Development and validation of a depression risk prediction model for rural elderly living alone

Depression is a prevalent psychological issue among rural elderly individuals living alone, severely impacting their physical and mental health.

To develop and validate a depression risk prediction model for rural elderly living alone based on the health ecological model, providing a scientific basis for early intervention.

Using data from the 2011 China Health and Retirement Longitudinal Study (CHARLS), we included 1,221 participants. Thedataset was randomly stratified into a training set (70%) and a validation set (30%). Predictors were screened via univariate analysis, followed by multivariate logistic regression to construct the nomogram model. Statistical analysis was performed using R Studio 4.4.1.Ten-fold cross-validation was used to assess the model's stability. Model performance was evaluated using the Receiver Operating Characteristic (ROC) curve, with the Area Under the Curve (AUC) calculated, along with calibration plots, the Hosmer-Lemeshow test, and Decision Curve Analysis (DCA).

Self-rated health, pain, frailty, nighttime sleep duration, poor sleep quality, life satisfaction, and visit frequency were identified as independent predictors of depressive symptoms. The model demonstrated excellent discrimination (AUC = 0.85 [95% CI: 0.83-0.88] in the training set and 0.83 [95% CI: 0.78-0.87] in validation), good calibration (Hosmer-Lemeshow test p = 0.47), and high clinical utility (net benefit > 10% in DCA).

The nomogram provides a reliable and intuitive tool for early screening of depressive symptoms in rural elderly individuals living alone, supporting targeted interventions.

Not applicable.