Knee osteoarthritis (KOA) combined with depressive symptoms is prevalent and leads to poor outcomes and significant financial burdens. However, practical tools for identifying at-risk patients remain limited. A robust prediction model is needed to address this gap. This study aims to develop and validate a predictive model to identify KOA patients at risk of developing depressive symptoms. The China Health and Retirement Longitudinal Survey (CHARLS) data were used for model development and the Osteoarthritis Initiative (OAI) for external validation. 18 potential predictors were selected using LASSO regression. 4 machine learning models—logistic regression, decision tree, random forest, and artificial neural network—were developed. Model performance was assessed using the area under the operating characteristic curve (AUC), calibration curves, and decision curve analysis. The most important features were extracted from the optimal model on external validation. A total of 469 individuals were included, with 70% used for training and 30% for testing. The random forest model achieved the best performance, with an AUC of 0.928 in the test set, outperforming logistic regression (AUC 0.622), decision tree (AUC 0.611), and neural network models (AUC 0.868). External validation revealed an AUC of 0.877 (95% CI: 0.864–0.889) for the adjusted random forest model. Pain severity was the most significant predictor, followed by the five-time sit-to-stand test (FTSST) and sleep problems. This study is the first in China to apply a predictive model for depressive symptoms in KOA patients, offering a practical tool for early risk identification using routinely available data.
Read full abstract