國中生血脂之個體及學校層次之多層次效應研究

血脂包含總膽固醇、HDL-C、LDL-C及三酸甘油脂。儘管青少年學童肥胖狀況已有文獻證明會隨著學校及附近環境不同而有所差異，但與肥胖有密切相關的血脂在此部分卻鮮少被研究，本文章為探討影響國中青少年學童血脂之個體因素以及學校及學校鄰區因素。以多步驟分層抽樣抽取高雄市共15間國中以及台東縣共8間國中，其中有74.6%、1939名青少年學童參與抽血。我們同時測量其人體測量學項目並以問卷詢問包含飲食、運動等習慣之個體因素。另外，也調查了學校及學校鄰區因素，包含了校內健康促進活動、飲食提供及運動設施，校外300公尺、600公尺以內飲食提供場所及運動休閒設備。我們以多層次迴歸分析去評估血脂的多層次效應。先以一般迴歸分析後，青少年學童攝取蔬果、從事運動與總膽固醇、LDL-C有顯著負相關，與HDL-C呈現顯著正相關。飲用飲料之行為則與LDL-C、三酸甘油脂之增加有關。利用多層次線性迴歸分析發現，國中青少年學童總膽固醇及HDL-C之變異分別約有9.4%以及19.6%可歸因於學校及學校鄰區層次變異。近一步來看，學校舉辦健康促進活動次數可以解釋總膽固醇知變異情形，而學校外300公尺內之運動場地數量也可以對HDL-C之變異做解釋。另外以多層次羅吉斯迴歸分析，則發現總膽固醇與HDL-C之過高(HDL-C則為過低)比率變異分別約有6.0%以及30.0%可歸因於學校及學校鄰區層次變異。LDL-C與三酸甘油脂則在多層次分析中則未達到顯著，此研究強調控制青少年學童血脂之策略不僅應從個人飲食及運動習慣著手，也要在其學校及學校鄰區多舉辦健康促進等活動、並且提供青少年學童良好的飲食環境及足夠的運動設施。

關鍵字

總膽固醇； HDL-C ； LDL-C ；三酸甘油脂；青少年學童；多層次迴歸分析

並列摘要

Blood lipids include triglyceride, high- (HDL), low-density and total cholesterols . Although studies about obesity of adolescent and children have been demonstrated to have variant among different schools, but little studies have been done about blood lipids that are closely associated with obesity. This study investigated individual- and school-level factors that may affect blood lipids of junior high school students. We use multi-stage stratified sampling method to recruit 1939 students who provided their blood samples (response rate, 74.6%) from 15 schools in Kaohsiung City and 8 schools in Taitung County, we also measured their anthropometric items and asked their individual-level factors including dietary and exercising habits by using questionnaire. School-level factors including health promotion activities, food-related providers, and sport facilities inside the school and around the schools within 300 and 600 meters have also been recorded. We used multilevel regression models to assess multilevel effects on blood lipids. At first by using general regression analysis, we found that there is significantly negative association between vegetable intake and exercising with total cholesterol and LDL-C, but is significantly positive association with HDL-C; and we also found that beverage-drinking may induce the increasing of LDL-C and triglyceride level. By using multilevel linear regression analysis, differences about 9.4% of total cholesterol and about 19.6% of HDL-C between junior high school students were attributed to the school-level variations. At further step, we found that the variation of total cholesterol can be explained by the frequency of health promotion activities hold by school and the variation of HDL-C can be explained by the amount of exercising facilities within 300 meter around school. Besides, by using multilevel logistic regression analysis, about 6.0% of higher proportion of total cholesterol and about 30.0% of lower proportion of HDL-C between junior high school students were attributed to the school-level variations. There were no significant attributions of school-level variation in LDL-C and triglyceride. This study emphasize that developing the strategies for controlling the high-level of blood lipids among adolescent not only including individual-level factors like dietary and exercising habit, but also school-level factors like holding health promotion activities, providing healthy-food stores, and enough for sport facilities inside school and around school