Richmond Tracy K, Elliott Marc N, Franzini Luisa, Kawachi Ichiro, Caughy Margaret O, Gilliland M Janice, Walls Courtney E, Franklin Frank A, Lowry Richard, Banspach Stephen W, Schuster Mark A
Division of Adolescent Medicine, Boston Children's Hospital, Boston, Massachusetts, United States of America.
Statistics, RAND Corporation, Santa Monica, California, United States of America.
PLoS One. 2014 Jan 13;9(1):e83254. doi: 10.1371/journal.pone.0083254. eCollection 2014.
Little is known about the contribution of school contextual factors to individual student body mass index (BMI). We set out to determine if school characteristics/resources: (1) are associated with student BMI; (2) explain racial/ethnic disparities in student BMI; and (3) explain school-level differences in student BMI.
Using gender-stratified multi-level modeling strategies we examined the association of school characteristics/resources and individual BMI in 4,387 5(th) graders in the Healthy Passages Longitudinal Study of Adolescent Health. Additionally, we examined the association of race/ethnicity and individual BMI as well as the between-school variance in BMI before and after adding individual and school characteristics to test for attenuation.
The school-level median household income, but not physical activity or nutrition resources, was inversely associated with female BMI (β = -0.12, CI: -0.21,-0.02). Neither school demographics nor physical activity/nutrition resources were predictive of individual BMI in males. In Black females, school characteristics attenuated the association of race/ethnicity and BMI. Individual student characteristics-not school characteristics/resources-reduced the between-school variation in BMI in males by nearly one-third and eliminated it in females.
In this cohort of 5(th) graders, school SES was inversely associated with female BMI while school characteristics and resources largely explained Black/White disparities in female weight status. Between-school differences in average student weight status were largely explained by the composition of the student body not by school characteristics or programming.
关于学校环境因素对个体学生体重指数（BMI）的影响，我们了解得甚少。我们着手确定学校特征/资源是否：（1）与学生BMI相关；（2）解释学生BMI中的种族/民族差异；（3）解释学生BMI在学校层面的差异。
在青少年健康纵向研究“健康通道”中，我们使用按性别分层的多层次建模策略，研究了4387名五年级学生的学校特征/资源与个体BMI之间的关联。此外，我们还研究了种族/民族与个体BMI之间的关联，以及在加入个体和学校特征前后BMI的校际差异，以测试衰减情况。
学校层面的家庭收入中位数与女性BMI呈负相关，但体育活动或营养资源与女性BMI无关（β=-0.12，CI：-0.21，-0.02）。学校人口统计学特征以及体育活动/营养资源均不能预测男性的个体BMI。在黑人女性中，学校特征减弱了种族/民族与BMI之间的关联。个体学生特征而非学校特征/资源使男性BMI的校际差异减少了近三分之一，并消除了女性BMI的校际差异。
在这个五年级学生队列中，学校社会经济地位与女性BMI呈负相关，而学校特征和资源在很大程度上解释了女性体重状况的黑白差异。学生平均体重状况的校际差异很大程度上是由学生群体的构成决定的，而非学校特征或规划。
