Abstract: FR-PO0543
Cluster Analysis of Routine Clinical Variables Identifies High-Risk Phenotypes for Rapid Progression of Diabetic Kidney Disease
Session Information
- CKM: Clinical - Trials, Epidemiology, and Biomarkers
October 23, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Cardiovascular-Kidney-Metabolic Health
- 602 Cardiovascular-Kidney-Metabolic Health: Clinical
Authors
- Teh, Yuan Kai, Khoo Teck Puat Hospital, Singapore, Singapore
- Low, Serena, Khoo Teck Puat Hospital, Singapore, Singapore
- Lim, Su Chi, Khoo Teck Puat Hospital, Singapore, Singapore
- Liu, Allen yan lun, Khoo Teck Puat Hospital, Singapore, Singapore
Background
Diabetic kidney disease (DKD) is a heterogeneous condition with substantial variability in the rate of kidney function decline. High metabolic burden, particularly insulin resistance is associated with accelerated DKD progression. We sought to identify individuals at high risk for rapid DKD progression using routine clinical variables, including age; hemodynamic measures, metabolic indices, and kidney parameters.
Methods
This was a retrospective cohort study on 3854 patients with type 2 diabetes. Data on demographics, clinical characteristics and medications was extracted from medical records. K-means clustering was performed based on the following classifiers: age, glycated haemoglobin (HbA1c), systolic blood pressure (SBP), Age, SBP, HbA1c, triglyceride-to-HDL cholesterol ratio (Tg/HDL), estimated glomerular filtration rate (eGFR); and urine albumin-to-creatinine ratio (UACR). Cox proportional hazards regression was done to examine the association between the clusters and follow-up kidney outcome defined as ≥ 40% decline in eGFR.
Results
The mean age was 57.3 ± 16.6. There were three clusters: Cluster 1 (no albuminuria and relatively preserved eGFR); Cluster 2 (albuminuria and low eGFR); and Cluster 3 (albuminuria and relatively preserved eGFR). After a mean follow-up period of 5.8 years (IQR 3.9-7.4), 13.2% and 6.2% of patients had ≥ 40% drop of eGFR and end stage kidney disease (ESKD) respectively. Using Cluster 1 as reference category and having adjusted for gender and ethnicity, Clusters 2 and 3 had increased hazards of ≥40% decline in eGFR with corresponding HRs 9.14 (95%CI 6.80, 12.28; p<0.001) and 2.43 (95%CI 1.89, 3.12; p<0.001). Similar findings were noted even after adjustment for sodium-glucose transporter 2 inhibitor (SGLT2i) and glucagon-like peptide-1 (GLP-1) receptor agonist.
Conclusion
Routinely available clinical variables can identify high-risk phenotypes associated with rapid DKD progression. This pragmatic approach may enable early identification of individuals who could benefit from intensified and targeted interventions to mitigate kidney disease progression.
Funding
- Commercial Support – Khoo Teck Puat Health Fund, Science-Translational and Applied Research (STAR 24101)