| Title: | Genetic determinants of serum creatinine and cystatin C independent of kidney function generate bias in filtration markers |
| Journal: | Kidney International |
| Published: | 19 Aug 2026 |
| Pubmed: | https://pubmed.ncbi.nlm.nih.gov/42617950/ |
| DOI: | https://doi.org/10.1016/j.kint.2026.07.018 |
Publication 20962
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
INTRODUCTION: Non-GFR determinants of filtration markers reduce the accuracy of estimated GFR (eGFR) and genetic factors may explain part of this reduced performance. Here, we identify genetic variants associated with serum levels of creatinine or cystatin C (but not GFR), compute their effect, and investigate candidate genes.</p>
METHODS: We used data from the UK Biobank (approximately 470,000 people) and the CKD-EPI creatinine (eGFRcr) and cystatin C (eGFRcys) equations. Using genome-wide association studies (GWAS), we defined strict criteria to classify loci as creatinine-specific or cystatin C-specific, refined with co-localization studies, and identified associated pathways. We computed biomarker and eGFR polygenic scores as the cumulative effect of biomarker-specific loci (genetic bias) and tested their associations to surrogate markers of identified pathways. Findings were validated by independent studies with measured GFR and mapped candidate genes using expression quantitative trait loci (eQTL) resources.</p>
RESULTS: We identified 52 creatinine-specific and 48 cystatin C-specific loci. Using eQTLs to identify candidate causal genes, enriched pathways were energy and muscle metabolism for creatinine, and inflammatory response, cancer, cysteine endopeptidase activity, wound healing, and immune modulatory functions for cystatin C. Polygenic scores showed that individual-level genetic bias in eGFRcr and eGFRcys had mean values of 3.8 (standard deviation 0.94 , range -1.5 to 7.9) mL/min per 1.73m2 and 0.6 (standard deviation 0.93, range -3.5 to 4.7) mL/min per 1.73m2 , respectively. Genetic bias was different per self-reported race for creatinine and correlated with surrogate markers of selected pathways (muscle mass, for creatinine; body mass index and C-reactive protein, for cystatin C). Lastly, we showed that the difference between polygenic scores derived from GWAS with and without adjustments for the biomarker-specific variants was associated with GFR bias in two independent cohorts (1304 and 939 individuals).</p>
CONCLUSIONS: Our study identified loci related to serum creatinine or cystatin C but not to GFR and quantified the populational and individual-level effect of genetic determinants on filtration markers and eGFRs.</p>
10 Authors
- Silvia M Titan
- Andrew D Rule
- Janet E Olson
- Joshua T Bublitz
- Tony C Luehrs
- Robert A Vierkant
- Hugues Sicotte
- Konstantinos N Lazaridis
- Vesna Garovic
- Alexandre C Pereira
1 Application
| Application ID | Title |
|---|---|
| 576095 | The association of genetic variants with kidney function and kidney diseases. |
Enabling scientific discoveries that improve human health