Lead
Until about ten years ago, the received wisdom was to delay allergen introduction: start eggs at twelve months, avoid peanuts until age three. Many parents were given that advice directly. It sounded cautious, protective, logical.
Since 2015, that wisdom has been quietly but fundamentally rewritten. The catalyst was the LEAP trial — a randomized controlled trial in high-risk infants that showed early peanut introduction reduced peanut allergy incidence by more than 80% [1]. Then, in Japan, the PETIT trial demonstrated that planned early introduction of heated egg in infants with eczema cut egg allergy incidence to roughly one quarter of the control rate [2].
"Delay equals safety" is the intuition that reversed. This article traces how that reversal happened, what the trials actually showed, and what it means for household decisions today.
The LEAP trial — what changed and why
LEAP — Learning Early About Peanut Allergy — was a large randomized controlled trial conducted in the United Kingdom by Du Toit and colleagues, published in the New England Journal of Medicine in 2015 [1]. The study enrolled 640 high-risk infants aged four to eleven months — defined as having severe eczema, egg allergy, or both — and randomly assigned them either to consume peanut protein regularly from enrollment or to avoid peanuts entirely until age five. The primary outcome was peanut allergy prevalence at age five [1].
The results were large by clinical trial standards. Among infants with a negative skin prick test at baseline, peanut allergy developed in 13.7% of the avoidance group versus 1.9% of the consumption group. Among infants with a positive baseline skin test, the figures were 35.3% versus 10.6% [1]. Relative risk reduction was approximately 80% — and this was not an observational finding. It came from a randomized controlled trial with allocation concealment.
The follow-on LEAP-On trial added a further finding: infants in the consumption group who then avoided peanuts for twelve months still retained their protection [1]. This suggests not a transient desensitization but a durable immunological tolerance: a state in which the immune system has learned to accept a substance without mounting an allergic or harmful response — the underlying mechanism most consistent with this outcome is the skin sensitization hypothesis: the idea that food proteins entering through inflamed or broken skin teach the immune system to react allergically, while early oral exposure teaches it tolerance instead, in which eczematous skin exposed to environmental food proteins drives sensitization, while oral exposure from an early age induces tolerance.
Before LEAP, introducing peanuts early to a baby with eczema was widely considered hazardous. After LEAP, the recommendation for the same population shifted to early introduction being actively preferred. A single RCT moving a clinical consensus is uncommon enough to be worth noting.
The PETIT trial — the same logic applied to egg in Japan
The domestic counterpart is the PETIT trial, led by the National Center for Child Health and Development in Tokyo. Natsume and colleagues published the results in The Lancet in 2017 [2]. The trial enrolled 121 infants aged four to five months with atopic dermatitis, treated their eczema aggressively as a concurrent intervention, and then randomized them to receive progressively increasing amounts of heated egg powder or a matched placebo [2].
The primary analysis found egg allergy in 8% of the active group versus 38% of the placebo group [2]. The interim analysis showed such clear efficacy that the trial was halted early. The two key design choices — treating the eczema first, and introducing the allergen incrementally in heated form — were both deliberate; PETIT was built on the same skin sensitization hypothesis as LEAP.
PETIT matters for Japanese families for a reason beyond its p-values: egg is the most common food allergen in Japanese infants, and this was a Japanese population trial. The findings translated directly into the 2021 edition of the Japanese Society of Pediatric Allergy and Clinical Immunology's Food Allergy Clinical Practice Guidelines, which now states that early intake of both eggs and peanuts can be expected to reduce allergy incidence in high-risk infants [3].
The EAT trial — a necessary complication
Not everything simplified into "introduce early and prevent allergy." Perkin and colleagues published the EAT (Enquiring About Tolerance) trial in the New England Journal of Medicine in 2016 [4]. It enrolled 1,303 breastfed infants from the general UK population and randomized them at three months to either introduce six allergenic foods early (peanut, cooked egg, cow's milk, sesame, white fish, and wheat) or to continue exclusive breastfeeding until six months [4].
The intention-to-treat: an analysis approach that includes all randomized participants in the groups they were assigned to, regardless of whether they actually followed the protocol analysis — covering all randomized infants regardless of adherence — showed no statistically significant difference in allergy incidence to any of the six foods (5.6% early introduction vs. 7.1% standard) [4]. However, the per-protocol: an analysis approach that includes only participants who actually completed the assigned intervention as planned, which can overestimate real-world effectiveness analysis — restricted to infants who actually consumed the foods as directed — found significantly lower rates in the early group: 0% vs. 2.5% for peanut allergy, 1.4% vs. 5.5% for egg allergy [4].
The gap between the two analyses tells its own story: introduction works, but sustaining it in real households proved difficult. A substantial proportion of families in the early-introduction arm could not maintain the protocol. The ITT-negative result does not mean early introduction is ineffective; the per-protocol result shows efficacy in those who completed the regimen. Together, they show that effectiveness in a trial setting and effectiveness in routine practice are not the same thing.
