Predictive model for mild cognitive impairment in older Chinese adults with depression

Neuropsychiatr. 2025 Jul 25. doi: 10.1007/s40211-025-00533-7. Online ahead of print.

ABSTRACT

BACKGROUND: Older adults with depression are at an increased risk of developing cognitive decline. This study aimed to develop and validate a risk prediction model for mild cognitive impairment (MCI) in older adults with depression in China.

METHODS: This study used 2020 China Health and Retirement Longitudinal Study (CHARLS) data, splitting the cohort (70:30) into training and validation sets. Least absolute shrinkage and selection operator (LASSO) regression with ten-fold cross-validation identified key predictors, and binary logistic regression examined MCI risk factors in older adults with depression. A nomogram was developed, with receiver operating characteristic (ROC) curves assessing discrimination, calibration curves for accuracy, and decision curve analysis (DCA) for clinical benefit.

RESULTS: This study included 3512 older adults with depression, 640 (19.9%) of whom had MCI. Binary logistic regression identified age, education level, marital status, residence, pain, internet use, and social participation as significant predictors of MCI in older adults with depression, and these factors were used to construct a nomogram model with good consistency and predictive accuracy. The area under the curve (AUC) values of the predictive model in the training set and internal validation set were 0.78 (95% confidence interval [CI] 0.75-0.80) and 0.75 (95% CI 0.71-0.78); the Hosmer-Lemeshow test results were P = 0.916 and P = 0.749, respectively. ROC analysis of the prediction model showed strong discriminatory ability, calibration curves demonstrated significant agreement between the nomogram model and actual observations, and DCA confirmed a favorable net benefit.

CONCLUSION: The nomogram constructed in this study is a promising and convenient tool for evaluating the risk of MCI among older adults with depression, facilitating early identification of high-risk individuals and enabling timely intervention.

PMID:40711693 | DOI:10.1007/s40211-025-00533-7