Validity of self-estimated and weighed dietary data for assessment of military rations

Validity of food intake values, self-estimated by subjects and weighed by researchers, was evaluated during the assessment of a new military ration. As an addition to food intake measured by the two methods, a third set of intake values was generated by combining food intakes from the self-estimated and weighted data. This combined method is proposed as another reference for assessing validity. The new ration, Meal-Ready-to-Eat (MRE), was the sole source of food for an experimental group of 27 soldiers engaged in a 34-day field exercise. A control group of 30 soldiers ate a freshly prepared (A ration) breakfast and dinner and an MRE lunch. Agreement of mean nutrient intake values from self-estimated and weighed data was high among both experimental and control groups (average coefficients of correlation for 16 nutrients, 0.948 and 0.884, respectively). Paired t-tests yielded no significant differences between self-estimated and weighed means for nutrient intake but were significantly different from the means obtained by the combined method (p less than .0001). Mean nutrient intakes determined by both the self-estimated and weighed methods approached 90% of the intakes found by the combined method.