Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 2/27/2007
Publication Date: 10/27/2007
Citation: Haytowitz, D.B., Lemar, L.E., Pehrsson, P.R. 2007. USDA’s Nutrient Databank system – a tool for handling data from diverse sources. 7th International Food Data Conference, Food Composition and Biodiversity, October 22-24, 2007, Sao Paulo, Brazil.
Technical Abstract: Objective: This talk describes key features of USDA’s Nutrient Databank System (NDBS), which allows processing of data from diverse sources, including USDA’s National Food and Nutrient Analysis Program, the food industry, the scientific literature, and food labels. Methods: NDL designed the NDBS as a 3-tiered data management system (Initial, Aggregation, and Compiled) with modules to facilitate the handling of the data. Raw data and documentation identifying the data source, sample description, sample handling, and analytical methods are migrated by either batch files or hand-entered into the “Initial” database. NDL scientists compare new data with old, test for outliers, and decide how to combine “Initial” data in the “Aggregation” database. Data can be grouped and weighted by a variety of parameters, including study, source, and market share. Depending on the type of data, various statistical algorithms are available to generate statistics, such as: mean, standard error, number of data points, and error bounds. At the “Compiled” step, names of food items are finalized and common measures selected. Full nutrient profiles and missing components are imputed according to scientific principles from similar foods or by using the formulation module, which uses existing data for limited nutrients and ingredient lists to estimate formulations and full nutrient profiles. A recipe module allows calculation of nutrient profiles based on recipe ingredients and proportions. The NDBS automatically documents how each value was derived. Prior to release, the completed nutrient profiles are sent for review and incorporates quality control checks at all levels. Finally, the data are disseminated. Results: The NDBS brings together a number of standalone applications into one integrated system. Data points and documentation are managed and maintained in one place. Conclusions: The NDBS permits the annual release of reliable, current data for a comprehensive suite of food components for a wide variety of foods through releases of SR on NDL’s web site: http://www.ars.usda.gov/nutrientdata. The SR also provides the nutrient data for the “What We Eat in America” component of the National Food and Nutrition Examination Survey (NHANES). Through these releases, NDL provides critical data for researchers, diet and health professionals, and consumers.