Testing duplicate diet sample collection methods for measuring personal dietary exposures to chemical contaminants

Dietary ingestion may be a significant pathway of human exposure to many potentially toxic chemicals. The U.S. Environmental Protection Agency-National Human Exposure Laboratory has made the development of methods for measuring personal dietary exposures a high priority for its dietary exposure research program. Of particular interest was the testing of methods that could be applied in the general population as one component of multipathway exposure measurement studies. This paper describes a controlled pilot study that was conducted to evaluate procedures for collecting and processing duplicate diet samples. Nine adult and three child participants volunteered to provide dietary information for 28 days, and duplicate portions of all foods consumed daily for seven consecutive days. Sample collection procedures were evaluated for participant collection and segregation of solid and liquid foods, and for identification and separation of high-fat and low-fat foods. Methods for compositing and homogenizing mixed diet samples were tested. Food records and questionnaires were tested to document the collected food and to evaluate procedures for assessing dietary changes and collection bias. Participant time and monetary needs were evaluated along with the approach for training and providing support to study participants. Participants were able to collect 96% of the meals they consumed, even with 33% of the meals consumed away from home. Food consumed in social settings was the most difficult to collect, and participants were unable or unwilling to collect foods in some social settings. Noncollection of meals and food items increased after the third day of collection. Mixed diet samples were successfully homogenized, with 1%-11% mean relative standard deviations for moisture, fat, protein, and ash analysis in replicate sample aliquots. The laboratory-measured caloric content of collected foods was an average of 12% (range: -24% to 36%) lower than estimates of energy intake using a food diary and 16% lower than estimated energy expenditure values.