中国西部地区小学生营养与食品安全教育的整群随机试验效果评价

Effectiveness on the education and food safety programs among senior primary students in 2 counties from west China-a cluster randomized trial

目的分析营养与食品安全教育对改善4～6年级小学生知识、态度和行为的有效性。方法通过整群随机试验，在我国西部2个国家级贫困县中抽取12所小学，随机分成干预组和对照组。以学生自填的知识、态度和行为问卷进行基线调查，在其基础上根据当地情况编制营养与食品安全教学的辅助材料，对干预组全体学生实施每周0．5学时共2学期的试点教学干预，以相同的调查工具对干预组和对照组进行终期调查。采用广义线性混合效应模型拟合个体及群组水平的随机效应及处理因素的固定效应。结果基线调查共抽取4～6年级学生378人，两组学生的知识、态度和行为得分差异均无统计学意义（P〉0．05），干预期间无失访群组。终期调查共抽取478人，干预组在个体水平上的各项得分均高于对照组（P〈0．01）；群组水平上知识与行为得分的前后变化均高于对照组，但态度得分变化差异较小。混合效应模型参数显示，干预是终期调查时影响知识得分的显著因子（P=0．015）、但对态度和行为得分无明显效应（P〉0．05）。结论教学干预可提高4—6年级小学生对营养与食品安全的认知，但态度和行为的改变还需更长时间和更大样本的观察。

Objective To evaluate the effectiveness of nutrition and food safety educational programs on primary students from grade 4 to 6 in impoverished areas of Western China, under a school-based cluster-randomized trial. Methods Twelve primary schools were selected from 2 impoverished counties in West China and assigned to intervention or control groups, randomly, with 6 schools in each group. Self-rating knowledge, attitude and practice questionnaires were used at both baseline and final stages. Textbooks and supportive materials were designed according to the characteristics of baseline data. All students in the intervention group were treated with targeted nutrition and food safety lectures of 0.5 hour per week for 2 semesters. Generalized linear mixed- effects model was applied to fit the random effects on individual and clusters as well as to fit the fixed effect of the programs. Results Three hundred seventy-eight students from grade 4 to 6 were randomly selected at the baseline study and the differences of scores on knowledge, attitude and practice between the 2 groups were statistically insignificant （P〉0.05）. No cluster was lost during the trial. In the final investigation, 478 students were randomly selected at the individual level. Scores on knowledge, attitude and practice among students in the intervention group were significantly higher, when comparing to the control group （P〈0.01）. At the cluster level, more schools in the intervention group showed significant changes on knowledge and practice, yet the change in attitude was less obvious. Data from the mixed-effects model demonstrated that the program served as an influentialfactor on scores related to knowledge after the intervention （P=0.015） but did not affect the scores on related attitude or practice （P〉0.05）. Conclusion Lectures seemed to have improved the cognition of nutrition and food safety among primary students from grade 4 to 6. However, long-term observation and larger sample size were needed to evaluate the changes on attitude and practice among the students.