Be Wary of Provocation-Neutralization Testing

Stephen Barrett, M.D.
August 29, 1997

Provocation and neutralization are used by “clinical ecologists” to diagnose and treat alleged sensitivities to foods and environmental chemicals. Clinical ecology is not a recognized medical specialty, is not advocated by standard medical textbooks, and is not a component of medical school or specialty training programs. Its proponents sometimes refer to themselves as specialists in “environmental medicine,” but this description is misleading.

During provocation, the patient is asked to report any symptoms that develop after various concentrations of suspected substances are administered under the tongue or injected into the skin. If symptoms occur, the test is considered positive and various concentrations are given until a dose is found that “neutralizes” the symptoms. Various other chemicals, hormones, food extracts, and other natural substances may be prescribed as “neutralizing” agents.

“Neutralization” superficially resembles the desensitization process used by allergists. However, allergists test and treat with substances that produce measurable allergic responses, whereas clinical ecologists base their judgments on subjective responses (what the patient reports).

Two scientific studies have demonstrated that provocation testing is not valid. Both found that patients reacted similarly to the test substances and placebo.

In 1971, two researchers reported on tests performed by five experienced clinical ecologists. Each of the patients had tested positive during provocation testing with special preservative-free extracts of food or alcohol, the contents of which were known to the clinical ecologist. During the experiment, the clinical ecologist was handed either the extract or a dilute salt-water solution (saline), the contents of which were known only to another physician who observed but did not participate in the procedure. Based on the patient’s reactions, the clinical ecologists were then asked to judge whether the administered material was the extract or the placebo. The extracts were correctly identified in 24 of 34 trials (70.6%). However, the salt-water solution relieved the patient’s symptoms in 28 of 40 trials (70%), indicating that symptom relief was not related to any allergy-causing substance in the extracts [1].

In the early 1980s, researchers at the University of California (UC) observed similar test results in a study funded by the Society for Clinical Ecology and the American Academy of Otolaryngic Allergy (another proponent group). The tests took place in the offices of seven clinical ecologists who had been treating the patients. During three-hour sessions, the patients received three injections of suspected food extracts and nine of normal saline. Sixteen patients were tested once, and two were tested twice. In nonblinded tests, these patients had consistently reported symptoms when exposed to food extracts and no symptoms when given saline injections. Under double-blind conditions, however, they developed symptoms with 16 (27%) the food-extract injections and 44 (24%) of the salt-water injections. The symptoms elicited by both types of injections were identical and included itching of the nose, watery or burning eyes, plugged ears, a feeling of fullness in the ears, ringing ears, dry mouth, scratchy throat, an odd taste in the mouth, tiredness, headache, nausea, dizziness, abdominal discomfort, tingling of the face or scalp, tightness or pressure in the head, disorientation, difficulty breathing, depression, chills, coughing, nervousness, intestinal gas or rumbling, and aching legs. The results clearly demonstrated that the patients’ symptoms were placebo reactions. The study also tested the claim that “neutralizing” doses of offending allergens can relieve the patient’s symptoms. All seven patients who were “treated” during the experiment had equivalent responses to extracts and saline.

  1. Kailin EW, Collier R. ‘Relieving” therapy for antigen exposure. JAMA 217:78, 1971.
  2. Jewitt DL and others. A double-blind study of symptom provocation to determine food sensitivity. New England Journal of Medicine 323:429-433, 1990.

This article was posted on August 29, 1997.