CHAPTER 13

Food Allergy

Up to this point, this book has been concerned primarily with the diagnosis and treatment of inhalant allergy (sensitivity to substances inhaled). But there are other types of allergy: ingestant allergy which is sensitivity to substances ingested, and contact allergy, which is sensitivity to items that contact the skin or mucous membranes. These terms are bandied about so freely by the allergist that it is easy to forget that the patient frequently does not understand the distinction and hesitates to ask for fear of appearing stupid. Most aspiring allergists start their program by treating only inhalant allergy, and in many cases only allergic rhinitis, as lower respiratory tract disease tends to involve additional problems other than allergy. Starting with the management of inhalant allergic rhinitis is a realistic approach. The success quotient for those treating this condition is high, and a series of successes both boosts confidence and helps in establishing a reputation for competence. It will not be too long, however, before a situation arises that simply does not fall within the parameters of allergic rhinitis caused by inhalants. It may be that all the symptoms of allergic rhinitis are present, but there appears to be no seasonal or exposure pattern. Alternatively, all the symptoms of allergic rhinitis are present but all the test results are negative. There are several possible explanations for this situation (discussed in Chapter 12), but one possible condition that should immediately be considered is food sensitivity.

For the rhinitis sufferer with negative inhalant test findings, the usual reaction to suggesting the possibility of food sensitivity is, “I’m not allergic to any foods!” This may be translated as, “I don’t have any gastrointestinal symptoms.” It may take a bit of convincing before the patient accepts the fact that an offender may produce symptoms in body locations other than that by which it entered the body. One explanation that is frequently accepted by patients is to remind them that the antihistamine-decongestant tablet taken to relieve nasal symptoms was ingested, not inserted into the nose, yet the result did not surprise the patient at all. In other words, the route of entry frequently has little to do with the response.

The first source of confusion is that there is not even a universally accepted definition of food allergy.¹ To the layperson, and even to the nonallergist physician, food allergy is usually defined as an adverse reaction to a food, in a specific patient, in the absence of the same reaction in other people who consume that food. To the allergist, this definition is too simple. The definition of food allergy depends on whether an immune mechanism produces the reaction, and even this definition is not uniform throughout the medical world.

Food sensitivity is known to involve all parts of the immune system, as well as mechanisms entirely outside the immune system. Among most general allergists in the United States, allergy (of whatever type) is defined as an adverse reaction mediated by immunoglobulin E (IgE). This is a Gell and Coombs type I reaction. A type I, anaphylactic, reaction is the only form of allergy easily capable of producing fatal results. These reactions represent only a small percentage of adverse reactions to food, but because of their sudden and severe nature, they have a high degree of visibility. Although type I reactions are easily identified by skin or in vitro testing, such tests are rarely needed, for these reactions are prompt, frequently violent episodes, that are usually quite obvious in their clinical pattern of cause and effect. Because of lifelong persistence, IgE-mediated food allergies are often referred to as fixed’food allergies. Although the patient’s sensitivity to a provoking food may diminish if avoidance is practiced for a period of years, it probably is never completely lost, and is easily reinvigorated by future exposure.

Fixed, type I food sensitivity has been estimated to represent between 5 and 20% of all food hypersensitivity. This broad range of estimates is a result of limitations that exist in recognizing other types of food hypersensitivity. In fact, the recognition of mechanisms of food allergy other than via IgE has been very slow to evolve. Breneman wrote, “For 30 years we have had a fixation on IgE as the answer to all food allergy questions. Why? Because we had two good tests for it (IgE RAST and skin tests).” Yet, “IgE explains only a small part of food allergy. Involvement of the entire immune system is evident if the more prevalent delayed-type food allergy is to be explained.”²

In the United States, allergic food reactions other than the anaphylactic type are usually designated hypersensitivity, and reactions outside the immune system are simply adverse reactions. To complicate matters further, in Europe, adverse reactions involving any or all Gell and Coombs categories are usually considered allergy, whereas specific adverse reactions outside the immune system are considered hypersensitivity. These definitions are by no means universally adhered to, however, and rarely does a contributor to the literature bother to define the terms being used when writing an article. This further compounds the difficulty of interpreting the significance of any study. In 1994, at a major pediatric conference, three articles were presented lamenting the lack of uniformity in defining food sensitivity and proposing carefully thought out definition formats. Unfortunately, no two of the articles were in agreement in their choice of terminology.³ Thus, it is not difficult to see why confusion reigns concerning the subject of food allergy, when the parameters of the problem have not even been adequately and consistently defined.

Because this book is designed for clinicians, and especially for physicians newly adding allergy to their practice, this chapter equates the terms for adverse reactions to foods: allergy, sensitivity, and hypersensitivity. These terms are taken to mean an abnormal reaction to a food, observed in one person, but not seen in the general population. The bottom line, after all, is relieving the patient’s symptoms. As will be evident in the portion on treating food reactions, this involves dietary manipulation, which serves to treat both immunologic and nonimmunologic reactions alike.

