Food Allergy Testing in Newborns
Routine food allergy testing is not recommended in newborns without symptoms, as testing cannot predict future allergies and has poor positive predictive value. 1
When Testing Should NOT Be Done
Do not perform allergy testing in asymptomatic newborns or infants, even with a family history of food allergy, as the tests are sensitive but poorly specific—having allergen-specific IgE does not indicate disease 1
Avoid testing before introducing highly allergenic foods (cow's milk, egg, peanut) in standard-risk infants, as there is insufficient evidence to support this practice 1
Never test based solely on family history in a close relative, unless the infant has already experienced symptoms with a specific food 1
When Testing IS Indicated
Testing should be considered only in specific clinical scenarios:
Infants with Moderate-to-Severe Atopic Dermatitis
Test children younger than 5 years with persistent moderate-to-severe atopic dermatitis for cow's milk, egg, peanut, wheat, and soy IF the eczema persists despite optimized topical therapy (moisturizers and appropriate-potency corticosteroids) OR if there is a reliable history of immediate reaction after ingesting a specific food 1, 2
Up to 37% of children younger than 5 years with moderate-to-severe atopic dermatitis have IgE-mediated food allergy 1
Critical caveat: Optimize skin care FIRST before attributing symptoms to food allergy—many cases are misattributed when the real issue is inadequate eczema management 1, 2
High-Risk Infants Before Peanut Introduction
Infants with severe eczema (requiring topical corticosteroids or calcineurin inhibitors for at least 12 of 30 days on two occasions) should undergo evaluation by an allergist between 4-6 months, which may include skin prick testing, before introducing peanut 3
Infants with diagnosed egg allergy in the first 4-6 months should also be evaluated before peanut introduction 3
Use skin prick testing rather than serum IgE in this population, as IgE testing results in considerably higher rates of sensitization and unnecessary oral food challenges 3
Infants with Suspected Reactions
Test when an infant has experienced anaphylaxis or immediate symptoms (hives, facial swelling, vomiting, respiratory symptoms) within minutes to hours of ingesting food, especially if symptoms occur repeatedly 1
Consider testing in infants with persistent gastrointestinal symptoms such as eosinophilic esophagitis, gastritis, enteritis, enterocolitis, or allergic proctocolitis 1
Testing Methods and Interpretation
Skin prick testing (SPT) should be performed with both negative and positive histamine controls, in duplicate, with maximum wheal diameter measured 3
For peanut SPT: wheal ≤2 mm indicates low risk, 3-7 mm indicates moderate-to-high risk, ≥8 mm indicates high likelihood of allergy 3
False-positive tests are common—ensure children are clinically allergic and not just sensitized before implementing dietary restrictions 1
Prevention Strategy: Early Introduction Without Testing
The current evidence-based approach prioritizes early introduction over testing:
Introduce allergenic foods (peanut, egg, dairy, wheat) around 6 months but not before 4 months in standard-risk infants without prior testing 1, 4, 3
For high-risk infants (severe eczema or egg allergy), introduce peanut between 4-11 months after appropriate evaluation, as this reduces peanut allergy risk by up to 80% 3
Do not delay introduction of any food beyond 6 months, as this increases risk of iron deficiency anemia, inadequate nutrient intake, growth deficits, and feeding problems 4, 5
Critical Warnings
Maternal dietary restriction during pregnancy or lactation is not recommended as a prevention strategy 1
Excessively restrictive diets can cause serious harm including weight loss, poor growth, calcium deficiency, hypovitaminosis, and kwashiorkor—only implement with documented, clinically relevant food allergies under medical supervision 2
The severity of allergic reactions cannot be predicted by food-specific IgE levels or skin test wheal size, though the likelihood of reaction is directly related 1, 6