Fostering data quality in food composition databases: applications and implications for public health

Public health practitioners and policy makers are critical end-users of food composition data. Public health applications of food composition data include assessing food and nutrient availability and intakes in populations; evaluation of programs to protect and improve nutritional status; research in diet and disease interrelationships; health education and promotion activities; assessment of risk from food-borne contaminants; preserving information on traditional foods; and predicting and evaluating the effects of policy decisions that affect the food supply. All of these applications require making assumptions about the quality of data available. When users of the data begin to explore the validity of the numbers, they quickly become aware of the need for transparent and thorough documentation of quality-related variables. The cost of developing country-specific food composition data is high, and may be prohibitive for developing countries; thus data are borrowed, adapted to local realities, and added to as information is available. Attention to non-nutrient biologically active compounds in foods and continually evolving laboratory techniques are continually improving the scope and quality of available data. The needs for high-quality data are world-wide and increasingly urgent; investing in this area will have significant payoff in terms of ability to protect and improve the publicʹs health over the next decades.