Abstract: FR-PO0256
Selection Bias Precludes Clinical Use of Kidney Function Polygenic Scores in CKD Populations
Session Information
- CKD: Omics, Systemic Stressors, and Targeted Pharmacotherapy
October 23, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: CKD (Non-Dialysis)
- 2201 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention
Authors
- Donovan, Killian, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Torres, Jason Matthew, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Zhu, Doreen, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Staplin, Natalie, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Che, Zhaojing, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Emberson, Jonathan R., University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Haynes, Richard, University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
- Herrington, William G., University of Oxford Nuffield Department of Population Health, Oxford, England, United Kingdom
Group or Team Name
- Renal Studies' Group
Background
Polygenic scores (PGS) that predict risk of low eGFR in the general population are available, but their associations with clinical outcomes in established CKD have not been well characterised. In cohorts defined by prevalent disease, selection bias may distort associations between exposures, including PGS, and clinical outcomes.
Methods
Analyses included participants of European ancestry from UK Biobank and three CKD cohorts (SHARP, CRIC and EMPA-KIDNEY) without ESKD at baseline. The exposure was a PGS for low eGFR derived from a GWAS of eGFR in the FinnGen cohort (n=307,398). Primary outcomes were eGFR slope (ml/min/1.73m2/year) and kidney disease progression (KDP), defined as first occurrence of dialysis, transplantation, sustained ≥40% decline in eGFR, or sustained eGFR <10ml/min/1.73m2. Associations with eGFR slope and KDP were estimated using linear regression and Cox proportional hazards models, respectively, and meta-analysed across studies.
Results
107,536 participants from UKB and 6,846 participants from CKD cohorts were included. In UKB, a higher PGS was associated with increased risk of KDP (HR 1.36 [95% CI 1.20 to 1.55] for those in the top 10% of the PGS versus the rest) and faster eGFR decline (-0.12 [-0.14 to -0.09] ml/min/1.73m2/year). In contrast, in CKD cohorts, a higher PGS was associated with lower risk of KDP (HR 0.76 [0.66 to 0.89]) and slower eGFR decline (+0.19 [-0.01 to 0.39] ml/min/1.73m2/year; Fig 1). In UKB, higher PGS was associated with a small reduction in uACR (-5.2% [-6.6 to -3.8]), whereas in CKD cohorts this association was substantially stronger (-34.7% [-43.4 to -24.6]).
Conclusion
A PGS for low eGFR was associated with faster eGFR decline and higher risk of KDP in the general population. However, in CKD populations, selection bias appears to induce a strong inverse association with uACR, reversing the observed associations. These findings suggest that applying PGS in populations selected for CKD may lead to misleading conclusions regarding associations with clinical outcomes.
Acknowledgment
Funding:
Kidney Research UK Clinical Training Fellowship TF_001_20220708
Wellcome Trust Core Award Grant Number 203141/Z/16/Z
NIHR Oxford BRC
Acknowledgements: The CRIC study was conducted by the study investigators and supported by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). The resources from the CRIC study reported here were supplied by NIDDK Central Repository (NIDDK-CR) and are available for request at https://repository.niddk.nih.gov.
Funding
- Private Foundation Support