Fruits and vegies in lunchboxes - accuracy of a prospective 24 hour food record for primary school children

Objective: To assess the accuracy of the lunchbox component of parent-completed food records for primary school children’s fruit and vegetable intakes. Design: We use a cross-sectional survey where parents completed 24 hour food records including a section about foods packed in their children’s lunchboxes for the next day. On the children’s arrival at school, trained observers compared actual lunchbox contents against the parentreported information. Subjects: 241 primary school children, representing a 76% response rate. Setting: A primary school in the Northern Rivers region of New South Wales, Australia. Statistical analyses: Pearsons correlations tested the agreement between the amounts of fruits and vegetables reported on the food record and observed in the lunchbox. Paired t-tests tested whether differences between the two methods were significantly different from zero. As many children had no fruits and/or vegetables in their lunchboxes, analyses were conducted for all children and for only those with some fruits and vegetables. Results: Amounts of fruits and vegetables observed in children’s lunchboxes were very similar to those reported by parents, with non-significant mean differences of less than 0.005 serves for both fruits and vegetables. Excluding the agreed zero counts, the lunchbox observations remained highly correlated with the parent-completed food records, with mean differences remaining minimal. Conclusion: Parents can accurately report fruits and vegetables packed in their children’s lunchboxes. While offering a high level of confidence in overall reported fruit intakes, further research is needed to confirm the accuracy of parent-reported overall vegetable intakes.