Misreporting of energy intake in the 2007 Australian Children's Survey: identification, characteristics and impact of misreporters.

Misreporting of energy intake (EI) is a common problem in national surveys. The aim of this study was to identify misreporters using a variety of criteria, examine the impact of misreporting on the association between EI and weight status, and to define the characteristics of misreporters in the 2007 Australian Children's Survey. Data from the 2007 Australian Children's Survey which included 4800 children aged 2-16 years were used to examine the extent of misreporting based on EI, physical activity level (PAL), age, gender, height and weight status. Three options for identifying misreporters using the Goldberg cut-offs were explored as was direct comparison of EI to energy expenditure (TEE) in a subset of children. Linear regression was used to determine the impact of misreporting on the association between EI and weight status. The prevalence of under-reporting among all children varied from 5.0% to 6.7%, and over-reporting from 1.6% to 3.0% depending on the option used. Direct comparison of EI to TEE revealed similar results. Regression analysis showed that excluding misreporters provided the best model to examine cross-sectional associations between EI and BMI. Characteristics associated with under-reporting included older age, female, higher BMI, higher PAL, living in an urban location, lower parental education level and feeling unwell on the survey day. Over-reporting was more common among children with a lower BMI and lower PAL. In conclusion, misreporting of EI is present among various subgroups of the 2007 Australian Children's Survey. The impact of misreporting on the association between EI and body weight should be recognised by users of this survey.