Abstract
OBJECTIVE: In England, standards for school meals included both foods and nutrients until 2015. School policies for packed lunches are generally food based; research is needed to determine whether these are adequate or whether a small number of nutrients would potentially improve their quality. DESIGN: From dietary data obtained using a weighed dietary assessment tool, a diet quality score (DQS) for packed lunches was calculated using the number of standards met out of twenty-one (eight foods and thirteen nutrients). Multilevel regression analysis determined the foods and nutrients contributing to variation in the DQS. SETTING: Eighty-nine primary schools across the four regions of the UK (England, Wales, Scotland and Northern Ireland). SUBJECTS: British schoolchildren (n 1294), aged 8-9 years, taking a packed lunch. RESULTS: The optimal model included all eight foods and seven of the thirteen nutrients, explaining 72 % of the variance in DQS. Folate, Fe and vitamin C, together with the eight food groups, explained 70 % of DQS variation. CONCLUSIONS: Ideally, policies for school packed lunches should include food-based standards plus recommendations based on a small number of nutrients.