Parents were told to delay allergy foods, and the research said the opposite

For years, parents were told to keep the risky foods out. Peanut. Egg. Shellfish. Wait, and wait longer if allergies ran in the family.

Then a trial funded by the National Institute of Allergy and Infectious Diseases tested that advice head on, and it did not survive. The Learning Early About Peanut Allergy study, published in 2015, gave peanut-containing food to infants at high risk of peanut allergy. NIAID’s own summary of the result says the approach was safe and “led to an 81 percent relative reduction” in peanut allergy. In January 2017, an expert panel NIAID convened published new guidance built on that finding.

If you’re reading this with a six-month-old in the next room and a jar of peanut butter in the cabinet, stop before you open it. The part that gets lost in the headlines is that the new guidance isn’t one instruction. It’s three different paths, and one of them starts with a blood test or a skin test rather than with food.

What follows is what the research showed, where it’s strong and where it’s thin, how the guidance is structured, and what anaphylaxis looks like. That last part you should know cold, before you ever need it.

What the peanut trial found, and what 81 percent means

The study enrolled 640 infants between 4 and 11 months of age, every one of them at high risk because they had severe eczema, egg allergy, or both. Researchers randomly assigned them either to eat peanut-containing food regularly or to avoid peanut entirely until age 5. Results appeared in the New England Journal of Medicine in 2015.

Among the 530 infants who started with a negative skin-prick test, 13.7% of the avoidance group had peanut allergy at 60 months, compared with 1.9% of the group that ate peanut. Among the 98 infants who already had a positive skin test at the start, the figures were 35.3% and 10.6%. There was no significant difference between groups in serious adverse events.

That is a large effect, and it’s worth being precise about how to read it. An 81 percent relative reduction describes the proportional drop. The absolute difference in the larger group was roughly 12 children out of every 100 who did not develop peanut allergy because of early introduction. Both numbers are true. The second one is the one a parent can picture.

A follow-up asked whether the protection was fragile. Researchers instructed all participants to avoid peanut for 12 months after the main trial ended. At 72 months, peanut allergy stood at 18.6% in the original avoidance group and 4.8% in the original consumption group, with no significant rise among the consumers during their year off. Early exposure didn’t just delay allergy. It changed the outcome.

A systematic review and meta-analysis published in JAMA in 2016 pooled the broader evidence. Across five trials covering 1,915 participants, early egg introduction at 4 to 6 months was associated with a lower risk of egg allergy, graded as moderate-certainty evidence. Across two trials covering 1,550 participants, early peanut introduction at 4 to 11 months was associated with lower peanut allergy risk, also moderate certainty. In absolute terms, the authors estimated about 18 fewer peanut allergy cases per 1,000 children in a population where 2.5% develop it. They downgraded certainty for imprecision and because the populations and interventions studied didn’t map perfectly onto everyone.

The 2017 guidance is a clinical pathway, not a recipe

This is where careful reading matters more than enthusiasm. The NIAID-sponsored expert panel published three addendum guidelines in January 2017, sorted by how much risk a child carries. They are not interchangeable.

Group one: severe eczema, egg allergy, or both

The panel defines severe eczema precisely. It means persistent or frequently recurring eczema that a health care provider judges severe, and that keeps requiring prescription-strength topical steroids or other anti-inflammatory treatment despite proper use of moisturizers. That’s a clinical judgment, not a parent’s impression of a dry patch.

For these infants, the panel says peanut-specific IgE blood testing, skin prick testing, or both should be strongly considered before any peanut is given. Referral to a specialist is the alternative. Where skin testing isn’t routinely available, the blood test may be the sensible first step.

What happens next depends on the result, and the pathway branches. A very low blood level or a tiny skin-test wheal points toward introducing peanut soon, either at home or under supervision in the office. A higher blood level sends the family to a specialist. A mid-range wheal calls for supervised feeding or a graded oral food challenge in a specialist setting. A large wheal means the child stays under specialist management. When testing clears the way, introduction can begin as early as 4 to 6 months, after other solid foods are already going well.

