Provocation of migraine by dietary components has been clearly described in the medical literature for over 100 years. Competing immunologic and metabolic concepts of pathogenesis have been proposed. The metabolic concept inroducted by Alex Russell^1,2,3 was based upon inherited enzyme deficiencies, in some apparently increasing the sensitivity of migraineurs to vasoactive substances consumed in food. Deficient activity of monoamine- and diamine-oxidases and of phenolsulphotransferase have been described. Phenolic amines have been suggested as triggers. The failure of tyramine administered alone to provoke migraine in children has dampened enthusiasm for this hypothesis, although Russell’s concept of a specifically metabolic and X-linked genetically determined form of hyperammonemic migraine has not been refuted. The immunologic concept assumes a delayed allergic mechanism. Egger proposed a two-stage process in migraine provocation: allergic reaction to foods increases intestinal permeability to vasoactive substances derived from food or gut flora.

Marteletti and his colleagues have found evidence of altered immune activation in pediatric and adult migraine. Following food challenge their subjects demonstrate an increase in circulating immune complexes and in total and activated T-cells.

Egger et al have published the only double-flind placebo-controlled trials of food intolerance in childhood migraine, confirming specific food sensitivities in 52% of children with severe, frequent migraine. An average delay of two days between exposure and symptom supports the thesis that provocation occurs in stages. Egger, McEwen and Stolla subsequently demonstrated that children with food-induced migraine could be desensitized to their food triggers by an immunologic hyposensitization procedure. At the study'’ end, 80% of children receiving active treatment and 25% of children receiving placebo were able to resume a full normal diet without experiencing migraine attacks (p=0.001).

These studies support a role of immunologic hypersensitivity in the genesis of migraine in food-intolerant children.

INTRODUCTION

Migraine headache and food intolerance are ancient phenomena, each mentioned in the Hippocratic texts. Pediatric migraine as a distinct disorder received relatively little attention until the middle of this century when Vahlquist established strict criteria for its definition. These were paroxysmal headache separated by pain-free intervals, associated with two fo the following four features; nausea or vomiting, visual aura, positive family history of migraine, unilateral distribution of throbbing pain.⁴ In studies conducted twenty years apart in different countries, Bille⁵ and Silanpaa⁶ found the prevalence of migraine among schoolchildren to be approximately 4%, using Valquist’s criteria.

A role for dietary components in provoking attacks of migraine was first clearly described in Living’s classic monograph of 1873, which included four cases of food-induced migraine⁷. During the first half of this century, numerous reports of an association between migraine and food appeared, most attributing headache to allergy^8-13. The weak association between food-induced migraine and total IgE levels or the results of cutaneous prick tests, however, led some authors to doubt the existence of allergic headache^14-17.

BIOCHEMICAL PROVOCATION OF MIGRAINE

Over the past three decades, competition between immunologic and pharmacologic mechanisms for food-induced migraine has received considerable attention. The pharmacologic concept was initiated by Hannington in 1967, when she proposed that food-borne tyramine, not anti- genic protein, was the trigger¹⁸. In subsequent reports, Hannington and her colleagues suggested that migraineurs are sensitive to tyramine because of a deficiency of monoamine oxidase in plateletsl^19,20. The defect in monoamine oxidase proved to be transitory, however, a result rather than a cause of the migraine state²¹, and the group's attention turned to a persisting deficiency of platelet phenolsulphotransferase as the underlying biochemical defect in migraine²². Phenolsulphatransferase not only inactivates phenylethylamines²³ but also metabolizes other foodderived phenols such as the flavonoids which may act as triggers for red wine headache²⁴. Additional candidates for the chief biochemical trigger of migraine have been advanced by researchers in Sweden, Canada and Germany, based upon response to exclusion diets. These include tryptophan, the precursor of serotonin²⁵, phenylalanine, the precursor of norepinephrine (which stimulates platelet serotonin release)²⁶ and histamine (which allegedly accumulates because of a deficiency of diamine oxidase)²⁷. The notion that food chemicals provoke migraine because of enzymatic deficiency implies an inborn error of metabolism, yet very few children with migraine have been studied biochemically. During two double-blind placebo-controlled trials of tyramine feeding, Forsyth and Redmond were unable to induce migraine headache in children²⁸. A similar study in adults also yielded negative results²⁹. It seams unlikely that monoamines alone are the principal. food triggers for pediatric migraine, although Russell's concept of a specifically metabolic and X-linked genetically determined form of hyperamnionemic migraine has not been refuted. Indeed, the vindication of its X-linked transmission supports its analogy as one form of classical migraine¹.

