Association of Variability of Monthly Continuous Glucose Monitoring‐Derived Metrics With Diabetic Kidney Disease in Type 1 Diabetes
Authors' conclusion
The SD of monthly TIR and TITR is strongly associated with DKD in T1D, whereas the CV of sensor glucose is not.
Study context
- Population
- T1D patients with 1-year consecutive continuous glucose monitoring (CGM) data; Diabetic kidney disease (DKD) defined as persistent urine albumin-creatinine ratio ≥ 30 mg/g or eGFR < 60 mL/min/1.73 m² confirmed by at least two measurements within 1 year
- Participants
- 282 patients
- Exposure
- Machine learning (ML) models (LightGBM, XGBoost, Random Forest) vs Logistic Regression for discrimination of DKD from longitudinal CGM metrics
- Outcome
- Discrimination of DKD
Methodological detailsmachine learning, SHAP analysis
- Machine learning
- SHAP analysis
Main finding
As reported in the abstract
| Group | AUROC | F1 score |
|---|---|---|
| LightGBM | 0.91 (95% CI 0.88-0.93) | 0.65 |
LightGBM
AUROC0.91 (95% CI 0.88-0.93)
F1 score0.65
Observational data: associations, not effects.
Further findings
| Outcome | P value |
|---|---|
| CV of sensor glucose between groupsDKD vs no DKD · did not differ significantly | P = 0.416 |
Influential features (SHAP)
- SD of monthly time in range (TIR) and time in tight range (TITR) were the most influential features
SD of monthly TIR and TITR in early DKD subgroup
- SD of monthly TIR and TITR remained significantly elevated
Original publication
Association of Variability of Monthly Continuous Glucose Monitoring‐Derived Metrics With Diabetic Kidney Disease in Type 1 Diabetes. Diabetes Obes Metab. Published 14 July 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →