Under- and Over-Reporting of Energy Intake in Slovenian Adolescents

Abstract

Objective

To examine under- and over-reporting of energy intake (EI) among adolescents and to compare relative food and nutrient intakes of under-reporters (UR), over-reporters (OR), and the whole population to acceptable reporters (AR).

Design

All adolescents completed food frequency questionnaires at regional health centers, and a subgroup also completed a 3-day weighed dietary protocol at home.

Setting

This study is a part of the first national representative study on dietary habits of Slovenian adolescents.

Participants

In total, 2,813 adolescents entering high school (10% of the population); participation rate was 95%.

Main Outcome Measures

Absolute EI and relative intakes of food and nutrients.

Analysis

Prodi (version 5.2 expert plus, Nutri-Science, Stuttgart, Germany, 2004) software was used to evaluate dietary intakes; t test, analysis of variance, and Mann-Whitney testing were used for differences between means, and chi-square was used for differences between proportions. Level of significance was set at P = .05.

Results

The prevalence of UR and OR was 34% and 10% among boys, and 27% and 11% among girls. Under-reporters reported lower energy percentage from carbohydrates, higher energy percentage from fats and proteins, and higher micronutrient densities than AR.

Conclusions and Implications

Under-reporting and over-reporting are widespread among Slovenian adolescents. Exclusion of UR and OR does not influence mean value of EI when assessing the diet of a group as a whole.

Introduction

Dietary surveys are an important tool in estimating food and nutrient intakes in populations. There are many different methods of dietary assessment. The food frequency questionnaire (FFQ) is considered the most appropriate method for dietary assessment in epidemiological studies,¹ as it estimates long-term habitual dietary intake. It is known that self-reported dietary assessment methods have a strong bias toward the underestimation of habitual energy intake (EI).² Many studies have shown underestimation of reported EI compared to measured energy expenditure (EE) with the doubly labeled water (DLW) technique or when calculating a specific cutoff limit for the ratio between EI and basal metabolic rate (EI:BMR).2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 The DLW technique is the gold standard for measuring EE, but it is expensive and is not applicable for large-scale surveys. Goldberg and colleagues defined minimum cutoff limits for EI, below which, when taking into consideration their energy requirements, individuals or populations are unlikely to survive.¹⁹

Adolescents are likely to under-report their EI.6, 8 Under-reporting of EI was found to be more common in subjects with higher body mass index (BMI), in adolescents as well as in adults.6, 8, 11, 12, 13 Female adolescents were found to under-report more often than male adolescents,3, 8, 11, 12, 13, 14, 15, 16 and under-reporting of EI estimated with dietary records increased with age in preadolescent and adolescent girls.7, 9

This study is part of a larger cross-sectional study of goiter prevalence and iodine supply in Slovenian adolescents entering high school. The subjects of the investigation in the cross-sectional study were anthropometric measures, iodine concentration in urine, thyroid size, thyroid volume, and dietary iodine intake.20, 21 In this study, dietary habits of Slovenian adolescents were examined. The objectives of this study were first, to identify under-reporters (UR) and over-reporters (OR) of EI and to estimate their reported relative food and nutrient intakes; second, to compare the relative intakes of UR and OR, as well as of the whole population, to the reported relative intakes of the acceptable reporters (AR); and third, to investigate whether the exclusion of UR and OR would affect the mean relative intakes of the population of Slovenian adolescents.