Baxter Suzanne Domel, Paxton-Aiken Amy E, Royer Julie A, Hitchcock David B, Guinn Caroline H, Finney Christopher J
J Acad Nutr Diet. 2014 Sep;114(9):1404-10. doi: 10.1016/j.jand.2014.04.024. Epub 2014 Jun 25.
Although many studies have relied on parental responses concerning children's school-meal participation, few studies have evaluated parental response accuracy. We investigated misclassification of fourth-grade children's participation in school-meal programs based on parental responses relative to administrative daily records using cross-sectional study data collected for 3 school years (2004-05, 2005-06, and 2006-07) for 1,100 fourth-grade children (87% black; 52% girls) from 18 schools total in one district. Parents reported children's usual school-meal participation on paper consent forms. The district provided administrative daily records of individual children's school-meal participation. Researchers measured children's weight and height. "Usual participation" in breakfast/lunch was defined as ≥50% of days. Parental responses misclassified 16.3%, 12.8%, 19.8%, and 4.7% of children for participation in breakfast, classroom breakfast, cafeteria breakfast, and lunch, respectively. Parental responses misclassified more children for participation in cafeteria than classroom breakfast (P=0.0008); usual-participant misclassification probabilities were less than nonusual-participant misclassification probabilities for classroom breakfast, cafeteria breakfast, and lunch (P<0.0001 for each) (two-proportion z tests). Parental responses concerning children's participation were more accurate for lunch than breakfast; parents overstated breakfast participation (both classroom and cafeteria) and lunch participation. Breakfast participation misclassification was not related to body mass index (P=0.41), sex (P=0.40), age (P=0.63), or socioeconomic status (P=0.21) (multicategory logistic regression controlling for school year, breakfast location, and school). Relying on parental responses concerning children's school-meal participation may hamper researchers' abilities to detect relationships that have policy implications for the child nutrition community. The use of administrative daily records of children's school-meal participation is recommended.
尽管许多研究依赖于家长对孩子学校用餐参与情况的反馈，但很少有研究评估家长反馈的准确性。我们利用为一个学区的18所学校的1100名四年级学生（87%为黑人；52%为女生）在3个学年（2004 - 05、2005 - 06和2006 - 07）收集的横断面研究数据，调查了基于家长反馈相对于行政每日记录对四年级学生参与学校用餐计划的错误分类情况。家长在纸质同意书上报告孩子通常的学校用餐参与情况。该学区提供了每个孩子学校用餐参与情况的行政每日记录。研究人员测量了孩子的体重和身高。早餐/午餐的“通常参与”定义为天数≥50%。家长反馈分别将16.3%、12.8%、19.8%和4.7%的孩子在早餐、教室早餐、自助餐厅早餐和午餐参与情况分类错误。家长反馈对参与自助餐厅早餐的孩子分类错误比对教室早餐的孩子更多（P = 0.0008）；对于教室早餐、自助餐厅早餐和午餐，通常参与者的错误分类概率低于非通常参与者的错误分类概率（每项P < 0.0001）（双比例z检验）。家长关于孩子参与情况的反馈在午餐方面比早餐更准确；家长高估了早餐（教室和自助餐厅）和午餐的参与情况。早餐参与情况的错误分类与体重指数（P = 0.41）、性别（P = 0.40）、年龄（P = 0.63）或社会经济地位（P = 0.21）无关（控制学年、早餐地点和学校的多类别逻辑回归）。依赖家长关于孩子学校用餐参与情况的反馈可能会妨碍研究人员发现对儿童营养界具有政策意义的关系的能力。建议使用孩子学校用餐参与情况的行政每日记录。
