Database Requirements to Meet Food Grouping Needs

There is a growing emphasis on the importance of food group patterns to further our understanding of diet and disease relationships. Despite the interest and undisputable need for these types of analysis, food and nutrient databases have not been developed to provide formats and outputs that could be easily manipulated to answer food pattern questions. To conduct food group analysis, foods are generally grouped together by similar nutrient content either to obtain meaningful comparison with dietary recommendations or to examine dietary change such as in intervention studies. Methodologies can vary considerably both in terms of mode of data collection such as food frequency, food records or recalls and in terms of analysis approach. General methodological approaches from both behavioral and epidemiological perspectives have been used to group foods. Nutrient and food analysis databases or software can be useful tools for food grouping and to a large extent can dictate the approach or strategy that is used to conduct food group analysis. A brief overview of methodologies used for food group analysis at the Penn State Diet Assessment Center to analyze dietary recalls is presented as well as some data on comparisons of food group approaches using the Nutrition Data System for Research (NDS-R). The flexibility of NDS-R provides the basis for sophisticated food pattern analysis to answer a variety of research questions. Food grouping issues along with other features of NDS-R and a summary of the requirements of databases for food group analysis are discussed.