That decision belongs to your child’s clinician. This article does not print the amounts, the preparations or the feeding schedule, and you should not reconstruct them from a forum post or a product label. The guidance contains specific quantities and forms. Those numbers only mean something next to your child’s test results, and the first step for this group is an appointment.

Group two: mild-to-moderate eczema

Around 6 months, after other solid foods have been introduced, in keeping with a family’s own preferences and food traditions. The panel does not require testing for this group. Home introduction is appropriate, though some families prefer a supervised feeding in the office, and that option stays open.

Group three: no eczema and no food allergy

Age-appropriate peanut-containing foods, introduced freely alongside other solids, on the family’s own terms. No testing.

How confident is the panel, really

The guidelines distinguish strong recommendations from weak ones in their own language. The word recommends signals a strong recommendation, and suggests signals a weak one. Group one’s guidance is mostly in the strong category, resting on moderate-quality evidence. Groups two and three sit in the weak category, resting on low-quality evidence.

That’s an unusual shape, and it’s worth naming. The children with the most at stake have the best evidence behind their pathway. The two lower-risk groups rest on weaker ground, which is exactly why the panel gives those families latitude to follow their own preferences.

The consumer-facing guidance has caught up. CDC now tells parents to introduce potentially allergenic foods when they introduce other foods, naming eggs, shellfish, nut butters and wheat. CDC also says that if a child has severe eczema or an egg allergy, parents should ask their doctor or nurse about when and how to safely introduce peanut. HealthyChildren.org, the AAP’s parent-facing site, says peanut allergy testing is recommended for babies with severe eczema or egg allergy, and that there’s no evidence waiting beyond 4 to 6 months prevents food allergy.

One practical note that gets missed: form matters as much as timing. Whole nuts and thick spoonfuls of nut butter are choking hazards for infants regardless of allergy risk, and the safe forms are a separate conversation worth having.

The other big trial, and why peanut and egg aren’t a template

The second major study is less famous and more complicated. Researchers in the United Kingdom recruited 1,303 exclusively breastfed three-month-olds from the general population, not a high-risk group, and randomized them to early introduction of six allergenic foods or to standard practice. The six were peanut, cooked egg, cow’s milk, sesame, whitefish and wheat. Results appeared in the New England Journal of Medicine in 2016.

In the intention-to-treat analysis, which counts everyone as assigned regardless of what they managed to do, food allergy developed in 7.1% of the standard group and 5.6% of the early-introduction group. That difference wasn’t statistically significant. The trial, on its primary measure, was negative.

The per-protocol analysis, limited to families who achieved the early introduction, told a different story: 2.4% versus 7.3% overall, with peanut allergy at 0% versus 2.5% and egg allergy at 1.4% versus 5.5%. There was no significant effect for milk, sesame, fish or wheat. The authors reported that introducing all six foods early was not easily achieved but was safe, and raised the question of whether prevention depends on how much a child eats.

Read honestly, that trial says two things at once. Adherence was the obstacle, and when families cleared it the signal showed up. But a null primary result is still a null result, and it keeps the confident claim narrow. Peanut and egg have the evidence. The other allergenic foods do not yet.

The JAMA review adds one more useful correction. It found high-certainty evidence that the timing of gluten introduction is not associated with celiac disease risk. Another delay rule, retired.

Anaphylaxis is a 911 call, and antihistamines are not the treatment

This is the section to read twice.

Anaphylaxis is a severe, life-threatening allergic reaction that can affect breathing and blood pressure within minutes. The American Academy of Allergy, Asthma & Immunology says symptoms typically start 5 to 30 minutes after exposure, though in some cases it can take more than an hour.

The signs, grouped by where you’d notice them:

• Breathing. Wheezing, coughing, a hoarse voice, chest tightness, trouble breathing, trouble swallowing, a tight or swollen throat.
• Mouth and face. Swollen lips or tongue, swelling at the back of the throat.
• Skin. Hives or welts, widespread redness, itching, swelling, or a face that goes pale or flushed.
• Stomach. Vomiting, diarrhea, cramping.
• Whole body. Weak pulse, dizziness, fainting, collapse, and what AAAAI describes as a feeling of impending doom.

