Clinical Review ABC of allergies

Food allergy

BMJ 1998; 316 doi: (Published 25 April 1998) Cite this as: BMJ 1998;316:1299

This article has a correction. Please see:

  1. Carsten Bindslev-Jensen

    The public perceives food allergy differently from doctors—especially in relation to its symptoms and prevalence. In controlled scientific studies a low prevalence of food allergy has been found in British and Dutch adults, whereas the percentage of people perceiving their illness as being food dependent is much higher. The prevalence in adults, confirmed by double blind, placebo controlled food challenge, has been estimated to be 1.4%. This is in contrast to findings in children, in whom the overall prevalence of IgE mediated food allergies is 5-7%.

    Prevalence (%) of adverse reactions to foods in adults*

    View this table:


    Adverse reactions to foods may be classified as due to either true food allergy or non-allergic food intolerance. In contrast, food aversion refers to symptoms that are often non-specific and unconfirmed by double blind, placebo controlled food challenge.

    Types of adverse reactions to foods

    • Food allergy due to IgE mediated mechanism (Coombs' classification, type I)

    • Food allergy not involving IgE, in which other immunological mechanisms are implicated (for example, type IV)

    • Non-allergic food intolerance (for example, pharmocological, metabolic, or toxic reactions to foods)

    • Food aversion (symptoms are often non-specific and unconfirmed by blinded food challenge)

    A true food allergy is a disorder in which ingestion of a small amount of food elicits an abnormal immunologically mediated clinical response. Food may cause allergic reactions by several mechanisms. The classic type I, IgE mediated reaction is the most thoroughly studied and potentially important in view of the risk of life threatening reactions in some people. Evidence is increasing, however, for an important role for delayed reactions (classic type IV mediated reactions). For example, eczema in children may be exacerbated by milk ingestion, and a small proportion of adults with severe contact dermatitis due to nickel may react to nickel in their diet.


    Common products containing preservatives

    Non-allergic food intolerance may be due to pharmacological, metabolic, and toxic causes. Pharmacological causes may provoke anaphylactoid reactions, flushing, hypotension, and urticaria. This can happen with foods with a high histamine content (for example, scombroid poisoning due to ingestion of brown oily fish (mackerel, tuna, etc) that has gone off). Tyramine in cheese or red wine may provoke or exacerbate migraine. Monosodium glutamate may provoke flushing, headache, and abdominal symptoms (the Chinese restaurant syndrome). Lactase deficiency in young children is an example of non-allergic food intolerance due to a metabolic cause, and it manifests as abdominal symptoms and chronic diarrhoea after ingestion of milk. Toxic reactions to foods may be due to contamination of food by chemicals or bacterial toxins.

    Much overlooked is the harmless, non-immunologically mediated, immediate perioral flare reaction (non-immunological contact urticaria) to, for example, benzoic acid from citrus fruits in children (especially those with atopic dermatitis). Parents and doctors may misinterpret this response in a child as an allergy and unnecessarily stop the child from eating citrus fruits. Food additives and colourings may elicit an acute flare up reaction of urticaria and, more rarely, gastrointestinal symptoms, with or without exacerbation of urticaria, asthma, or rhinitis. Additives include benzoates, salicylates, sulphites, and tartrazine and other colourings. The diagnosis of these reactions should be suspected in patients who develop symptoms on exposure to foods that contain preservatives—for example, meat pies, sausages and other preserved meats, dried fruits that contain sulphite, and many commercially tinned and bottled foods. Preservatives may also be sprayed on to salads to maintain freshness and are commonly present in alcoholic drinks and coloured fruit drinks. There are no diagnostic tests for reactions to preservatives or colourings. Diagnosis depends on suspicion and the use of elimination diets or blinded challenges with capsules containing preservatives and placebo capsules, or both of these approaches.