IMMUNOLOGIC EVENTS IN THE GENESIS OF MIGRAINE

Marteletti and his colleagues have found evidence of immunologic disturbance following food challenges in patients with ostensibly food-related migraine, specifically an increase in circulating immune complexes and activated T-cells^30,31 and a decline in circulating levels of IL-4 and IL-6 accompanied by an increase in gamma-IFN and GM-CSF³². They have also demonstrated protection against precipitation of migraine attacks by oral administration of sodium cromoglycate, a stabilizer of mast cell membranes³³. Prophylactic benefits of sodium cromoglycate in adult migraine have been demonstrated by Mansfield et al in a double-blind placebo-controlled trial³⁴ and by Monro et al^35,36. Paganelli found that ingestion of allergenic foods by atopic individuals produces an increase in circulating immune complexes containing food protein, which can be attenuated by pretreatment with cromolyn sodium³⁷. Doering has proposed that failure of migraineurs to clear food-containing circulating immune complexes may precipitate an immunologically mediated headache and that susceptibility to immunecomplex phenomena cannot be detected by prick tests or IgE measurements³⁸.

Egger has attempted to weld together immunologic and pharmacologic mechanisms in migraine with his proposal that food allergic reactions cause an increase in small intestinal mucosal permeability which allows excessive absorption of vasoactive substances from the gut, derived either from food or from the endogenous flora³⁹. His theory receives indirect support from the work of Andre and of Dupont in Paris. Each has demonstrated that ingestion of food allergens by atopic children causes an increase in para-cellular permeability of the small intestine to biochemical substances such as the disaccharide lactulose, which are ordinarily not absorbed from the intestinal tract. Dupont found a weak correlation between prick test results and increased permeability in response to food challenge, but a strong clinical correlation between provocation of allergic symptoms and an increase in permeability on challenge^40,41. Andre was able to show that pre-treatment with cromolyn attenuated the permeability increase⁴² and concluded that the increase in permeability in response to food is more sensitive and specific than prick tests or RAST and by itself constitutes an accurate diagnostic test of food allergy⁴³. If food-induced eczema is considered a model for immunologic food allergy, then the inconsistent relationship between prick test or RAST results and clinical response to food challenge is found in atopic eczema as well as migraine and does not constitute evidence against an immunologic basis for migraine. In contrast the protective effect of sodium cromoglycate in both conditions suggests a pathogenetic role for gut mast cells.

OLIGOANTIGENIC DIETS FOR MIGRAINE

In 1970, McEwen and Constantinopoulos published the results of a prospective trial of diet in so-called "intrinsic" asthma⁴⁴. Three years later, Professor Soothill of the Hospital for Sick Children, Great Ormond Street, London, began investigating the role of non-atopic dietary hypersensitivity in a number of common diseases of childhood, including migraine. Soothill accepted six principles for dietary trials of nonatopic food sensitivity which had been set down by McEwen. These are:

(1) The essential baseline for further investigation is a symptom-free patient on a formal diagnostic diet.

(2) Because food intolerances are often multiple, the diagnostic diet must be limited to a small number of foods which are unlikely to provoke intolerance (oligoantigenic).

(3) Because non-atopic sensitivity often provokes prolonged and fluctuating symptoms, the diagnostic diet must be administered for sufficient time to allow remission to occur and b_~ clearly recognized, usually two to three weeks.

(4) Because the symptoms of food intolerance are often delayed, testing by dietary reintroduction of foods which have been avoided must be restricted to one new food per week, which is eaten daily during the challenge period.

(5) As the dose-response curve of food intolerance is bell-shaped, challenge with Virget foods should be done using normal quantities, not excessive quantities.