In 1994 I was diagnosed with celiac disease, which led me to create Celiac.com in 1995. I created this site for a single purpose: To help as many people as possible with celiac disease get diagnosed so they can begin to live happy, healthy gluten-free lives. Celiac.com was the first site on the Internet dedicated solely to celiac disease. In 1998 I created The Gluten-Free Mall, Your Special Diet Superstore!, and I am the co-author of the book Cereal Killers, and founder and publisher of Journal of Gluten Sensitivity.
The only plants demonstrated to have proteins that damage the small intestines of people with celiac disease are those from wheat, rye, and barley (and the man-made wheat-rye cross called triticale). Although oats had generally been considered harmful until 1996, several high quality studies published since then indicate that oats are not harmful either in celiac disease or dermatitis herpetiformis. Some physicians choose not to accept these findings or else point out that there is some potential problem of contamination of oats by wheat. The contamination question has not yet been adequately researched, but may be overemphasized. The three harmful species are members of the grass family and are quite closely related to one another according to various schemes of plant classification (taxonomy). However, not all members of the grass family damage the intestines of celiac patients. Rice and corn, for example, are apparently harmless.
Many other grains have not been subjected to controlled testing or to the same scrutiny as wheat, rye, barley, oats, rice, and corn in relation to celiac disease. In fact, only wheat and oats have been extensively studied in controlled experiments with the most up-to-date methods. If we accept corn and rice as safe, however, and this seems reasonable to me, then members of the grass family that are more closely related to these species (on the basis of taxonomy) than to wheat are likely to be safe. Such grasses include sorghum, millet, teff, ragi, and Jobs tears, which appear to be reasonably closely related to corn. In some cases, there are protein studies in support of this conclusion, although the studies are not sufficiently complete to provide more than guidance. Scientifically controlled feeding studies with celiac patients would provide a better answer. However, such studies are not likely to be carried out in the next few years because of high costs and the difficulty of obtaining patient participation (such studies would likely involve intestinal biopsy). In lieu of feeding studies, further studies of protein (and DNA) would provide the next best way to evaluate my suggestion that millet, sorghum, teff, ragi, and Jobs tears are not likely to be toxic in celiac disease, although even such studies are hampered at present by a lack of knowledge of which sequences in the wheat gluten proteins are harmful. There is evidence that a few sequences are harmful, but not all possibilities have yet been tested.
The scientific name for bread wheat is Triticum aestivum var. aestivum--the first part of the name defines the genus (Triticum) and the second part, the species (aestivum). Species falling in the genus Triticum are almost certain to be harmful to celiac patients. Grain proteins of these species include the various types characteristic of the gluten proteins found in bread wheats (including the alpha-gliadins) that cause damage to the small intestine in celiac disease. Durum wheats (Triticum turgidum var. durum) used for pasta are also harmful to celiac patients. Some Triticum species of current concern include Triticum aestivum var. spelta (common names include spelt or spelta), Triticum turgidum var. polonicum (common names include Polish wheat, and, recently, Kamut), and Triticum monococcum var. monococcum (common names include einkorn and small spelt). I recommend that celiac patients avoid grain from these species. Also, given their very close relationship to bread and durum wheats, I think it is unlikely that these grains would be safe for those with classical allergic responses to wheat.
Rye (Secale cereale) and barley (Hordeum vulgare) are toxic in celiac disease even though these two species are less closely related to bread wheat than spelta and Kamut. They belong to different genera, Secale and Hordeum, respectively, and lack alpha-gliadins, which may be an especially toxic fraction.
There have been anecdotal reports suggesting a lack of toxicity in celiac disease for spelta and Kamut, along with anecdotal reports of the opposite, at least in the case of spelt-celiac patients who have been harmed by eating it. Controlled tests would be necessary to draw a firm conclusion, although they hardly seem necessary insofar as spelt and Kamut should be considered forms of wheat.
The diagnosis, sometimes self-diagnosis, of celiac disease is occasionally made without benefit of reasonably rigorous medical or clinical tests, especially intestinal biopsy. Individuals who are diagnosed in this way without rigorous testing may not actually have celiac disease. Claims that particular foods cause this latter group no problems in relation to their celiac disease could cause confusion.
Furthermore, celiac patients who report no problems in the short run with spelt or Kamut might experience relapse later. There is now adequate evidence that when celiac patients on a gluten-free diet (that is, a diet free of any proteins or peptides from wheat, rye, and barley) have wheat reintroduced to their diets, times-to-relapse vary enormously among individuals, ranging from hours to months, or even years. And this is for wheat, presumably the most toxic of all cereal grains to celiac patients.
Additionally, the relapse may not be accompanied by obvious symptoms, but be recognized only by physicians through observation of characteristic changes in the small intestinal tissues obtained by biopsy. The reasons for the enormous variability of response times are not known. It may be speculated that the variability has something to do with the degree of recovery of the lining of the small intestine on a gluten-free diet, the degree of stress that the patient had been experiencing (including infections), and individual genetic differences.
As I have indicated, all known grain species that cause problems for celiac patients are members of the grass family. In plant taxonomy, the grass family belongs to the Plant Kingdom Subclass known as monocotyledonous plants (monocots). The only other grouping at the Subclass level is that of dicotyledonous plants (dicots). Some other species about which celiac patients have questions actually are dicots, which places them in very distant relationship to the grass family. Such species include buckwheat, amaranth, quinoa, and rape. The seed of the last plant listed, rape, is not eaten, but an oil is pressed from the seeds that is commonly used in cooking. This oil is being marketed as canola oil. Because of their very distant relationship to the grass family and to wheat, it is highly unlikely that these dicots will contain the same type of protein sequence found in wheat proteins that causes problems for celiac patients. Of course, some quirk of evolution could have given rise in these dicots to proteins with the toxic amino acid sequence found in wheat proteins. But if such concerns were carried to a logical conclusion, celiac patients would have to exclude all plant foods from their diets. For example, buckwheat and rhubarb belong to the same plant family (Polygonaceae). If buckwheat were suspect for celiac patients, should not rhubarb, its close relation, be suspect as well?
It may be in order to caution celiac patients that they may have undesirable reactions to any of these foods--reactions that are not related to celiac disease. Allergic reactions may occur to almost any protein, including proteins found in rice, but there is a great deal of individual variation in allergic reactions. Also, buckwheat, for example, has been claimed to contain a photosensitizing agent that will cause some people who have just eaten it to develop a skin rash when they are exposed to sunlight. Quinoa and amaranth may have high oxalate contents-approaching those of spinach and these oxalate levels may cause problems for some people. Such reactions should be looked for, but for most people, buckwheat, quinoa, or amaranth eaten in moderation apparently do not cause problems. (Buckwheat is sometimes found in mixture with wheat, which of course would cause a problem for celiac patients.) It seems no more necessary for all people with celiac disease to exclude buckwheat from their diets because some celiac patients react to it than it would be for all celiac patients to exclude milk from their diets because some celiac patients have a problem with milk.
In conclusion, scientific knowledge of celiac disease, including knowledge of the proteins that cause the problem, and the grains that contain these proteins, is in a continuing state of development. There is much that remains to be done. Nevertheless, steady progress has been made over the years. As far as I know, the following statements are a valid description of the state of our knowledge:
A list of my publications with pertinence to celiac disease follows. Cross-references to the literature for most of the points discussed above can be found in these publications.
Grains in Relation
to Celiac (Coeliac) Disease by Donald D. Kasarda.
An annotated copy: http://wheat.pw.usda.gov/topics/