Development and Validation of a Predictive Depression Model in Cancer Survivors Using CHARLS Data

Wang P, Jin X, Luo S, Li C, Liu Y, Liu X and Wang X

Department of Gynecology, The First Clinical Medical College of Lanzhou, China
Department of Pediatric Orthopedics, Shengjing Hospital of China Medical University, China
Institute of Neuroscience, Chongqing Medical University, China
Institute of Health Science Center, Ningbo University, China
Department of Pediatric Surgery, Wuhan Children's Hospital (Wuhan Maternal and Child Healthcare Hospital), Tongji Medical College, Huazhong University of Science & Technology, China
These authors contributed equally to this work

^*Correspondance to: Xiaohui Wang 

 PDF  Full Text Research Article | Open Access

Abstract:

Background: Depression significantly impacts the recovery and quality of life of cancer survivors. This study developed and validated a prognostic model to assess depression risk using data from the China Health and Retirement Longitudinal Study (CHARLS) and five external medical centers. Methods: Data from 574 cancer survivors in the CHARLS dataset were analyzed, with 216 diagnosed with depression. An additional cohort of 503 cancer patients from five medical centers served as the external validation group. Predictive factors included demographic characteristics, lifestyle habits, and medical history. Least Absolute Shrinkage and Selection Operator (LASSO) regression was used to refine the variables, and multivariate logistic regression identified significant predictors. Model performance was evaluated using the C-index, calibration curves, Hosmer-Lemeshow test, Decision Curve Analysis (DCA), and Clinical Impact Curve (CIC). Results: Five independent predictors of depression were identified: Rural residency, self-reported health, arthritis, memory efficiency, and life satisfaction. The model demonstrated excellent discrimination in the CHARLS cohort (C-index 0.854; 95% CI: 0.821-0.887), test group (C-index 0.902; 95% CI: 0.784-0.915), and external validation group (C-index 0.827; 95% CI: 0.762-0.892). The Hosmer-Lemeshow test yielded a value of 0.910. Calibration curves, DCA, and CIC analyses further confirmed its predictive accuracy and clinical relevance. Conclusion: The developed model is a precise and clinically significant tool for predicting depression risk among cancer survivors. It enables early identification of high-risk individuals, facilitating timely and appropriate interventions.

Keywords:

Depression; Cancer; Prediction model; Nomogram; CHARLS

Cite the Article:

Wang P, Jin X, Luo S, Li C, Liu Y, Liu X, et al. Development and Validation of a Predictive Depression Model in Cancer Survivors Using CHARLS Data. Clin Oncol. 2024;9:2084..