📄 中文摘要
超重和肥胖在儿童和青少年中仍然是美国主要的公共卫生问题，受到行为、家庭和社区因素的影响。研究旨在识别美国青少年超重和肥胖的多层次预测因素，并比较统计模型、机器学习模型和深度学习模型的预测性能、校准度和亚组公平性。分析数据来自2021年国家儿童健康调查，共涉及18,792名10至17岁的儿童。超重和肥胖的定义基于BMI分类，预测因素包括饮食、身体活动、睡眠和父母压力等。这项研究为理解青少年超重和肥胖的复杂性提供了新的视角，并为制定干预措施提供了数据支持。
美国10-17岁儿童超重和肥胖的多层次决定因素：基于2021年国家儿童健康调查的统计与机器学习方法的比较评估
📄 English Summary
Multilevel Determinants of Overweight and Obesity Among U.S. Children Aged 10-17: Comparative Evaluation of Statistical and Machine Learning Approaches Using the 2021 National Survey of Children's Health
Childhood and adolescent overweight and obesity are significant public health concerns in the United States, influenced by behavioral, household, and community factors. This study identifies multilevel predictors of overweight and obesity among U.S. adolescents and compares the predictive performance, calibration, and subgroup equity of statistical, machine learning, and deep learning models. Data from the 2021 National Survey of Children's Health, involving 18,792 children aged 10-17 years, were analyzed. Overweight and obesity were defined using BMI categories, with predictors including diet, physical activity, sleep, and parental stress. The findings provide new insights into the complexities of adolescent overweight and obesity and offer data-driven support for intervention strategies.