accuracy of disease prediction. Expanding this approach to more diverse populations and validating findings through experimental studies will be important next steps.
Journal Reference
Xiang, R., et al. Genome-wide analyses of variance in blood cell phenotypes provide new insights into complex trait biology and prediction. Nat Commun 16, 4260 (2025), DOI: https://doi.org/10.1038/s41467-025-59525-4, https://www.nature.com/articles/s41467-025-59525-4
New Genetic Discoveries Offer Hope for Osteoarthritis Treatments