Objective: To study the determinants and nature of dietary underreporting in a free-living population.
Design: Cross-sectional study of nutritional and behavioural characteristics.
Subjects: 1030 weight-stable subjects, 501 women and 529 men older than 15 y, included in the Fleurbaix Laventie Ville Santé study.
Measurements: Dietary intake was assessed using a 3 dy dietary record. Self assessed body weight and height were also recorded. Behavioural and socio-economic data were obtained from a questionnaire. Underreporters were defined as people with a reported ratio of energy intake to estimated basal metabolic rate lower than 1.05.
Results: Underreporting concerned 16% of the population and was significantly more frequent in obese than in non obese subjects (P < 0.001). Underreporting was significantly associated with a high socio-professional class (P < 0.05), having dieted at least once (P < 0.01) and to be in dietary restraint (P < 0.05). Furthermore, the contribution of protein to energy intake was significantly higher in underreporters than in non underreporters, independently of weight status.
Conclusions: These data underline that underreporting may bias the assessment of energy and macronutrient intake, particularly in studies on obesity and dietary restraint. Questions about weight concern, dieting and dietary restraint may be useful to identify subjects who underestimate their food intake.