Continuing needs for food consumption data for public health policy

There are five broad categories of food-related public health problems in the U.S. for which survey data on food consumption are needed. These relate to reproduction, growth and development, chronic disease, food safety, food insecurity and problems specific to the elderly. The collection and analysis of food consumption data has become increasingly difficult for three major reasons. First, broad societal trends (e.g., ethnicity and use of food away from home) and changes in the food production, processing and marketing sectors have complicated the task. Second, the traditional concerns of monitoring (tracking population means and prevalences) are no longer the only objectives; there is a growing demand for data on habitual intake of individuals and variability in habitual intakes (e.g., to estimate the frequency of acute exposures). Third, data are required at several levels of aggregation (e.g., nutrients, food categories, commodities and name-brand foods) and a high frequency of non-consumption at lower levels of aggregation places high demands on sample size. It is suggested that the current large national surveys be supplemented with special purpose surveys that meet individual agency needs, and that principles and methods of state and local monitoring be further developed and implemented.