目標:探討學童從一至六年級之身體質量指數(BMI)的發展軌跡類型及國小入學後首次超重的時間,並進一步找出相關因素。方法:運用兒童與青少年行為長期發展研究計畫,2001年就讀國小一年級至2006年時完整追蹤之研究樣本,且有學童及家長之身高及體重資料,共1609人。採用生態模式之概念,從個人、家庭、學校、社區等不同層級,找出與學童BMI發展軌跡類型有關之因素。運用群體基礎軌跡模型和多項式邏輯迴歸模型,進行統計分析。為探討國小入學後首次超重時間及相關因素,特將就讀國小入學前已超重者予以排除,留下1249人進行存活分析。結果:1.學童從國小一至六年級之BMI發展軌跡類型,因性別而有顯著差異。2.男性樣本之BMI發展軌跡,可區分為正常稍輕組(40.62%)、持續正常組(34.69%)、過重變肥胖組(18.15%)和持續肥胖組(6.54%)四種類型;女性樣本之BMI發展軌跡可以區分為持續稍輕組(31.04%)、持續正常組(40.18%)、持續過重組(22.03%)和持續肥胖組(6.76%)四種類型,各組BMI之平均值隨時間呈現上升趨勢。3.與BMI發展軌跡類型為超重有關之變項,在男性樣本為:課後做運動、自覺學業表現、家人互動、父母超重情形、父親教育程度;在女性樣本為:看電視及用電腦、家人互動、同儕互動、父母超重情形。4.以國小入學後首次超重時間而言,男性(5.24年)比女性(5.44年)來得早;居住在新竹縣者(5.27年)比臺北市者(5.39年)來得早。5.與國小入學後首次超重時間較早有關的因素,在男性樣本為:父母有超重、與同儕互動較高、與社區互動較低;在女性樣本為:父母有超重、青春期發育較早、自覺同儕互動較高、居住在新竹縣者,家庭為低收入組者。結論:國小學童的BMI發展軌跡類型,及國小入學後首次超重的時間,皆有性別差異。針對於國小學童過重及肥胖的問題,需從一年級,甚至學齡前就開始關注,並及早推動健康促進政策及介入措施。根據本研究發現,建議兒童課後多做運動、減少看電視及用電腦、留意與家人的互動內容、注意與同儕互動內容、和多與社區互動。
Objective: To investigate developmental trajectory patterns of body mass index (BMI) and onset of overweight from the 1st to 6th grade. Methods: Data was obtained from the Children and Adolescent Behaviors in Long-term Evolution project. Participants included 1,609 students who were followed from the 1st to 6th grade (2001-2006). Data included height and weight of students and their parents. The concept of an ecological model was used to indentify the individual, family, school, and community level factors associated with BMI developmental trajectories among students. Group-based trajectory models and multinomial logit models were used in the statistical analysis. To explore the onset of overweight and related factors during elementary school years, we excluded students who were overweight at the beginning of 1st grade, leaving a total of 1249 participants for the survival analysis. Results: 1.There were significant sex differences in BMI developmental trajectory patterns from the 1st to 6th grade. 2. Among boys the four BMI developmental trajectory patterns included the normal that became slightly underweight group (40.62%), the persistently normal group (34.69%), the overweight that became obese group (18.15%), and the persistently obese group (6.54%). Among girls, the BMI developmental trajectory patterns included the persistently slightly underweight group (31.04%), the persistently normal group (40.18%), the persistently overweight group (22.03%), and the persistently obese group (6.76%). The mean BMI in each group demonstrated an upward trend over time. 3. In boys, developmental trajectories of overweight were significantly associated with after-school exercise, academic performance, family interactions, having overweight parents and father's education level. In girls, developmental trajectories were significantly associated with watching television and using the computer, family interactions, peer interactions, and having overweight parents. 4. The onset of overweight during elementary school among boys ( 5.24 years) was earlier than that in girls (5.44 years). In addition, the onset of overweight occurred earlier in those living in Hsinchu County (5.27 years) compared to Taipei city (5.39 years). 5. In boys, early onset of overweight during elementary school was associated with having overweight parents, higher peer interactions, and less community interactions. In girls, early onset of overweight was associated with having overweight parents, earlier puberty, higher peer interactions, living in Hsinchu County, and coming from a low-income household. Conclusion: Among elementary school students, different patterns of BMI developmental trajectories exist in boys and girls. Similarly, the onset of overweight in elementary school differs by gender. Children who are overweight and obese in the 1st grade, or even in preschool, are an important target for health promotion policies and interventions. Based on our findings we recommend that children be encouraged to participate more in after-school exercise, reduce their time spent watching TV and using computers, pay attention to the content of their interactions with family and peers, and increase their community-level interactions.