Methodology for developing a nutrient and food pattern equivalents database for selected branded foods in the USDA National Household Food Acquisition and Purchase Survey-1

Authors: Bowman, Shanthy

Submitted to: Journal of Food Composition and Analysis
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/6/2021
Publication Date: 10/9/2021
Publication URL: https://handle.nal.usda.gov/10113/7502642
Citation: Bowman, S.A. 2021. Methodology for developing a nutrient and food pattern equivalents database for selected branded foods in the USDA National Household Food Acquisition and Purchase Survey-1. Journal of Food Composition and Analysis. 105:104567. https://doi.org/10.1016/j.jfca.2021.104167.
DOI: https://doi.org/10.1016/j.jfca.2021.104167

Interpretive Summary: This study describes the methodology used for the development of a nutrient and food pattern equivalents database for about 3,500 branded foods collected in the U.S. Department of Agriculture (USDA) National Household Food Acquisition and Purchase Survey (FoodAPS-1) that could not be directly linked to foods in the USDA nutrient and food pattern equivalents databases available then. The current study was initiated in 2017 to develop nutrients and food pattern equivalents for the 3,500 branded foods. Food label data; Information Resources Incorporated scanner data; and USDA databases Food and Nutrient Database for Dietary Studies (FNDDS) 2013-2014, Food Patterns Equivalents Databases (FPED/FPID) 2013-2014, and National Nutrient Database for Standard Reference Legacy Release (SR-Legacy) were used to develop the final database. Nutrition Facts Label serving size data were used to compute energy and food label nutrients per 100 grams. The ingredients listed in the food labels were then linked the appropriate SR-Legacy or FNDDS food codes to obtain nutrient profiles. The amounts of each of the ingredients and their Healthy Eating Index (HEI) nutrients were computed by matching calorie and nutrients per 100 grams computed from the Nutrition Facts Labels. Similarly, the 37 FPEs were computed by linking the ingredients from the previous step to the FPID or FPED, as appropriate. The HEI nutrients and FPEs for the branded foods were computed by totaling respective nutrients and FPEs of the ingredients. Calories and nutrient computed from Nutrition Facts Labels and that computed by this study methodology were in close agreement. As an example, a comparison of HEI nutrients computed from frozen enchilada and bean entrée food label and that computed using this study method were as follow: energy 119 vs. 114 kilocalories, total fat 4.4 vs. 4.4 grams, saturated fat 0.3 vs. 0.4 gram, and sodium 288 vs. 282 milligrams. The frozen enchilada and bean entrée contained 0.22 cup eq. of total vegetables, 0.4 oz. eq. of soy and legumes estimated as protein foods. The study methodology is suitable for estimating HEI nutrients and 37 FPEs for branded foods.

Technical Abstract: The study describes the methodology for developing a database that contains nutrients and 37 Food Patterns Equivalents (FPEs) needed to compute the Healthy Eating Index (HEI) for about 3,500 branded foods collected in the U.S. Department of Agriculture (USDA) National Household Food Acquisition and Purchase Survey (FoodAPS-1) that could not be directly linked to USDA nutrient and food pattern equivalents databases available in 2014. Data sources used in this study included food label data; Information Resources Incorporated scanner data; and USDA Food and Nutrient Database for Dietary Studies (FNDDS) 2013-2014, Food Patterns Equivalents Databases (FPED/FPID) 2013-2014, and National Nutrient Database for Standard Reference Legacy Release (SR-Legacy). This study illustrates the stepwise process for computing nutrient and food pattern profiles, by using frozen enchilada and bean entrée as an example. The ingredients were obtained from the ingredient list in the food labels. Nutrient profiles and amounts of each of the ingredients were computed by matching nutrients from the Nutrition Facts Labels to the USDA nutrient database foods. Similarly, by linking the ingredients to FPED or FPID, the 37 FPEs were computed. Nutrients and FPEs present in ingredients were totaled to get that for frozen enchilada and bean entrée. Calories and HEI nutrients computed by this method matched closely with that obtained from the Nutrition Facts Labels. The estimated kilocalories of 114 is close to 119 kilocalories computed from the food label. A comparison of computed HEI nutrients from the food label vs. that estimated using this methodology were: total fat 4.4 vs. 4.4 grams, saturated fat 0.3 vs. 0.4 gram, and sodium 288 vs. 282 milligrams. Although food labels do not contain food patterns equivalents, this study demonstrated that by linking label ingredients to appropriate, publicly available, SR-Legacy, FNNDS, and FPED/FPIDs, and by solving for ingredient amounts, both nutrients and FPEs can be estimated for branded foods.