Which allergy tests are worth having and what a positive result actually shows — along with the tests sold directly to the public, which do not work and sometimes do harm.
What testing is for
Allergy tests do not scan for allergies. There is no test that takes a sample and returns a list of what someone is allergic to, and the tests that claim to do so are dealt with at the end of this article.
What the real tests do is check a specific suspicion. The suspicion comes first, from an account of what happened: what was eaten, touched or encountered, how long before symptoms appeared, what the symptoms were, and whether the same thing has happened again. The test then supports or weakens that account.
UK guidance is firm on the consequence. Testing should be limited to the things a convincing history already points to, and tests should not be carried out on foods a person is currently eating without any trouble. Testing broadly, in the hope of finding something, produces results that mislead.
Skin prick testing
This is the most common test, and the quickest. Drops of allergen extract are placed on the skin of the forearm — the upper back in young children — and the surface of the skin is pricked through each drop with a small lancet. It goes into the top layer of skin only. It is not an injection, despite how it is sometimes described.
Two controls are placed alongside the allergens. A negative control shows what the skin does with no allergen at all, and a positive histamine control shows that the skin is capable of reacting. Without both, a result cannot be read properly.
The result is read 10 to 15 minutes later by measuring the raised wheal at each site. A wheal 3 mm or more larger than the negative control counts as positive, with a smaller threshold sometimes used in infants. Itching and a raised bump are the intended outcome rather than a side effect — the test works by provoking a small, contained version of the reaction.
Several medicines suppress the response and make the test unreadable. Antihistamines are the main one, and short-acting antihistamines are usually stopped about three days beforehand; the clinic gives the exact instruction, which varies with the drug. Tricyclic antidepressants, and topical steroids applied to the test area, interfere too. Pregnancy, unstable asthma, and treatment with beta blockers or ACE inhibitors are all reasons for caution and should be mentioned in advance.
Blood tests for specific IgE
This measures the amount of IgE antibody in the blood directed at one particular allergen. It is worth being precise about that, because it is often described loosely: it is not a measure of total IgE, and it is not a percentage. Each allergen is requested and reported separately.
It is the test used when skin prick testing is unsuitable — where eczema or dermographism leaves no readable skin, or where antihistamines cannot be stopped — and it is frequently the only allergy test available in general practice.
It is not simply a less accurate substitute. Specific IgE testing and skin prick testing have broadly similar strengths and the same central limitation, described below.
Patch testing
Patch testing looks for something different: a delayed reaction in the skin, which does not involve IgE at all and works on a slower mechanism than the tests above.
Patches carrying a panel of substances are worn on the back, typically for 48 hours, and the skin is examined when they are removed and again a couple of days later, since the reaction takes time to appear. It is the test for contact dermatitis, and the usual culprits are nickel, fragrances, preservatives, hair dye and rubber chemicals.
Intradermal testing
Here the allergen is injected into the skin, deeper than a prick test. It is more sensitive and carries more risk of provoking a substantial reaction.
Its use is narrow: chiefly suspected drug allergy, penicillin being the common example, and insect venom. It is a hospital procedure carried out by specialists with resuscitation facilities to hand, and it is not something undertaken casually.
Food challenge
For food allergy, the definitive test is to give the food itself, in gradually increasing amounts, under medical supervision with resuscitation equipment available.
It is used to settle the question where tests and history disagree, and to establish whether an allergy has been outgrown. Because it deliberately risks a reaction, it belongs in hospital and nowhere else. The same logic makes supervised exclusion and reintroduction, usually guided by a registered dietitian, the main route to diagnosing the food allergies that do not involve IgE and therefore do not show up on the tests above.
What a positive result does not mean
A positive skin prick or specific IgE result shows sensitisation: the immune system has produced IgE against that substance. Allergy means something more — sensitisation plus actually reacting on exposure.
A great many people are sensitised to things they eat or meet regularly with no trouble whatever. UK guidance puts the positive predictive value of these tests as low as 50%, which is to say a positive result taken on its own is close to a coin toss as a predictor of genuine allergy. The negative predictive value is good, so these tests are far more useful for reassurance than for confirmation. A clear negative, set against a doubtful history, is strong evidence that the suspected allergy is not there.
There are specific traps. Sensitisation to pollens and house dust mite can produce false positives to foods including shellfish and tree nuts, because the antibodies cross-react with similar proteins.
The practical consequence matters more than the statistics. A positive result with no supporting history is not grounds for avoiding a food for life, and a long list of positive results from broad testing is a recipe for a needlessly narrow diet. Test results are read with the history, by someone who knows how to weigh the two together.
Tests sold directly to the public
A substantial market sells allergy and intolerance testing without any medical referral, and the tests on offer are largely ones that do not work. UK dietetic guidance names them and does not recommend any of them.
IgG food antibody tests are the most widely sold. They measure a different antibody from the one involved in allergy, and the reason they fail is simple: IgG antibodies to foods are present in people with no food allergy at all, and broadly reflect having eaten the food. A high result indicates exposure, not intolerance.
Hair analysis has no basis for allergy testing. Hair is genuinely analysed for other purposes, such as detecting drug use or heavy metal exposure, which lends the method a borrowed credibility it does not carry here. Kinesiology, which judges allergy by testing muscle strength, has no evidence behind it. Electrodermal or Vega testing relies on skin conductivity, which bears no established relationship to allergy or intolerance. Pulse testing treats a change in heart rate as the signal, and heart rate responds to anxiety, movement, stress and hydration. ATCAT produces results that are inconsistent and do not reproduce.
The harm is not just the fee. These tests characteristically return long lists of foods to be avoided, and acting on such a list means cutting out foods for no reason, with a real risk of nutritional deficiency where the restriction is wide or prolonged. That risk is greatest in children, who should not be put on a restricted diet except under the supervision of a paediatric dietitian.
Where a commercial report has already led to foods being dropped, a GP or a registered dietitian can work out which exclusions have any foundation and get the rest back into the diet.
How testing is arranged
The route is a GP, who can order specific IgE tests directly in many cases and refer on to an allergy clinic for skin prick testing, patch testing, drug or venom testing, or a food challenge. Referral is also the route to immunotherapy, which requires a confirmed diagnosis of the specific allergen driving symptoms.
Testing is worth pursuing where the trigger genuinely is not known, where a reaction has been severe, where a food allergy is suspected in a child, or where identifying the allergen would change what is done about it. Where the trigger is already obvious and the symptoms are the ordinary seasonal kind, the useful next step is usually treatment rather than a test.
Changed