One detail saves lives: a severe reaction can happen without any skin symptoms at all. Parents watch for hives, and hives are the one sign that may not show up. A child with hives alone may be having a mild reaction. A child vomiting and struggling to breathe is not, whatever their skin is doing.

Epinephrine is the treatment. AAAAI’s instruction on the most common and most costly mistake is blunt: “Don’t take an antihistamine or wait to see if symptoms get better.” Antihistamines may settle hives. They do nothing for a closing airway or a dropping blood pressure. Epinephrine does, and every minute of hesitation is a minute the reaction keeps moving.

Call 911. If an auto-injector is available, give it right away and have someone call while you do, which is what MedlinePlus, the National Library of Medicine’s consumer health service, advises for a serious allergic reaction. A child who gets epinephrine still needs emergency evaluation, because reactions can come back after they appear to settle.

If your child has been prescribed an epinephrine auto-injector, it needs to be wherever the child is. School. Daycare. A grandparent’s house. The sports bag. AAAAI recommends that family members and other adults close to the child learn how to use it, and that an anaphylaxis action plan be on file at school, work or camp. A prescription sitting in a kitchen drawer is not a plan.

The gap here is measurable. In a survey of 38,408 US children published in Pediatrics in 2018, 19.0% of food-allergic children had made at least one food-allergy-related emergency department visit in the previous year, 42.0% had made one at some point in their lives, and only 40.7% had a current epinephrine auto-injector prescription. Among food-allergic children, 42.3% reported at least one severe food allergy.

Allergy, intolerance, and the gap between reported and confirmed

These two words get used interchangeably and they describe different machinery.

A food allergy is an immune system response. The National Institute of Diabetes and Digestive and Kidney Diseases draws the line directly, stating that lactose intolerance is not the same thing as a milk allergy. Milk allergy is an immune system disorder. Lactose intolerance is a digestive problem, where the small intestine can’t break down all the lactose a person consumes, producing bloating, gas and diarrhea. NIDDK estimates about 36 percent of people in the United States have lactose malabsorption, and notes that not everyone with it develops symptoms.

The stakes differ, and so does the management. AAAAI points out that someone with an intolerance may be able to handle small amounts of a food without reacting, while a confirmed allergy means strict avoidance. Treating an allergy like an intolerance is dangerous. Treating an intolerance like an allergy removes foods from a child’s diet for no reason.

The scale of that second error is larger than most people assume. That same 2018 Pediatrics survey estimated current food allergy prevalence among US children at 7.6%, after excluding 4% of children whose described reaction history didn’t fit an IgE-mediated allergy. But more than 11% of children were perceived by their parents as food-allergic. NIAID puts food allergy at roughly 8 percent of American children, and nearly 11 percent of adults.

The most common confirmed allergies in that survey were peanut at 2.2%, milk at 1.9%, shellfish at 1.3% and tree nut at 1.2%. AAAAI counts nine foods behind most allergic reactions: cow’s milk, eggs, fish, peanuts, sesame, shellfish, soy, tree nuts and wheat.

The distance between 7.6% and 11% isn’t parents being careless. It’s that real symptoms get attributed to the wrong cause, and that a commercial market has grown up to fill the gap.

Why a food sensitivity panel isn’t an allergy diagnosis

The NIAID expert panel addressed testing directly, and its advice runs against the instinct to order more. In the course of evaluating an infant for peanut introduction, the panel advised against ordering broad food allergen panels and against IgE testing for foods other than peanut.

The mechanism behind that advice is the thing to understand. A positive allergy test shows sensitization, meaning the immune system has produced antibodies to a food. It does not establish that the child will react when they eat it. Order a panel covering twenty foods and you will likely get positives that predict nothing, and each one tends to become a restriction.

AAAAI says proper diagnosis is extremely important and that studies have shown many suspected food allergies turn out to be caused by something else. The diagnostic path AAAAI describes involves skin testing, blood testing, and where needed a supervised food challenge with an allergist.

