PREDICTIVE MODEL FOR OPTIMAL SELF-MANAGEMENT STRATEGIES IN ATHLETES WITH METABOLIC SYNDROME

Abstract

Background: Effective self-management of metabolic syndrome (MS) is crucial for maintaining optimal health and performance, especially in athletes. Developing a predictive model for health self-management behaviors in individuals with MS can significantly enhance intervention strategies and outcomes. Objectives: To develop a risk prediction model for health self-management behaviors tailored to individuals with MS, focusing on a population in southern Xinjiang, with an emphasis on applicability to athletes. Method: Utilizing physical examination data from 10,614 participants of a national health examination in southern Xinjiang from February to July 2022, 906 individuals diagnosed with MS were included. The dataset was divided into an 80% training set and a 20% validation set. Logistic regression (LR), decision tree (DT), and backpropagation (BP) neural network methods were employed to construct the models. Model effectiveness was evaluated based on AUC, Kappa value, accuracy, specificity, and sensitivity. The Delong test was used to compare the AUC values across the models. Results: Lasso regression identified five high-risk factors as significant predictors: knowledge of MS prevention and treatment, symptom management self-efficacy, tension, loss of control, and subjective support. The constructed models using LR, DT, and BP neural networks exhibited strong predictive performance, with AUC values of 0.839 (95%CI: 0.791, 0.887), 0.873 (95%CI: 0.828, 0.919), and 0.849 (95%CI: 0.801, 0.897) in the validation set, respectively. The differences among the models' AUC values were not statistically significant according to the Delong test. Discussion: This study provides a robust framework for assessing the risk of poor health self-management behaviors in individuals with MS, with potential applications for optimizing health management strategies in athletic populations. By focusing on key predictive factors, sports medicine professionals can better tailor interventions to support athletes in managing MS effectively, thus maintaining their health and enhancing their performance.