The identification of Key Foods for food composition research

The United States Department of Agriculture's (USDA) National Food and Nutrient Analysis Program (NFNAP) was initiated to update existing component values and to add data on new foods and components to reflect today's marketplace and needs for data. The USDA Nutrient Database for Standard Reference contains data for about 6040 foods for over 100 compounds. To develop a full nutrient profile for each food costs approximately $12 000 (six analytical samples x $2000 per sample). To determine food sampling priorities, the Nutrient Data Laboratory (NDL) has used the Key Foods approach to generate a list of 666 foods. This method utilizes existing nutrient profiles and nationally representative food consumption survey data collected by USDA in the Continuing Survey of Food Intakes by Individuals 1994-1996 (CSFII) and by The U.S. Department of Health and Human Services (USDHHS) in the National Health and Nutrition Examination Survey (NHANES). One premise of the project is that more samples will be collected and prepared for those foods which provide important amounts of nutrients of public health significance to the diet and not every sample will be analyzed for all the nutrients currently in NDL's nutrient databases. Even though the list of 666 Key Foods is much more manageable, procedures to set priorities for analysis are still needed. To accomplish this, two approaches were developed. One is based on a point system, and the other on nutrient consumption data. Based on an analysis of the two approaches, the nutrient consumption approach was chosen to be the primary method of selecting foods for analysis. This paper reports details of the two methods to modify the existing Key Foods list to determine new and specific priorities for NFNAP efforts. This program represents a comprehensive approach to collect baseline nationally representative data. Results will be used to update the USDA Nutrient Database for Standard Reference and to establish future priorities for frequency of updates. (C) 2002 Published by Elsevier Science Ltd.