Food may act on the body through any of the four immunologic mechanisms defined by Gell and Coombs (which are explained in detail in Chapter 2).⁴ In addition to the type I reaction already described, types II, III, and IV have been demonstrated to occur in food allergy In a study of 54 infants with eczema and positive double-blind cow’s milk challenge results, and by comparing multiple simultaneous methods of testing, Isolauri’s group⁵ found that only 15% were type I, whereas 26% were type IV, 35% were either type II or type III, and 24% were of mixed types. There have been several other recent reports demonstrating similar high prevalence of non-IgE-mediated food allergies.^6–11 Type III reactions have been theorized to be the most common immune mechanism in food allergy, though to date few studies have been able to give precise values to the degree to which each Gell and Coombs type occurs. Type IV (cell-mediated) reactions may also be fairly common in food allergy, but the delay of hours to days in appearance of symptoms makes clinical correlation extremely difficult. For clinical use, accurately identifying the type of reaction is not practical, because it can only be done by subjecting patients to multiple forms of testing for the same foods. The underlying mechanism of the immunologic reactions is shown in Table 13-1.

FIXED AND CYCLIC CLINICAL FOOD ALLERGY CATEGORIES

Although each type of immune reaction may involve a different route by which a food may produce an allergic reaction in a patient, these reactions have been clinically divided into fixed and cyclic types. Fixed food reactions are defined as those that always occur when the offending food is ingested in any quantity (even minute amounts). These reactions are often rapid in onset and may be severe. Some cases of delayed symptom onset, producing chronic, rather than acute, illness, have also been ascribed to IgE-mediated reactions. These late-onset type I reactions may involve a prominent late-phase reaction, such as is seen in asthma, but are still incompletely understood. As previously noted, all IgE-mediated sensitivity is normally sustained throughout life, although it may weaken somewhat after several years if there has been no exposure. Fixed reactions are now considered synonymous with IgE-mediated type I reactions.

An interesting, and very common, clinical aspect of cyclic food sensitivity is the development of masking. Many patients with cyclic food allergies develop a tendency to eat the offending food at every meal, and frequently between meals. Like drug addicts, they “crave” the food because a regular dose of the offending food temporarily relieves some of their symptoms. This temporary improvement is offset by the fact that, overall, the patient’s condition is worsened by the offending food, and if the food is withdrawn from the patient’s diet, a considerable improvement ensues. This does eventually occur, but in the early stages of withdrawal, the patient frequently complains of an increase in symptoms and must be encouraged to persevere until relief is noted (generally after 4 to 7 days). A food can be strongly suspected of causing cyclic allergy with masking when the patient s immediate reaction to a discussion of food sensitivity is, “Don’t take away my chocolate!” (or whatever food it may be). Masking is also suspect when patients always eat just before sleep, or habitually awaken for a snack.

In addition to provoking the immunologic reactions already described, food is able to affect the body through a variety of pathways not involving the immune system. These reactions may be all but impossible to distinguish, on the basis of symptoms alone, from immunologic reactions. Although many chemical mediators may be involved in various adverse food reactions, the most frequent mediator is histamine, which is contained in mast cells throughout the body. Histamine can be released by both immunologic and nonimmunologic reactions. For example, when certain foods (e.g., strawberries and tomatoes) are ingested, they induce the release of histamine without involving the immune system. Other foods, such as aged cheese or spoiled fish, contain preformed histamine that is released on ingestion. Table 13-2 lists some of the foods that may induce the release of histamine and related substances on ingestion. Thus food, unlike inhalant allergens, is capable of affecting the body through a wide variety of mechanisms, many of which produce very similar reactions.

In addition to the histamine-triggered induction of adverse reactions to foods described previously, certain enzyme deficiencies may make foods incompletely digestible, and food-borne toxins may cause gastrointestinal injury, resulting in symptoms that may be virtually indistinguishable from those of food allergy. Lactase deficiency is a well-known example. These are also noted in Table 13-2.¹²

