Construction and validation of a depression risk prediction model in middle-aged and elderly patients with diabetes
10.3760/cma.j.cn115682-20250224-00848
- VernacularTitle:中老年糖尿病患者抑郁风险预测模型的构建与验证
- Author:
Lei YANG
1
;
Yaping HAO
;
Yuxiao TANG
;
Juntao CHI
;
Lingyan ZHAO
;
Guiqin GU
;
Liang WANG
Author Information
1. 烟台毓璜顶医院内分泌科，烟台 264000
- Publication Type:Journal Article
- Keywords:
Diabetes mellitus;
Middle-aged and elderly;
Depression;
Prediction model;
China Health and Retirement Longitudinal Study
- From:
Chinese Journal of Modern Nursing
2025;31(29):3976-3983
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To construct and validate a depression risk prediction model for middle-aged and elderly patients with diabetes.Methods:Data were extracted from the fifth wave (2020) of the China Health and Retirement Longitudinal Study (CHARLS). A total of 900 diabetic patients were identified, and after excluding those with missing data or invalid questionnaires, 769 patients were included in the analysis. Patients were randomly divided into a training set and a validation set in a 7∶3 ratio. Univariate analysis and logistic regression analysis were performed to screen the optimal predictors of depression in diabetic patients, and a nomogram model was developed. The predictive performance of the model was assessed by the area under the receiver operating characteristic curve ( AUC). Model calibration and accuracy were evaluated using bootstrap resampling, calibration plots, and the Hosmer-Lemeshow test. The clinical utility was further assessed by decision curve analysis (DCA) and clinical impact curves (CIC) . Results:Among the 769 patients, 366 (47.59%) had depression. Logistic regression analysis showed that place of residence, pain, difficulty in toileting, difficulty in bathing, sleep duration, physical exercise, life satisfaction, and children's satisfaction were independent predictors of depression in diabetic patients. A nomogram was constructed based on these variables, yielding an AUC of 0.775. At the optimal cutoff value of 0.557, the model demonstrated a sensitivity of 59.1% and a specificity of 84.8%, indicating good discriminative ability. The Hosmer-Lemeshow test showed (χ 2=15.821, P=0.105), suggesting good agreement between predicted and observed outcomes. In the validation set, the AUC was 0.778, with Hosmer-Lemeshow (χ 2=8.557, P=0.575). DCA and CIC indicated favorable clinical applicability of the model. Conclusions:The depression risk prediction model constructed in this study demonstrated good predictive performance. It can assist clinicians in early identification of high-risk individuals with diabetes and provide a theoretical basis for targeted interventions.
