目的：探索利用社会生态模型，从“学生-学校-家庭”三个层面对学生肥胖进行干预是否有效。方法：利用分层整群抽样方法选择北京市房山区城市小学、乡村小学、城市中学和乡村中学各4所，共16所学校7～18岁学生共得到有效样本3 175人，每层4所学校各选取2所学校共1 661人，另2所学校作对照组，利用社会生态模型从“学生-学校-家庭”三个层面进行为期3个月的肥胖干预。干预方式包括健康教育、环境改善等，干预内容涉及肥胖相关知识、健康饮食与运动等。检测干预前后身体形态指标。结果：干预可以整体改善超重肥胖（OR=1.12，P<0.05），降低腰围、腰臀比（waist-hip ratio，WHR）等中心性肥胖指标（adjusted difference=0.63、0.02，P<0.05）；女生腰围和WHR降低（adjusted difference=0.52、0.02，P<0.05）并改善了超重肥胖（OR=1.18，P<0.05），男生腰围和WHR降低（adjusted difference=0.73、0.01，P<0.05），女生效果优于男生；城市地区WHR降低（adjusted difference=0.01，P<0.05），乡村地区腰围和WHR降低（adjusted difference=1.05、0.02，P<0.05）并改善了超重肥胖（OR=1.43，P<0.05），乡村效果优于城市；小学WHR降低（adjusted difference=0.02，P<0.05）并改善了超重肥胖（OR=1.21，P<0.05），中学WHR降低（adjusted difference=0.02，P<0.05）， 小学效果优于中学；干预措施有效降低超重肥胖学生的超重肥胖率（χ²=11.01，P<0.01）。结论：“学生-学校-家庭”社会生态模式干预方法短期内可改善腰围、WHR等中心性肥胖指标，可以作为一种有效干预手段推广使用。

Objective： To study whether the socio-ecological model based on “student-school-family” three-level strategy is effective in obesity prevention.Methods: A total of 3 175 students aged 7 to 18 from 16 schools (4 urban primary schools, 4 rural primary schools, 4 urban secondary schools and 4 rural secondary schools, of which 2 intervention schools were selected, respectively ) were recruited by stratified cluster sampling method. A threemonth intervention using “student-school-family” socio-ecological model was conducted through health education and environment improvement. The intervention contents included knowledge on obesity, healthy diet and physical activities. Their anthropometric indexes were recorded. Results: The intervention prevented obesity(OR=1.12, P<0.05), and was effective in waist circumference (WC) and waist-hip ratio (WHR) (adjusted difference=0.63, 0.02, P<0.05). WC and WHR were reduced in girls (adjusted difference=0.52 & 0.02, P<0.05), and obesity was prevented in girls (OR=1.18, P<0.05). WC and WHR were reduced in boys (adjusted difference=0.73, 0.01, P<0.05). WHR were reduced in urban areas (adjusted difference=0.01, P<0.05). WC and WHR were reduced (adjusted difference=1.05, 0.02,P<0.05) and obesity was prevented (OR=1.18, P<0.05) in rural areas. WHR were reduced (adjusted difference=0.01, P<0.05) and obesity was prevented (OR=1.21, P<0.05) in primary schools. WHR were reduced in secondary schools (adjusted difference=0.02, P<0.05).The intervention effect was better in girls than in boys, in rural areas than in urban areas, and in primary schools than in secondary schools. The overweight and obesity prevalence went down after the intervention (χ2=11.01，P<0.01).   Conclusion:Intervention strategy is effective in central obesity indexes such as WC and WHR, and it can be used widely.