As Ms. Moliver indicates, there is no single, universally accepted definition of a "radical" dietary regimen, and what is judged as "radical" will depend in part on one's frame of reference. Our working definition of a "radical" dietary regimen is as follows: 1) Dietary guidelines forbid or substantially restrict consumption of certain food groups that are staples of the U.S. diet, and 2) proponents claim that the regimen is effective in the treatment of specific diseases. By this definition, the macrobiotic regimen is "radical": It forbids consumption of a food group (dairy) that currently forms an integral part of USDA dietary guidelines and substantially restricts consumption of a second USDA food group (fruits); proponents claim this regimen can effectively treat cancer. Patients can pursue macrobiotic diets while adhering to conventional cancer therapy. However, proponents claim macrobiotics can cure cancer in patients who forgo or discontinue conventional therapy (1). As Ms. Moliver suggests, the risk that patients will refuse conventional treatments applies to any complementary and alternative medical therapy for which similar anecdotal reports of cancer "cures" have been publicized.

We agree that the diet consumed by many Americans is far from ideal and that it is a mistake for patients to rely solely on medical interventions while disregarding potentially beneficial dietary and lifestyle changes. We encourage eating a well-balanced diet, with adequate consumption of fruits and vegetables as suggested by USDA guidelines. As we noted in our article, a carefully formulated macrobiotic diet can provide adequate nutrition, and it seems reasonable, if they wish, for many patients with cancer to use macrobiotics as an adjunct to conventional therapy. However, for cachectic patients who have difficulty maintaining adequate intake of calories and nutrients, we discourage dietary regimens that substantially restrict consumption of food groups that can provide nutritional benefits.

Ms. Moliver states that it is possible to eat macrobiotically without consuming soy phytoestrogens. However, isoflavonoids (derived primarily from soy) are not the only phytoestrogens found at high levels in macrobiotic diets. An analysis of the urine of women following macrobiotic diets showed a mean concentration of lignans (phytoestrogens found in whole grains, legumes, vegetables, and seeds) that was more than 10 times the mean concentration found in women following typical western or Japanese diets (2, 3). Therefore, even with soy-free macrobiotic diets, we remain concerned about the theoretical potential for adverse effects of excessive phytoestrogen consumption on progression of breast or endometrial cancer.