    “One man's meat is another man's poison”

    Symptoms and signs of adverse reactions to foods

    Patients with true IgE mediated food reactions generally identify either one or a limited number of specific foods that provoke symptoms, usually within minutes. A characteristic feature is the oral allergy syndrome—itching and swelling in the mouth and oropharynx followed, on further intake, by concomitant symptoms and signs from two or more of the following organ systems (the gastrointestinal tract, skin, and respiratory system). Life threatening reactions may include exacerbation of asthma, laryngeal oedema, and anaphylaxis with cardiovascular collapse.

    Factors suggesting classic IgE mediated food allergy

    • Specific food(s) can be identified

    • Timing of symptoms is closely associated with food intake

    • Symptoms are typical and involve more than one organ (for example, oral itching or swelling, nausea, vomiting, abdominal pain, diarrhoea, asthma, rhinitis, urticaria, angio-oedema, anaphylaxis)

    • Patient has a personal or family history of other atopic disorders

    Offending foods

    Many foods have been claimed to cause allergy, but controlled studies show that a limited number of foods are responsible for the vast majority of cases. Common allergenic foods include milk, eggs, and peanuts in children; and fish, shellfish, nuts (especially peanuts), and fruit in adults. When clinically insignificant cross reactions are excluded, most patients react clinically to a few foods only. Food allergy may also result from exposure to food in the workplace.

    Common foods provoking food allergy

    View this table:

    Occupational food allergy

    View this table:

    Patients allergic to birch pollen may have cross reactions with some foods (eg, apples, peaches)


    The importance of a careful case history cannot be overemphasised. The history, supported by diagnostic tests, should point towards a few possible offending foods or groups of foods. A diagnostic diet period is helpful. Usually, a highly restricted diet is not necessary—elimination diets based on essential amino acids are expensive and unpalatable, resulting in low compliance. A diet period of two weeks is usually sufficient, but a more prolonged period may be necessary, especially in the case of atopic dermatitis. To ensure that patients get sufficient nutrition while excluding suspected foods from their diet, the help of a clinical dietitian with experience of food allergy should be enlisted.


    Standardised food extracts are rarely available for use in skin prick testing to diagnose food allergy. However, a few food extracts have been validated in clinical trials in children and adults by using a double blind, placebo controlled food challenge as the gold standard. Foods that have been validated in this way include cod, peanuts, cows' milk, hens' eggs, shrimps, and soy beans. In many cases it is better and more convenient to use fresh fruits for skin prick testing. A drop of liquid food or a piece of solid food is placed on the forearm and pricked through (the “prick-prick” method).


    Cross reactivity between birch pollen and apple in patient with springtime hay fever and oral allergy syndrome after ingestion of apple


    Fresh fruit can be used for skin prick testing for fruit allergy

    The same reservations expressed for skin prick testing—namely, poorly standardised food allergen extracts—are also true for the various methods for determining serum concentrations of allergen-specific IgE against food. Another major problem with the newer and technically highly sensitive methods is that they detect the many clinically insignificant serological cross reactions, in which IgE raised against and directed towards epitopes on, for example, grass pollen, also binds to wheat proteins, but without any clinical significance of the finding.

    Results of allergen-specific IgE should be interpreted with caution, especially weakly positive results in patients with high serum concentrations of total IgE


    Top: Effect on symptoms of introducing restricted diet (elimination of suspected offending food) then reintroducing normal diet then returning to restricted diet in a patient who was eventually confirmed as being allergic to wheat and rye in a double blind, placebo controlled food challenge. Bottom: Effect on symptoms of a very strict elimination diet in a patient with atopic dermatitis. Although symptoms decreased initially, they had returned to normal levels by week 10 (although the diet was still being maintained). If at week 8 the patient had been given an open challenge or had returned to a normal diet, a food related exacerbation would have been suspected and false conclusions drawn. Especially in diseases with a high degree of spontaneous fluctuations in severity of symptoms a double blind, placebo controlled food challenge is mandatory, and care must be taken to avoid overinterpretation of the results

    The significance of reactions to patch testing is currently being evaluated in several centres. However, before any new test is included for routine diagnosis, it should be validated in clinical trials with a double blind, placebo controlled food challenge as the gold standard.