COMPLICATING FACTORS

Cumulative Reactions and Cross-Reactivity

Like inhalants, ingestants (foods) contain a large variety of biologic antigens, any number of which may be sensitizing. Different foods may contain several such antigens in common. As with inhalants, the number of such antigens shared by different foods determines the degree to which these foods immunologically cross-react. In the case of botanical foods (i.e., grains, fruits, vegetables, and foods prepared from them) cross-reactivity is common, although frequently unrecognized by either patient or physician. For those without extensive botanical knowledge, specific references may be helpful. Appendix 1 contains a cross-reactivity list providing information that is both useful and interesting. For example, how many laypersons would recognize potato and eggplant as close relatives? This cross-reactivity among foods may easily result in an accumulation of similar antigens in the system from several different foods to which the patient is sensitive. The resultant unrecognized increase in the total load of allergens affects the pattern of cyclic allergy, which is affected by both dose and frequency of ingestion. Patients may think that they are eating only a limited amount of a particular food, or eating it at infrequent intervals, but if they are in fact eating other cross-reacting foods of the same food family, the result may be the same as if large quantities of the primary offending food are eaten regularly. It is the total amount of the reactive allergen in the system at one time, and/or the frequency with which the body is exposed to the allergen, that determines the intensity of symptoms produced by cyclic food allergy. To determine the eating habits of the patient, not only must individual consumed foods be identified, but cross-reacting foods must also be considered, and the total quantity of each cross-reacting food family group should be estimated.

In the case of nonbotanical foods (meat from mammals, fish, birds, and other animals), studies of cross-reactivity are more limited. In this food group, IgE-mediated food allergies, mainly those caused by crustaceans and mollusks, are well known. As previously noted, IgE-mediated food sensitivity is a relative rarity and usually presents no difficulty in diagnosis, as reactions are usually prompt and noticeable. Cyclic food responses are more subtle, and hence more difficult to recognize. The non-IgE-mediated antigenic reactions involving meats have not been extensively studied, and hence knowledge of cross-reactivity within this group is limited. When considering cross-reactivity in the nonbotanical food group, it is largely necessary to use clinical ingestion trials, investigating only nonanaphylactic foods that are being eaten.

Appendix 5 lists some of the common staple animal foods in the typical American diet. It must be stressed that this represents only a starting point from which to explore cross-reactivity However, it is a reasonable approach, as opposed to the “substitute” foods suggested in some of the commercial literature, so that persons allergic to common meats do not have to try locating whale or hippopotamus! When cross-reactivity between meats other than those listed is to be considered, a reasonable approach is to use the same principle already applied to the foods previously described. One can check the scientific name of the animal in question in an encyclopedia or on the Internet, at http://animaldiversity.ummz.umich.edu/index.html If the genus of the two foods being compared is the same, there is probably significant cross-reactivity. If the genera are different, there is less chance that they cross-react. It must be borne in mind that this is simply a guide. There may at times be some cross-reactivity between even very distantly related organisms, and people are unique in their allergic sensitivity. However, consideration of cross-reactivity based on the scientific taxonomy of a food provides a very useful starting point.

Concomitant Food Reactions

In addition to cross-reactivity between various types of botanical food, cross-reactivity frequently exists between inhaled pollens and ingested foods of the same botanical family. Thus, the ingestion by a patient with active inhalant allergy of a cross-reacting food may produce a greater reaction than expected. This is called a concomitant food reaction. Such clinically observed cross-reactivities between inhalants and foods have usually been confirmed by radioallergosorbent test (RAST) inhibition studies (Fig. 13-1), which use a precise in vitro competition technique for confirming the presence of cross-reacting antigens.

The classic example of inhalant and food cross-reactivity is that between grasses and cereal grains (which are, in reality, grasses). The phenomenon has also been demonstrated between such apparently diverse items as apple (the fruit) and birch (the pollen), as well as between ragweed and members of the gourd family (watermelon, cantaloupe). A more extensive list of cross-reacting inhalants and foods is found in Appendix 1. The clinical significance of concomitant reactions is that all allergens, although they enter the body by different routes, contribute to the total allergic load. For example, a person mildly sensitive to both apple and birch pollen might have no reaction on eating an apple in the fall, but when birch trees are pollinating in spring, eating the same amount of apple might precipitate a severe reaction. This type of reaction is one more complexity to confuse and confound the unwary allergist and the allergic patient.

As if the previous problems were not enough, simply identifying a basic food to which a patient is allergic does not necessarily clarify all the possible ramifications of the situation. With inhalants, the offending substance is delivered to the nasal mucosa in its natural, unaltered state. Although some foods are usually eaten raw, more commonly foods are processed in some manner before being eaten. This may be by cooking, fermenting, marinating, or any number of methods designed to prolong the storage or enhance the palatability of the food. All such procedures are capable of altering the antigens present in the food, and the degree and pattern of such alteration are not predictable. Although some of the antigens of the basic food remain, others are changed or destroyed. In addition, new antigens may be created during processing. For example, a patient sensitive to cow’s milk would be expected to also show sensitivity when eating cheese, but the sensitivity might be less than that displayed when whole cow’s milk is ingested. Also, other allergens may have been created by alteration of the antigens of cow’s milk during processing into cheese, so that some patients may react to cheese, but not to milk. This alteration of food antigens by preparation is one of the factors that has limited the effectiveness of testing for food allergies using “pure” food material prepared by reference laboratories. The material used in testing does not necessarily represent the material to which the patient is actually exposed.

