Background
A global public health issue, chronic kidney disease(CKD) may worsen more quickly if depression symptoms overlap. In this study, a nomogram model was developed and validate to predict depression in Chinese CKD patients between the ages of middle-aged and old.
Methods
A 7:3 random split of the 1571 participants in the China Health and Retirement Longitudinal Study was made into training and validation sets. After doing Least Absolute Shrinkage and Selection Operator(LASSO) and multivariate binary logistic regression analysis to discover determinants of depression symptoms. These predictors were used to create a nomogram, which was then evaluated for discriminative power, predictive performance, and clinical applicability using receiver operating characteristic (ROC) curves, calibration curves, Hosmer-Leme show tests, and decision curve analysis (DCA).
Results
The nomogram model included 10 predictors, including gender, marital status, place of residence, education level, life satisfaction. pain, sleep disorders, self-reported health, as well as comorbid chronic diseases. The Area under the curve(AUC) values of the training and validation sets were, in turn, 0.889 (95% CI: 0.869–0.908) and 0.869 (95% CI: 0.836–0.902), the values of Hosmer–Lemeshow test were p = 0.113 and p = 0.259. The calibration curves and the Hosmer-Lemeshow test results were used to verify the nomogram model's predictive capabilities. Additionally, the decision curve analysis (DCA) curves illustrated a high net clinical benefit provided by the predictive model.
Conclusions
We developed and validated a depression risk model for middle-aged and elderly CKD patients. Clinicians can accurately screen middle-aged and older CKD patients having depressive symptoms using the evaluation instrument, which is important for early intervention.