ACCURACY OF DIETARY RECALL USING THE USDA 5-STEP MULTIPLE PASS METHOD IN A MULTI-ETHNIC POPULATION: AN OBSERVATIONAL VALIDATION STUDY.

Authors: Conway, Joan; Ingwersen, Linda; Moshfegh, Alanna

Submitted to: Experimental Biology
Publication Type: Abstract Only
Publication Acceptance Date: 2/1/2003
Publication Date: 4/1/2003
Citation: Conway, J.M., Ingwersen, L.A., Moshfegh, A.J. 2003. Accuracy of dietary recall using the USDA 5-step multiple pass method in a multi-ethnic population: an observational validation study. Experimental Biology Meetings.

Interpretive Summary: None

Technical Abstract: Because of the need to test methods of dietary assessment for accuracy, we tested the ability of an ethnically diverse population to recall food intake. We observed food intake for one day in African-American (AA) and Caucasian-American (CA) male (M) and female (F) subjects. They selected all meals and snacks from a wide variety of foods and were permitted to eat as much or as little as desired. The USDA Multiple Pass Method for dietary recall was administered by telephone the following day. Subjects were provided with food measuring devices and a copy of the USDA Food Model Booklet. Survey Net, the companion food coding system, was used to calculate actual food consumption from foods selected and plate waste. Nutrient intake for the group as a whole was overestimated by 9-10%.Using SAS ANOVA for mixed models the differences between actual and recalled were significant (p less than 0.0001) for energy, protein, carbohydrate, and fat intakes for all groups. There were no differences in the accuracy of reporting macronutrient intakes between ethnic groups or between males and females. While some authors suggest that a precision of 10% between actual and recalled energy intake is ideal, these data suggest that under laboratory controlled conditions the USDA Multiple Pass Method for dietary recall can be used effectively to assess food intake in a multi-ethnic population.