Diet interviews of subject pairs: How different persons recall eating the same foods

Objective To compare qualitative descriptions of the same food items eaten by different persons using 24-hour dietary recall interviews. Design Eleven pairs of subjects were interviewed twice using 24-hour dietar y recalls such drat each member of the pair described the same day's foods. Each hair shared a home and ate at least 2 meals together daily. After eac h interview, subjects were asked to identify the foods reported during the interview that they observed tale other member of their pair consuming and to note when a particular food was the only item of that type available in the house. Qualitative descriptions of the foods were compared, differences in descriptions were noted, and calculations were made of the potential en ergy error produced if a subject: erred in reporting a food item. Subjects/setting Subjects were randomly selected from a database of persons who have participated in other studies at the Beltsville Human Nutrition R esearch Center. Ten pairs were husbands and wives and 1 pair was sisters. E ach pair reported eating at least 2 meals per day together. Dietary recall interviews were done at the Research Center and were conducted by a trained dietitian in a quiet room free of distractions. Results Discrepancies in qualitative food descriptions were identified for every subject pair interviewed. Men were found to be more likely to omit fo od items than women, snack items were more likely to be omitted than meal i tems, meat items were likely to be described inaccurately, and first interv iews were likely to contain more errors than second interviews. Applications/conclusions Thus analysis shows which types of food items are most likely to be omitted or inaccurately described, and that dietetics pro fessionals may improve the accuracy of dietary intake interviews by asking questions related to meat, milk, and snacks very carefully. The analysis al so showed reductions in recall inconsistencies from the first recall to the second recall, suggesting that the learning associated with repeated inter views may be helpful in accurately identifying what a person consumes.