No abstract

Accurate assessment of dietary intake among preschool-aged children is important for clinical care and research, for nutrition monitoring and evaluating nutrition interventions, and for epidemiologic research. We identified 25 studies published between January 1976 and August 2000 that evaluated the validity of food recalls (n = 12), food frequency questionnaires (n = 9), food records (n = 2), or other methods (n = 2). We identified four studies that evaluated the reproducibility of food frequency questionnaires. Validity studies varied in validation standard and study design, making comparisons between studies difficult. In general, food frequency questionnaires overestimated total energy intake and were better at ranking, than quantifying, nutrient intake. Compared with the validation standard, food recalls both overestimated and underestimated energy intake. When choosing a method to estimate diet, both purpose of the assessment and practicality of the method must be considered, in addition to the validity and reproducibility reported in the scientific literature.

Accurate assessment of infant feeding is needed for clinical practice and research. We identified 32 studies that evaluated the validity of direct observation, test weighing, or doubly labeled water methods. Correlations with validation standards were highest for doubly labeled water and test weighing, and lowest for observation. Cost and availability of isotope may limit the doubly labeled water method to research studies, whereas observation may be useful for clinical practice. Test weighing could be applied to either setting, but it may be practical to sample less frequently over 24 hours. Validity results and intended use of the measurement should be considered when selecting a method.