Obtaining Nutritional Information for Dietary Assessment

Abstract

<bold>Objective: </bold>Determine the adequacy of food composition data for sweetened beverages and snack foods, and improve the nutrient database for a previously validated FFQ (the Beverage and Snack Questionnaire,BSQ). <bold>Methods:</bold> To illustrate the impact of using only USDA Standard Reference-24 to perform nutrient analyses of the BSQ, we populated two nutritional databases. The first database included only BSQ foods whose nutrient profiles were available from USDA SR-24. The second database included nutrient profiles obtained from USDA SR-24 augmented with data from other databases, food manufacturers, and scientific publications. Nutritional analyses of the BSQ were performed separately for each database, and mean estimates for energy, protein, fat, and carbohydrate were compared. We also compared the nutritional databases by selecting a single food to define the nutrient profile of each line item, and subsequently by selecting multiple foods to define line items. <bold>Results:</bold> Comparing BSQ nutritional analysis using only information from USDA-SR24 vs. analysis using this information augmented with other sources revealed minor differences (approximately 0-1%) for mean energy and macronutrients in 13 of 19 line items. However, deviations up to 100% with respect to macronutrient composition were observed for other line items including fruit drinks, flavored waters, and frozen desserts. Nutritional analysis using single vs. multiple foods to define line items was similar, although significant deviations (up to 100% were observed for various line items such as candies, cookies, and fruits. <bold>Significance:</bold> If USDA SR-24 is the sole source of food composition data, the accuracy of nutritional assessments could be compromised, potentially leading to underestimating effects of sweetened beverages and snack foods on nutritional status and energy balance. Understanding how food composition data and line item definitions impact the performance of FFQs may improve the accuracy of dietary assessment tools, and advance understanding of the effects of foods on nutritional status.