ADVERSE REACTIONS TO FOODS
AND THEIR RELEVANCE TO
PSYCHIATRIC DISORDERS

John W. Crayton, MD, Professor of Psychiatry, Loyola University Medical School and Chief, Section of Biological Psychiatry, Hines Veterans Administration Hospital, Maywood, Illinois.

Caffeine affects our nervous systems, we all know, and certain foods can affect nerve transmission as well. These and other effects of food on behavior have raised an interesting question: could adverse reactions to foods contribute to the development of mental illnesses such as schizophrenia and depression?

Wheat is a staple of modern agriculture and of the American diet. One researcher, however, has suggested that wheat ingestion in genetically predisposed individuals may aggravate schizophrenia. Epidemiologic studies reveal that schizophrenia is seldom seen in cultures in which wheat consumption is low, but the incidence rises when wheat is introduced into the culture. These epidemiologic studies, of course, could not control for important confounding variables, but they have stimulated interest in more-focused studies of schizophrenic patients, which have shown an association between wheat ingestion and the exacerbation of schizophrenic symptoms. The mechanism by which wheat aggravates schizophrenia is not known, but researchers at the National Institutes of Health have reported that peptide fragments of wheat gluten can enter the brain of experimental animals and affect specific receptor sites in the brain. Altering the activity of these receptor sites could alter behavior. On the basis of studies such as these, some clinicians have recommended a trial of a wheat-free diet for patients with schizophrenia.


Wheat is a staple of modern agriculture and of the American diet. . . . schizophrenia is seldom seen in cultures in which wheat consumption is low, but the incidence rises when wheat is introduced into the culture.


Evidence for another possible connection between schizophrenia and wheat comes from studies of the nerve supply to muscles. Increased branching of the nerve supply to muscles shows up in severe mental illnesses like schizophrenia and depression. It also occurs in neurological disorders. With Lou Gehrig’s disease (ALS) there are two groups of affected people: first, those with a very rapid downhill course but without the branching nerves; second, those who live 30 or 40 years and have an enormous number of branches. People with wheat sensitivity also have this branching, which suggests that there may be a subpopulation of psychiatric patients who are sensitive to wheat and who have nerve damage as a result of that sensitivity.

In depressed patients, anorexia is a common response but increased appetite can also occur, often with specific cravings for chocolate and carbohydrates. Carbohydrate-loading tends to increase the amount of the amino acid tryptophan entering the brain. Tryptophan is the precursor of the neurotransmitter serotonin, and since low levels of serotonin have been proposed as a mechanism mediating some forms of depression, the carbohydrate craver could be attempting to raise serotonin levels via this mechanism. Evidence that central serotonin levels can influence the proportions of carbohydrate and protein selected in the diet also supports an important relationship between serotonin metabolism and diet.

In 1984 I completed an in-hospital study of 56 subjects consisting of two populations: 17 patients who had significant behavioral problems, such as fatigue and mental confusion, which they attributed to food; and 39 controls without symptoms. The foods most commonly listed in the literature on food reactions are sugar, milk, wheat, corn, chocolate, eggs, peanuts, and beef. Of these I selected, based on the patient’s histories, the following for double-blind testing: milk, wheat, corn, chocolate, eggs, or peanuts. The hospital area used was as close to being an environmentally "safe" unit as we could easily manage. Only spring water was used, and the powdered food was put into capsules. The volunteers were tested for eight days: first on placebo for four days to wash out their gastrointestinal tracts, and then on the placebo or the foods that the symptomatic subjects said they reacted to. The subjects were tested at one hour, four hours, and eight hours with different tests: coordination, reaction times, memory, and mood assessment. With 2,000 data points for each patient, statistically significant differences showed up in reactions to both wheat and chocolate.


The foods most commonly listed in the literature on food reactions are sugar, milk, wheat, corn, chocolate, eggs, peanuts, and beef.


Is there a biological marker for "food reactors"? When an antigen, such as a particular bacterium or foreign chemical, enters the body, a part of the immune system can form a very specific antibody designed to combine with that antigen into an antigen-antibody complex. In addition, this antigen-antibody complex binds to a serum protein called "complement." The complement system consists of 12 or 15 proteins that work in cascade-like sequence to produce inflammation. People with symptoms had high levels of these immune complexes. A drop in complement levels tended to follow a food challenge, suggesting that serum complement had been "used up" in the production of immune complexes.

Complement levels were below normal in more than half of the food reactors, and there were many more abnormal values in symptomatic patients than in "normals." With two kinds of complement, all 17 reactors had low levels. The probability of that result happening by chance is less than one in a thousand.

Thus we were able to demonstrate a statistically significant symptomatic reaction to food in about two-thirds of the subjects that came in with a history of food-related problems. Perhaps with a larger "dose" of food, more subjects would have responded. We also found that there is something immunologically different about people who have behavioral reactions to foods. They have low levels of complement and higher levels of immune complexes than asymptomatic subjects. However, we do not yet have a reliable diagnostic test. We did show that reactors did not have elevated immunoglobulin E, the antibody involved in classical allergy. They do not have allergy in the classic sense. Instead, we can call it "food hypersensitivity."

What relationship does this syndrome of food sensitivity have to the more conventional or classic psychiatric diagnoses? It seems possible that a subgroup of patients diagnosed as having schizophrenia, depression, or anxiety disorders may also have a food sensitivity disorder that contributes to their condition. What is the treatment? For food reactors, avoiding particular foods is the treatment of choice. For patients with both mental difficulties and hypersensitivity, treatment of both conditions is often indicated.


Thus we were able to demonstrate a statistically significant symptomatic reaction to food in about two-thirds of the subjects that came in with a history of food-related problems. . . . We also found that there is something immunologically different about people who have behavioral reactions to foods.


The research facility at Loyola University and Hines Veterans Administration Hospital is well equipped to study the relationship between brain function and exposure to environmental substances, including food. The laboratory is equipped with computerized brain-mapping equipment to monitor brain activity in individuals exposed to foods and chemicals and also has extensive biochemical laboratories for measurement of bodily changes produced by these substances. Facilities are available to do "state of the art" immunological studies to explore the interaction of the immune system and the brain. These laboratories are available for the evaluation of people who have food and chemical sensitivities and symptoms of depression, anxiety, panic, or fatigue. In addition to clinical evaluation, selected subjects may be eligible for one of our research studies into the mechanisms of these reactions. To obtain more information, write me at P.O. Box 1011, Hines, Illinois 60141. As always, I am grateful for the continuing interest and support of NOHA.

Article from NOHA NEWS, Vol. XVII, No. 1, Winter 1992, pages 2-3.