Construction and Validation of a Machine Learning-Based Risk Prediction Model for Sleep Disorders Among Middle-Aged and Elderly Residents in Xining Region
FAN Xiaowei
DONG Bixuan
HAN Shukui
REN Yitao
CHEN Hongru
LI Ze
DANG Zhancui
LI Bin
Abstract：Objective To construct a risk prediction model for sleep disorders among middle-aged and elderly people in Xining region.Methods A total of 673 residents aged ≥ 45 years were included.Feature selection was performed using LASSO regression,Boruta algorithm,and random forest(RF).Models were constructed using logistic regression,decision tree(DT),RF,eXtreme gradient boosting(XGBoost),and support vector machine(SVM).Bootstrap resampling method(1 000 repeated samplings)was used for internal validation.The models' performance was comprehensively evaluated using receiver operating characteristic(ROC)curve,Brier scores,calibration curves,and decision curve analysis.Sensitivity analysis assessed the impact of outlier processing,missing value imputation,and feature selection methods on the optimal model's performance.Results Feature selection identified gender,comorbidities,annual income,depression,and anxiety as key predictors of sleep disorders.Among the five machine learning models,the logistic regression model performed best with an area under the curve(AUC)of 0.777(95％CI:0.767-0.779)and a Brier score of 0.188,demonstrating good discrimination and calibration performance.Interpretability analysis showed that depression contributed most significantly to sleep disorder prediction.Sensitivity analysis demonstrated that outlier processing,missing value imputation,and feature selection all influenced model performance.Conclusion This study constructed a risk prediction model for sleep disorders applicable to middle-aged and elderly people in Xining region,providing a practical tool for clinicians to identify high-risk populations and providing a basis for early prevention and intervention of sleep disorders.
Keywords：machine learningprediction modelsleep disordersmiddle-aged and elderlyXining
Publication Date：2026-03-13
Online Publishing Date：2026-09-12(First online date of this platform, not the publication date of the document)
Pages：9( 788-796 )