Analysing the Total Diet

An example of efficient use of scarce resources is the collaboration between, or even consolidation of, food composition and food contaminant laboratories. Although in the minds of many people, food composition means nutrient composition, we have never been exclusive in this Journal. In this issue there are two papers on mercury in fish, and one on toxic elements in coffee. Food composition can and should be inclusive of all intrinsic, deliberately added, and incidental components, including environmental contaminants, additives, and bioactive non-nutrients. The classification of these food components is getting more and more difficult. Nutrient elements, like zinc and copper, can be contaminants; vitamins, like riboflavin and various tocopherols, can be food additives; and bioactive non-nutrients, like non-provitamin A carotenoids and flavonoids, have as much status as conventional nutrients, if not more. Even certain components historically classified as anti-nutrients are being hailed as beneficial for certain people at certain doses. This challenges our component classifications systems, and our analytical methodologies. But it also argues strongly for establishing food composition databases and dietary assessment software products that contain all these types of data. Currently few do. Even international organizations such as FAO have separate databases for nutrients, additives and contaminants. USDA also has independent data sets for some food components. Most national food composition databases restrict their compon- ent lists to just a few nutrients, even when data for more components (including additional nutrients) are available in the laboratories. As I write this, I am recalling all the arguments against combining data sets of nutrients, contaminants and additives. Many people will say that the strategy, from sampling plans to dissemination and use of the data, must be specific to the purpose, and the purpose of a contaminant database is di!erent from that of a nutrient database. Well, yes . . .and no. Sampling plans, laboratory instruments and reagents, qa/qc procedures, data compilation, statistical evaluations and representations, and all the proper documentation procedures, are mostly common. Contaminants are spurious in the foods supply. Yes, so we date- and location-mark, and otherwise properly document our samples and the values. Date-marking in particular is a lesson that food composition world can learn from the contaminant world. The food identification/classification system and the associated nomenclature and terminology do not need to be different. The INFOODS food component identification system (tagnames) readily accommodates additives and contaminants. The data compilation systems can be identical. We would surely have a powerful tool if a food consumption survey could serve the dual purpose of a risk/exposure assessment and a nutrient intake assessment. The economy of rationalizing resources is clear. Scientifically, however, who knows what we have missed in diet-related morbidity and mortality by not having these combined data sets in our software. The 2nd International Total Diet Study Workshop, scheduled for mid-November 2001 in Brisbane, will bring together toxicologists and food composition specialists (mainly from OCEANIAFOODS) to cover their common areas of activity. It remains to be seen how well we can meld, both theoretically and in practice. Iʹll keep you posted.