Background/Objective: The objective of this study is to determine the relative validity of reported energy intake (EI) derived from multiple 24-h recalls against estimated energy expenditure (EEest). Basal metabolic rate (BMR) equations and physical activity factors were incorporated to calculate EEest. Subjects/Methods: This analysis was nested in the multidisciplinary PhysicaL Activity in the Young study with a prospective study design. Peri-urban black South African adolescents were investigated in a subsample of 131 learners (87 girls and 44 boys) from the parent study sample of 369 (211 girls and 158 boys) who had all measurements taken. Pearson correlation coefficients and Bland–Altman plots were calculated to identify the most accurate published equations to estimate BMR (P less than 0.05 statistically significant). EEest was estimated using BMR equations and estimated physical activity factors derived from Previous Day Physical Activity Recall questionnaires. After calculation of EEest, the relative validity of reported energy intake (EIrep) derived from multiple 24-h recalls was tested for three data subsets using Pearson correlation coefficients. Goldberg's formula identified cut points (CPs) for under and over reporting of EI. Results: Pearson correlation coefficients between calculated BMRs ranged from 0.97 to 0.99. Bland–Altman analyses showed acceptable agreement (two equations for each gender). One equation for each gender was used to calculate EEest. Pearson correlation coefficients between EIrep and EEest for three data sets were weak, indicating poor agreement. CPs for physical activity groups showed under reporting in 87% boys and 95% girls. Conclusion: The 24-h recalls measured at five measurements over 2 years offered poor validity between EIrep and EEest.
European Journal of Clinical Nutrition / Vol. 65, pp.910-919