Frailty Prediction Model for Elderly Diabetic Peripheral Neuropathy Patients

Elderly patients with diabetic peripheral neuropathy (DPN) are significantly impacted by frailty, yet frailty prediction models for this population remain underexplored. This study aims to develop and internally validate a frailty prediction model for elderly patients with DPN.

A cross-sectional study design was employed, and 400 elderly DPN patients were recruited from a tertiary hospital in Guangdong Province, China, between December 2024 and July 2025. Logistic regression was employed to identify frailty risk factors and develop a prediction model and nomogram for elderly DPN patients. We evaluated the performance of the model using the area under the receiver operating characteristic (ROC) curve, abbreviated as AUC, and was further assessed through the Hosmer-Lemeshow test and calibration curves. The clinical utility of the model was assessed by decision curve analysis (DCA). Internal validation was performed using 1000 bootstrap resamples to reduce the risk of overfitting.

Among the 400 patients, 113 (28.25%) patients had frailty. Six factors were identified as significant predictors: age, marital status, regular exercise, PSQI score, MNA-SF score, and HADS-D score. We constructed a nomogram based on these factors. Internal validation demonstrated good performance in both discrimination and calibration, and DCA confirmed the model's clinical applicability.

The nomogram developed in this study provides an effective tool for the early identification of elderly DPN patients at risk of frailty, thereby informing tailored preventive and intervention strategies. External validation will be conducted in future studies, and future studies will assess the model's generalizability across different regions and healthcare systems. The main predictors identified in this study include age, marital status, regular exercise, PSQI score, MNA-SF score, and HADS-D score, which significantly contribute to frailty risk in elderly DPN patients.