Direct-to-consumer “food sensitivity” panels measure a different antibody than the one driving allergic reactions, and they are not part of that pathway. The cost of acting on one isn’t just money. Pulling foods out of a growing child’s diet affects nutrition and social eating, and it removes the regular exposure that helps maintain tolerance. The broader market in consumer testing is covered elsewhere on this site.

Six things to settle before the first allergenic food

Bring this to the visit. It takes four minutes and it replaces weeks of searching.

1. Which group is my child in? Ask the clinician to say it plainly: severe eczema, egg allergy, mild eczema, or none of those. Everything downstream depends on the answer.
2. Does my child need testing before peanut, and if so, which test and when?
3. If we’re introducing at home, what form should the food take? The safe form for an infant is not a whole nut or a spoonful of thick butter.
4. What should I watch for afterward, and for how long? Get a specific list and a specific window.
5. If she reacts, what exactly do you want me to do, and does she need a prescription?
6. At what point do I call you, and at what point do I call 911 instead? Settle this while nobody is panicking.

Outside that conversation, a reaction that involves breathing, swelling of the tongue or throat, repeated vomiting with widespread hives, or collapse is not a phone call to the office. It’s 911.

What to take away

The reversal is real, and it’s good news. Early introduction of peanut and egg has better evidence behind it than delay ever did.

But the headline and the instruction are not the same thing. If your baby has severe eczema or a known egg allergy, the first step is an appointment, not a spoon, and testing comes before food. For every other family, the practical move is to bring allergenic foods in alongside other foods rather than saving them for later, in a form that can’t choke a baby, one at a time so you can see what happened.

And learn the signs of anaphylaxis now, while your hands are steady. Breathing trouble, a swelling tongue or throat, vomiting with widespread hives, collapse. Epinephrine, then 911, in that order if the auto-injector is within reach.

Sources

1. National Institute of Allergy and Infectious Diseases. “Guidelines for Clinicians and Patients for Food Allergy.” niaid.nih.gov
2. National Institute of Allergy and Infectious Diseases. “Food Allergy.” niaid.nih.gov
3. NIAID-Sponsored Expert Panel (hosted by the American Academy of Allergy, Asthma & Immunology). “Addendum Guidelines for the Prevention of Peanut Allergy in the United States.” aaaai.org
4. Centers for Disease Control and Prevention. “When, What, and How to Introduce Solid Foods.” cdc.gov
5. MedlinePlus, U.S. National Library of Medicine. “Anaphylaxis.” medlineplus.gov
6. American Academy of Allergy, Asthma & Immunology. “Anaphylaxis.” aaaai.org
7. American Academy of Allergy, Asthma & Immunology. “Food Allergy.” aaaai.org
8. National Institute of Diabetes and Digestive and Kidney Diseases. “Definition & Facts for Lactose Intolerance.” niddk.nih.gov
9. American Academy of Pediatrics (HealthyChildren.org). “Starting Solid Foods.” healthychildren.org
10. PubMed. “Du Toit G, et al. "Randomized trial of peanut consumption in infants at risk for peanut allergy." New England Journal of Medicine, 2015. PubMed. DOI: 10.1056/NEJMoa1414850.” doi.org
11. PubMed. “Du Toit G, et al. "Effect of Avoidance on Peanut Allergy after Early Peanut Consumption." New England Journal of Medicine, 2016. PubMed. DOI: 10.1056/NEJMoa1514209.” doi.org
12. PubMed. “Perkin MR, et al. "Randomized Trial of Introduction of Allergenic Foods in Breast-Fed Infants." New England Journal of Medicine, 2016. PubMed. DOI: 10.1056/NEJMoa1514210.” doi.org
13. PubMed. “Ierodiakonou D, et al. "Timing of Allergenic Food Introduction to the Infant Diet and Risk of Allergic or Autoimmune Disease: A Systematic Review and Meta-analysis." JAMA, 2016. PubMed. DOI: 10.1001/jama.2016.12623.” doi.org
14. PubMed. “Gupta RS, et al. "The Public Health Impact of Parent-Reported Childhood Food Allergies in the United States." Pediatrics, 2018. PubMed. DOI: 10.1542/peds.2018-1235.” doi.org