What question did this study set out to answer?

May 13, 2026Open Access

Development and Validation of a Nomogram for Predicting Chronic Kidney Disease in Older Patients with Type 2 Diabetes Mellitus and Cardiovascular Disease

Key Points

The aim is to develop and validate a nomogram predicting chronic kidney disease in older patients with type 2 diabetes mellitus and cardiovascular disease.
Developed a predictive model from electronic health records of 28,443 older inpatients with type 2 diabetes and cardiovascular disease.
Population divided into training set (n=21,234) and validation set (n=7,209).
Utilized univariate and multivariate logistic regression analyses to select predictive variables, visualized with a nomogram.
Identified seven independent predictors of chronic kidney disease: systolic blood pressure, uric acid, hemoglobin, glycated hemoglobin, white blood cell count, triglyceride glucose, and UA/Cr ratio.
Nomogram exhibited AUC of 0.845 in training set and 0.853 in validation set, indicating strong discriminative ability.
Decision curve analysis showed the model provides net benefit across various threshold probabilities.

Abstract

Background: Type 2 diabetes mellitus (T2DM) and cardiovascular diseases (CVD) are significant global health challenges, particularly in older adults where their coexistence exacerbates disease burden and leads to adverse outcomes. Chronic kidney disease (CKD), a severe complication of T2DM, is increasing in prevalence and significantly contributes to morbidity and mortality. Early detection of CKD in older patients with T2DM and CVD is crucial for timely intervention. Methods: This study developed a predictive model using electronic health record data from 28,443 older inpatients with T2DM and CVD at the Affiliated Banan Hospital of Chongqing Medical University (2018– 2025). The study population was divided into training (n=21,234) and validation (n=7,209) sets. Predictive variables were selected through univariate logistic regression, LASSO regression, and multivariate logistic regression analyses. The model was visualized using a nomogram and validated using ROC curves, calibration curves, and decision curve analysis (DCA). Results: The model identified seven independent predictors of CKD: systolic blood pressure, uric acid (UA), hemoglobin, glycated hemoglobin, white blood cell count, triglyceride glucose, and UA/Cr ratio. The nomogram demonstrated good discriminative ability with AUCs of 0.845 (95% CI: 0.836– 0.853) in the training set and 0.853 (95% CI: 0.838– 0.867) in the validation set. Calibration curves showed good agreement between predicted and observed risks. DCA indicated that the model provided a net benefit across a range of threshold probabilities, highlighting its clinical utility. Conclusion: The developed nomogram provides a practical tool for clinicians to predict CKD risk in older patients with T2DM and CVD using readily available clinical data. This model can facilitate early identification and intervention for high-risk patients, potentially improving outcomes. Keywords: type 2 diabetes mellitus, cardiovascular disease, chronic kidney disease, predictive model, nomogram

Bookmark

View Full Paper