    Confirmation with oral food challenge

    Double blind, placebo controlled food challenge may be needed to confirm the medical history of and positive diagnostic tests for food allergy. Most published studies show that in an average of 50% of patients whose medical history plus positive skin prick test result or positive IgE result suggest food allergy, allergy can be confirmed by a double blind, placebo controlled food challenge. Using fresh foods masked in a vehicle is better than using freeze dried foods in capsules. In selected cases an open challenge (that is, not double blind or placebo controlled) may be used; if the results are negative then the patient is not allergic to the offending food, whereas a positive result should be confirmed by a further, double blind, placebo controlled food challenge. Food challenges should be conducted only by staff with specialist training and in the presence of a physician (or a paediatrician, for children aged under 16). They should be conducted cautiously, with incremental doses and with the immediate availability of adrenaline (epinephrine) and other resuscitative measures in view of the small risk of a serious allergic reaction.


    Reaction pattern during titrated double blind, placebo controlled food challenge in patient allergic to eggs. With maximum challenge (50 mg of egg), the patient reacted with a systemic reaction within two minutes of challenge. Blood pressure was maintained. The next day, exacerbation of the patient's atopic dermatitis occurred


    The only treatment for food allergy is avoidance of the offending food. Training patients to avoid a particular food often requires the help of a dietitian, clear written instructions, and advice about the labelling of foods. Many patients outgrow their clinical reactivity to a food (90% of infants allergic to milk do so by the age of 3, and half of patients who are allergic to eggs do so, but most patients allergic to peanuts or cod do not). The diagnosis should therefore be re-evaluated yearly.

    Adrenaline is life saving in cases of anaphylaxis and should be administered as early as possible. It is administered with a user friendly device (see later chapter on anaphylaxis), with careful instruction of patients and, in the case of children, their parents, schoolteachers, etc. Other antiallergy drugs, including cromoglycate and glucocorticoids, have been investigated in clinical trials with conflicting results and are generally unhelpful. Their use should be restricted to selected cases only, with specialist advice. Antihistamines are effective in relieving the symptoms of the oral allergy syndrome but may mask initial warning symptoms of a more severe reaction and should therefore not be used.

    Prophylaxis with breast feeding or with documented, hypoallergenic hydrolysates is effective against development of allergy to cows' milk and of atopic dermatitis but will not prevent later development of inhalant allergies

    Food aversion

    Symptoms that cannot be confirmed by double blind, placebo controlled food challenge may none the less be very distressing for patients and are likely to reflect a heterogeneous and largely unexplained group of disorders that include food aversion (“food fads”). Such patients may present with atypical and non-specific symptoms. Although they consider their symptoms to be food induced, they are often unable to identify specific foods or they report foods that are not typical for inducing IgE mediated allergy. Early specialist referral and exclusion of an IgE dependent mechanism (and potential for serious reactions) may be reassuring for the patient and their general practitioner.

    Diseases without proved association to food intake

    • Multiple chemical sensitivities

    • Chronic fatigue syndrome

    • Rheumatoid arthritis

    • Hyperactivity disorders

    • Depression

    • Crohn's disease

    • Serous otitis media

    The possible role of food allergy in other diseases or behavioural disorders is difficult to establish, although association is often easy to exclude on the basis of the history and the results of diagnostic tests. It is unhelpful to dismiss out of hand the possibil ity that a patient's symptoms are provoked by food. Equally it is inappropriate to interpret a clinical presentation as food allergy in the absence of any indication of an immunological disorder.

    Further reading


    Carsten Bindslev-Jensen is associate professor in the department of dermatology at Aarhus Marseilisborg Hospital, Aarhus, Denmark.

    The ABC of allergies is edited by Stephen Durham, honorary consultant physician in respiratory medicine at the Royal Brompton Hospital, London. It will be published as a book later in the year. BMJ 1998; 316: 1299-1302