Reproducibility of a food menu to measure energy and macronutrient intakes in a laboratory and under real-life conditions

Abstract

Given the limitations associated with the measurement of food intake, we aimed to determine the reliability of a food menu to measure energy intake (EI) and macronutrient intake within the laboratory and under free-living conditions. A total of eight men and eight women (age 25·74 (sd5·9) years, BMI 23·7 (sd2·7) kg/m2) completed three identical in-laboratory sessions (ILS) and three out-of-laboratory sessions (OLS). During the ILS, participants hadad libitumaccess to a variety of foods, which they chose from a menu every hour, for 5 h. For the OLS, the foods were chosen from the menu at the start of the day and packed into containers to bring home. There were no significant differences in total EI (6118·6 (sd2691·2), 6678·8 (sd2371·3), 6489·5 (sd2742·9) kJ; NS) between the three ILS and three OLS (6816·0 (sd2713·2), 6553·5 (sd2364·5), 6456·4 (sd3066·8) kJ; NS). Significant intraclass correlations (ICC) for total energy (r0·77,P < 0·0001), carbohydrate (r0·81,P < 0·0001), dietary fat (r0·54,P < 0·0001) and protein (r0·81,P < 0·0001) intakes for the ILS and significant ICC for total energy (r0·85,P < 0·0001), carbohydrate (0·85,P < 0·0001), dietary fat (0·72P < 0·0001) and protein (0·80,P < 0·0001) intakes for the OLS were noted. The average within-subject CV for total EI was 18·3 (sd10·0) and 16·1 (sd10·3) % for the ILS and OLS, respectively, with a pleasantness rating for foods consumed of 124 (sd14) mm out of 150 mm (83 %). Overall, the food menu produces a relatively reliable measure of EI inside and outside the laboratory. The results also underscore the difficulties in capturing a representative image of food intake given the relatively high day-to-day variation in the amount and composition of foods consumed.