The Challenge of
Fifty years of dietary research have shown that the disease experience and risk factor status of vegetarians often differ markedly from those of others. This almost invariably seems to be in the direction of lower risk. The question of which specific nutrients or foods, either used or avoided by vegetarians, affect the risk of chronic disease has been much more difficult to answer. It will be clear to the reader that despite decades of research, what is not known still far outweighs what is known with confidence. There are many “maybes” and “probables,” and rather fewer “almost certains.”
This situation must raise questions about the adequacy of the research methods being used to define the effects of particular foods or nutrients, and indeed the methodologic challenge is quite severe. Diet is a complex matter that involves choices among hundreds or even thousands of foods. These each contribute dozens of well-known nutrients/vitamins/minerals in different combinations, and this is aside from the large number of phytochemicals that, though often poorly defined chemically, may possess important biological activity.
The complexity of diets presents analytic challenges on several fronts. First, we expect study subjects to provide accurate or at least useful information about what they eat. Yet it is clear that the ability of most people to accurately summarize what they routinely eat is limited. In addition, accurate information about the nutrient content of particular foods is not always available. The Department of Agriculture database (U.S. Department of Agriculture, 2003), for instance, is very extensive, but different varieties of certain vegetables may differ in nutrient/vitamin content. Maturity of the product and storage time can also be important. Then there are the many phytochemicals whose content in particular foods is poorly characterized. This has been brought forcefully to our attention when we