It has already been stressed that the portal of entry for an allergen may have little effect on the site or type of eventual reaction. Although it is true that inhalant allergens primarily affect the respiratory tree, even this is not always the case. Food allergens show no such limitations in their range of action. Before the allergic reaction can occur, the foods must be digested, and the digested antigens absorbed into the circulation and carried to the target organ. This much is not difficult to understand, but why allergens circulating throughout the body should affect certain organs and not others is still not understood. Virtually any organ or organ system can serve as a target organ, and the organ is not specific to the food. In other words, cow’s milk may produce a rash in one person and asthma in another. The target organ in the first person is the skin, and in the second it is the lungs. The reaction is specific to the patient, and if milk causes asthma in one patient, milk will normally continue to cause asthma in the same patient. It will not switch to causing a rash. Of course, there may be more than one target organ from the beginning. Milk, for example, might cause both asthma and diarrhea as soon as the patient becomes sensitized. If this is the case, the symptom pattern may fluctuate in severity over time, varying with the amount of milk consumed and the interval between exposures, but rarely will the pattern change. Therefore, new symptoms developing in another target organ usually indicate sensitivity to a new, and different food.

Virtually any organ system in the body may be a target organ for adverse reactions to food. The symptoms depend on the nature of the target organ, not on the type of food, and are not greatly influenced by the type of immunologic or nonimmunologic mechanism causing the reaction. As described in Chapter 2, and reiterated in the portion of this chapter on the nature of food reactions, most food reactions involve the release of histamine, or other similar mediators, into the tissues of the target organ, which in turn, produce tissue edema and inflammation. It can be quite difficult to determine the specific mechanism involved, except for the clinical difference between fixed and cyclic reactions. If the target organ is the external carotid artery complex, for example, a migraine headache will result. If the target organ is the intestinal tract, cramping and diarrhea may be expected. The reactions within each tissue at a cellular level are very similar, but each organ responds differently to inflammatory injury.

The “Leaky Gut Syndrome”

Yet another factor may even further complicate the picture of food allergy. Some patients report being “allergic to everything.” However, clinical studies on food allergy, worldwide, have been in agreement that rarely are patients primarily allergic to more than five or six foods. These “universally allergic” patients, however, appear to react adversely to an immense range of foods, typically to almost every food that is tested. Frequently, such reactions have not been a lifelong problem but have developed within a short period of time. This apparent contradiction in patterns, which is not uncommon, can be easily explained.

The mature intestinal tract, functioning properly, is a sophisticated immunologic filter. Food substances are not absorbed from the intestine into the circulation until digestion has broken them down into such small macromolecules that the immune system tends to consider them as building blocks rather than potential offenders. The intestinal tract, however, is easily shocked, and when such shocks occur, the gut wall becomes much more permeable and absorbs larger macromolecules of food, which then produce sensitization, and true allergic reactions. Many factors can shock the intestinal tract to this degree. They include infection or parasites, immaturity, and immunologic insults, including food allergy.¹²

Infection is a complex subject, but it is necessary only that the allergist realize that infection can give rise to increased gut permeability. In cases in which almost all food allergy tests are positive, some evaluation to identify fungal, bacterial, or parasitic causes should be considered, especially if the history suggests potential reasons for one of these to develop.

The infant has not yet developed a mature intestinal tract, and as a result, varying degrees of leakage of macromolecules are common. This is the reason why a small child frequently strongly manifests allergic clinical problems, although skin testing and/or in vitro tests indicate little or no response. Skin testing and/or in vitro testing is reliable for IgE-mediated allergy, but major amounts of IgE usually have not developed in infants, even if they have the genetic makeup to allow such an abnormality to appear. Prolonged or frequent exposure to allergens is required for significant amounts of specific IgE to develop. As stated elsewhere, it is often of little value to test a child for pollen allergy before the age of 5 or 6 years, and even sensitivity to perennial allergens (e.g., dust, mold, and pet epidermals) is usually minimal before the age of 2 or 3 years. On the other hand, the intestinal tract at birth, and for several months thereafter, is “porous.” The allergic child with gut immaturity may be expected to demonstrate a wide range of food allergy symptoms, including nasal and eustachian tube congestion, in addition to eczema and gastrointestinal complaints. It has been frequently observed that the vast majority of recurrent otitis media cases, appearing before age 1, can be controlled simply by eliminating cow’s milk from the diet. For the child with major allergy symptoms of any type, analyzing the diet and replacing the major staples (in the infant, usually the formula) with another basic food type frequently reduces or eliminates the problem. It is reassuring to the parents to note that most such early food allergy problems eventually resolve as the intestinal tract matures, the child is treated with a rotation diet, and permanent damage to the affected target organ is prevented by appropriate medical and surgical therapy.