Longitudinal changes in the accuracy of reported energy intake in girls 10-15 y of age

Background: Dietary records are often used to estimate individual energy needs and population energy requirements. However, significant underreporting of total energy intake (EI) has been found when EI is compared with total energy expenditure (EE) measured by doubly labeled water. Objective: This study aimed to determine whether the accuracy of reported EI decreases from middle childhood to adolescence. Design: In this longitudinal study of 26 healthy girls, El and EE were measured at ages 10, 12, and 15 y. Accuracy of reported EI (EI/EE x 100%) was calculated at each age. At study entry, girls had a mean (+/- SD) body mass index (in kg/m(2)) of 16.8 +/- 1.9 and percentage body fat of 24.0 +/- 4.6%. Measurements of EI were a 7-d dietary record and those of EE were by doubly labeled water. Results: As they got older, girls tended to report EI less accurately: the average accuracy was 88 +/- 13% at age 10 y, 77 +/- 21% at age 12 y, and 68 +/- 17% at age 15 y. The declines in reporting accuracy from age 10 y to age 12 y and from age 10 y to age 15 y were statistically significant (P = 0.03 and P = 0.001, respectively). Reporting accuracy also declined from age 12 to age 15 y but not significantly. When percentage body fat was added to the model, results were essentially unchanged. Conclusion: Because of the decline in EI reporting accuracy with age, the use of EI data obtained from dietary records in adolescent girls will result in substantial underestimation of energy needs.