Abstract: SA-OR015
A Novel Urine Metabolite Risk Score for Predicting Cardiovascular Disease in Black Adults with CKD
Session Information
- CKD: Advancing Epidemiology, Risk Factors, and Prevention
November 08, 2025 | Location: Room 362A, Convention Center
Abstract Time: 05:20 PM - 05:30 PM
Category: CKD (Non-Dialysis)
- 2301 CKD (Non-Dialysis): Epidemiology, Risk Factors, and Prevention
Authors
- Li, Changwei, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- Zhang, Ruiyuan, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- Sun, Yixi, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- Liu, Yizhuo, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- Huang, Zhijie, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- Grams, Morgan, New York University Grossman School of Medicine, New York, New York, United States
- Kelly, Tanika, University of Illinois Chicago, Chicago, Illinois, United States
- Chen, Jing, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
- He, Jiang, The University of Texas Southwestern Medical Center, Dallas, Texas, United States
Background
Cardiovascular disease (CVD) is the leading cause of death in patients with chronic kidney disease (CKD). However, existing CVD risk prediction models developed in the general population perform poorly in CKD. While blood proteomic tools have shown promise in CKD-specific risk prediction, they rely on invasive sampling. Urinary metabolites offer a non-invasive alternative that may reflect underlying CVD risk. We aimed to develop a urine-based metabolite risk score (MRS) to predict incident CVD among Black participants of the Chronic Renal Insufficiency Cohort (CRIC) study.
Methods
Baseline 24-hour urine samples from 776 Black CKD patients without CVD were profiled using untargeted metabolomics. Incident CVD, defined as the onset of congestive heart failure, myocardial infarction, or stroke, was ascertained over a median follow-up of 10.16 years. Elastic net regression identified metabolites associated with incident CVD, and their coefficients were used to construct the MRS. Model training and testing was conducted with an 80:20 data split, followed by leave-one-out cross-validation to generate individual MRS values. Associations between the MRS, incident CVD, and traditional cardiovascular risk factors were evaluated using Cox and linear regression models. Discrimination was assessed using the concordance index (C-index), and sensitivity analyses included adjustment for APOL1 genotype.
Results
The final MRS included 57 urinary metabolites and was significantly correlated with traditional CVD risk factors (e.g., kidney function, cardiometabolic markers; all P<0.05). The MRS was strongly associated with incident CVD (HR per SD increase=2.06, 95% CI: 1.81-2.34; P<0.0001), with moderate discriminatory performance (C-index=0.708). Adding traditional cardiovascular risk factors improved discrimination (C-index=0.780), and the MRS remained an independent predictor of CVD (HR=1.25, 95% CI: 1.00-1.56; P=0.05). Findings remained robust in sensitivity analyses adjusting for APOL1 status.
Conclusion
This novel urine-based MRS independently predicted incident CVD in Black CKD patients, even after accounting for traditional cardiovascular risk factors. Its non-invasive nature supports its potential use in clinical settings to enhance early CVD risk stratification in high-risk populations.