Allergenicity versus Sensitivity

Dr Hugh Sampson is professor of pediatrics and head of the Jaffe Food Allergy Institute at Mount Sinai Medical Centre in New York. He explains the difference between being ?sensitised? to a food and having an allergic reaction to it - not at all the same thing.

When assessing whether a patient is likely to suffer from an allergy (or other allergies apart from those already known about) an allergist will frequently test for foods that a patient has never knowingly eaten, or is eating without obvious problems.

They are most likely to do so with children with moderate to severe atopic eczema or asthma who do not exhibit acute symptoms after eating, and with children with a known food allergy.

The likelihood of multiple allergies

Around 35% of infants with milk or egg allergy do go on to develop other food allergies while approximately 33% of children with peanut allergy will develop allergy to at least one tree nut.

Negative tests are therefore very reassuring as they suggest that the specific food tested can be eaten safely. (Of course this does not apply in allergic disorders which are not due to IgE antibodies, such as those whose symptoms are limited to the gut.)

Positive test - but no reaction

The problem arises when a patient (adult or child) tests positive to a food that he or she is eating with no apparent problems.

Although no diagnostic test in medicine is 100% accurate, the discrepancy between a positive laboratory test for a food and the absence of clinical symptoms (eg a reaction) is especially disconcerting when dealing with food allergy. Much of this confusion stems from a misinterpretation, or ‘over-interpretation’ of what the test can actually do.

A positive skin prick test to egg or milk simply indicates that a person has IgE antibodies to milk or egg - that the patient is ‘sensitised’ to milk or egg. The positive prick skin test does not necessarily mean that the patient will experience an allergic reaction to milk or egg. Overall less than half of those with positive skin tests to a food will develop allergic symptoms if they eat that food.

The larger the reaction (the skin test wheal or raised area) the more likely it is that someone will react to the food, but no wheal size size is 100% definitive.

Other Factors

When determining whether someone is likely to react to a food, the allergist must weigh a number of factors, including a detailed history, skin prick test results and food-specific IgE antibody levels. Even after considering all these factors, the allergist may still not know whether a patient will react to a specific food, and a food challenge may be recommended.

Sensitivity to untried foods or foods which do not cause a reaction

Many parents wonder how their children can have positive skin tests or blood tests to foods that they have never eaten, since you cannot make IgE antibodies against something that your immune system has never met. Because many foods are made up of related proteins (ie botanically related, such as legumes: peanuts, peas, green beans, lentils etc) the skin or blood test may not fully discriminate between various members of food families. Consequently tests to a food botanically related to an allergen to which the patient is known to react may be positive, even though the patient will not react to that food when they eat it.


This tendency for IgE antibodies to ‘bind’ to several different related foods is called ‘Cross-reactivity’.

About 90% of peanut-allergic patients can eat all other members of the legume family (peas, beans, soya, lentils, chickpeas) without allergic symptoms, even though they may have positive skin tests to many of these foods.

About 90% of milk- or egg-allergic patients can eat beef or chicken, respectively, even though their skin tests are frequently positive to both.

Grass-pollen-allergic patients often test positive to grains such as wheat, oat and corn, but they can almost always eat those foods without problems. Certain pollen proteins are similar to food proteins - ragweed pollen proteins are similar to the proteins found in melon and bananas; birch pollen proteins to those in apple, plum, carrot and kiwi.

Consequently, ragweed pollen allergic, autumn hay fever patients may have positive tests to melon and bananas but not experience any symptoms when the food is eaten. Similarly birch pollen-allergic, spring hay fever or rhinitis sufferers may have positive tests to apple, plum, carrot or kiwi without experiencing symptoms when the food is eaten.

Other routes of sensitisation

In addition it is possible that infants become exposed to food proteins from inevitable and unsuspected places in the environment.

Researchers have suggested a number of possibilities: processed food contaminated with other foods, inhalation of food protein in vapour particles from cooking or in house dust, small amounts of food proteins passed in breast milk, residual food on parents’ or siblings’ hands contacting the skin of babies with eczema. And possibly by contact with food protein in the mother before the child is born.

As noted above, many of these exposures are inevitable, and researchers cannot agree whether some (food protein in breast-milk or the mother’s system prior to birth, for example) may be protective rather than harmful.

Other tests

IgG antibodies are made by the immune system to help protect us from infection; however, IgG antibodies also are made against the foods we eat and are found in most individuals.

Levels of IgG antibodies specific to certain foods may be higher in some individuals who have various gastrointestinal disorders, but this does not necessarily signify an allergy to a specific food.

Although certain allergy tests are very accurate at detecting and quantifying IgG (allergic) antibodies, they must be interpreted by someone who is highly skilled in the diagnosis of food allergy and who understands the clinical limitations of these tests.

A variety of other tests are used to diagnose food allergy, but at this point, they must be considered ‘unproven’ until well-controlled clinical trials demonstrate their value in identifying specific food allergies.

Over the past decade research has enabled the allergist to diagnose more accurately which patients will actually react to a food and when they are likely to have ‘outgrown’ their food allergy, but in many cases, the physician-supervised food challenge is still necessary to provide the patient with a reliable diagnosis.

Positive skin tests or blood tests to a food that someone is eating regularly with no allergic symptoms (including chronic eczema, hives or asthma) are never a good reason to eliminate the food from the diet. Only by considering the patient’s history and allergy test results can the physician decide what foods should be excluded.

Courtesy of Food Allergy News, the newsletter of the American Food Allergy and Anaphylaxis Network (FAAN). More information from (001) 800 929 4040

First published in 2006

If this article was of interest you will find many other articles on unlikely allergies and allergy connections here – and links to many relevant research studies here.

For more on the more 'mainstream' allergies check in to our 'allergy and intolerance home page' – and for ideas on alternative foods go here.

Top of page