The EAT trial also clarifies a distinction the LEAP enthusiasm can obscure: the high-risk population (infants with eczema or existing food allergy) and the general population are different. LEAP and PETIT showed large, consistent effects in high-risk groups. EAT, testing general-population infants, found effects that depended on adherence and were not detectable at the intention-to-treat level. Current guidelines reflect this distinction — early introduction is strongly recommended for high-risk infants, and for the general population the guidance is to introduce complementary foods, including allergens, at the appropriate developmental stage without deliberately delaying them [3,5].
Implementation — where the difficulty lies
Following the LEAP and PETIT data, the American Academy of Pediatrics issued an addendum to its 2017 guidelines, and the 2019 clinical report by Greer and colleagues stated explicitly that "there is no convincing evidence that delaying the introduction of allergenic complementary foods beyond four to six months of age prevents atopic disease," and that early peanut introduction prevents peanut allergy [5]. The 2021 Japanese Food Allergy Clinical Practice Guidelines align with this direction [3].
What the guidelines do not simplify away is the implementation challenge. EAT's per-protocol dropout rates illustrated this, and real households face the same obstacles:
- Assessing risk level. Infants with severe eczema or a known food allergy should not have allergens introduced based on household judgment alone — specialist evaluation first is the recommended sequence [3].
- Continuing regular exposure. A single introduction is not the mechanism. The immunological tolerance documented in LEAP and PETIT was built through repeated, ongoing consumption — typically several times per week at small amounts [1,2]. "Tried it once" is not equivalent to "introduced it."
- Recognizing and managing reactions. Early introduction does not guarantee no reaction. The prerequisite is being able to notice and respond to an allergic reaction if it occurs.
This means that while the direction of the guidance is now clear, the boundary between what households can manage independently and what requires medical supervision remains somewhat indistinct. Infants with eczema, a family history of food allergy, or any prior food reaction should move toward a clinician before attempting early allergen introduction at home.
Practical anchors for household decisions
Drawing the above together, four principles hold up across the evidence:
- There is no evidence-based reason to avoid common allergenic foods during the 4–6 month complementary feeding window [3,5]. Deliberate avoidance in the hope of reducing allergy risk runs counter to current evidence.
- Control eczema actively. PETIT's trial design treated eczema aggressively as a concurrent intervention, and the skin sensitization hypothesis suggests that untreated eczema is a route into food sensitization independent of oral exposure timing [2,3].
- Record the first introduction. Food introduced, approximate amount, time of day, and any observed reaction — even negative ("nothing happened") — is information that becomes valuable if a reaction occurs later and a clinician needs to reconstruct the history. A recording app like Memori or a written note will do. The four data points are: what, how much, when, what happened.
- For high-risk infants, consult before introducing. Severe eczema, a known food allergy, or a family history with a high prior probability of allergy are all reasons to arrange a medical evaluation first.
Summary
The "delay equals safety" rule that governed allergen introduction until roughly a decade ago has been reversed by randomized trial evidence [1,2]. In high-risk infants — those with eczema or existing food allergy — planned early introduction of peanut and egg substantially reduces the likelihood of developing those allergies. Japan's PETIT trial extended the same finding to egg in a Japanese population [2]. The EAT trial simultaneously showed that intention-to-treat effects in the general population depend on adherence, and that converting trial-level evidence into household-level practice is its own challenge [4].
The landscape has changed. That change does not mean households should proceed unilaterally — it means knowing enough to ask the right question at the next pediatric visit.
References
- Du Toit G, Roberts G, Sayre PH, et al. Randomized trial of peanut consumption in infants at risk for peanut allergy. N Engl J Med. 2015;372(9):803–813. doi:10.1056/NEJMoa1414850. PMID: 25705822.
- Natsume O, Kabashima S, Nakazato J, et al. Two-step egg introduction for prevention of egg allergy in high-risk infants with eczema (PETIT): a randomised, double-blind, placebo-controlled trial. Lancet. 2017;389(10066):276–286. doi:10.1016/S0140-6736(16)31418-0. PMID: 27939035.
- Ebisawa M, Ito K, Fujisawa T, eds. Food Allergy Clinical Practice Guidelines 2021. Japanese Society of Pediatric Allergy and Clinical Immunology; 2021.
- Perkin MR, Logan K, Tseng A, et al. Randomized trial of introduction of allergenic foods in breast-fed infants. N Engl J Med. 2016;374(18):1733–1743. doi:10.1056/NEJMoa1514210. PMID: 26943128.
- Greer FR, Sicherer SH, Burks AW; Committee on Nutrition; Section on Allergy and Immunology. The Effects of Early Nutritional Interventions on the Development of Atopic Disease in Infants and Children: The Role of Maternal Dietary Restriction, Breastfeeding, Hydrolyzed Formulas, and Timing of Introduction of Allergenic Complementary Foods. Pediatrics. 2019;143(4):e20190281. doi:10.1542/peds.2019-0281. PMID: 30886111.