All topics › Feeding breastfeeding formula
Does introducing peanut and egg early reduce food allergy, and who should do it?
For decades, parents were told to delay allergenic foods. Then a landmark trial reversed the advice for peanut. Does introducing peanut and egg early actually prevent food allergy — and does the answer differ for high-risk babies versus everyone else?
For infants at high risk of peanut allergy — those with severe eczema, an egg allergy, or both — introducing peanut early (around 4–6 months) substantially reduced peanut allergy by age 5 in a randomized trial: among initially non-sensitized high-risk infants, 137 in 1,000 developed peanut allergy with avoidance versus 19 in 1,000 with early consumption [1]. For the general population the evidence is weaker: the EAT trial's primary result was inconclusive — 71 vs 56 in 1,000 for any food allergy, not statistically significant [3]. Current UK guidance says peanut and egg need not be delayed or treated differently from other foods from around 6 months, and that deliberately excluding them beyond 6–12 months may increase risk — but babies with early eczema or suspected allergy should get medical advice first [4].
The LEAP trial randomized 640 infants aged 4–11 months who all had severe eczema, egg allergy, or both — the highest-risk group for peanut allergy — to either consume peanut regularly or avoid it until age 5 [1]. At 60 months, among infants who started out not sensitized to peanut: 137 in 1,000 in the avoidance group had peanut allergy versus 19 in 1,000 in the consumption group — 118 fewer per 1,000 [1]. Among infants who started out mildly sensitized, the gap was even larger in absolute terms: 353 versus 106 in 1,000 — 247 fewer per 1,000 [1]. There was no significant difference in serious adverse events [1].
A follow-up (LEAP-On) then told families to avoid peanut for a year — and the protection persisted: at 72 months, peanut allergy was 186 in 1,000 in the original avoidance group versus 48 in 1,000 in the original consumption group [2].
Rating: A — strong, for high-risk infants. This is the trial that changed guidelines worldwide. But it was only high-risk infants: it does not automatically prove the same benefit for babies with no eczema and no egg allergy.
The EAT trial randomized 1,303 exclusively breastfed general-population infants to early introduction of six allergenic foods (peanut, egg, milk, sesame, fish, wheat) from 3 months versus standard introduction at around 6 months [3]. The primary, pre-specified result — food allergy by age 3 in everyone randomized — was 71 in 1,000 with standard introduction versus 56 in 1,000 with early introduction: not statistically significant (P=.32) [3].
Among only the families who actually managed to follow the demanding early-introduction regimen, the results looked much better: any food allergy 73 vs 24 in 1,000, peanut allergy 25 vs 0 in 1,000, egg allergy 55 vs 14 in 1,000 [3]. But per-protocol analyses are biased by design — families who can stick to a demanding feeding schedule differ systematically from those who can't — so these numbers support the idea without proving it [3].
Rating: B — moderate, downgraded because the primary result was inconclusive and the favorable numbers come from the biased per-protocol analysis.
The PETIT trial (Japan, n=147) tested early introduction of heated egg in infants with eczema who had not yet reacted to egg, alongside aggressive eczema treatment — a combined regimen, not egg alone [6]. The trial was stopped early; in the primary analysis, egg allergy developed in 8 in 100 (5/60) of the heated-egg group versus 38 in 100 (23/61) of the placebo group [6]. But the egg arm had six hospital admissions versus none in placebo, and the trial was small and stopped early [6]. This is evidence for a specific high-risk, medically supervised regimen — not a DIY egg routine for every baby.
Rating: C — low, for the general reader's purposes: small, stopped early, safety events in the treatment arm, and inseparable from the concurrent eczema treatment.
Note the US high-risk recommendation includes pre-introduction testing, while UK guidance routes high-risk families through medical advice rather than routine testing — a genuine difference in approach, not in the underlying science.
If your baby has significant eczema or a diagnosed/suspected food allergy: this is the group the strong evidence is about. Talk to your GP or an allergist before introducing peanut — the guidelines are explicit on this [4, 5]. Don't try to replicate LEAP at home with a high-risk baby without medical input.
If your baby has no eczema and no allergy concerns: the evidence doesn't show a proven benefit of very early (3–4 month) introduction for your baby specifically, and the trial did not identify a safety signal for early introduction; guidance says there's no reason to delay peanut or egg past around 6 months [3, 4]. Introducing them as ordinary foods from around 6 months is the reasonable middle ground.
The burden is real and unstudied. Early introduction done "properly" means regular, repeated exposure to specific foods in specific forms for months — the EAT trial itself showed many families couldn't keep it up [3]. Nobody has measured parental anxiety about reactions, preparation time, or cost as trial outcomes. If the regimen is stressing your family out, that cost is real even though it's unquantified — and for low-risk babies, the proven benefit of a strict early regimen is uncertain anyway.
Relationships, work, and money. In practice the preparation and worry tend to fall on whichever caregiver does the feeding — no trial has measured how allergen introduction affects partners, sleep, return to work, or household food budgets. If you're juggling work and childcare, a simple version (peanut butter in porridge a few times a week from around 6 months, per guidance [4]) is more sustainable than a strict LEAP-style regimen, and for low-risk babies the evidence doesn't demand the strict version anyway.
Choking safety. Whole peanuts and chunks of peanut butter are choking hazards; use smooth peanut butter thinned with warm water, or peanut powder mixed into familiar food [5]. Introduce allergens one at a time, in small amounts, so you can spot a reaction — and keep offering them regularly once tolerated.
| Population | Outcome | Avoidance / standard | Early introduction | Difference |
|---|---|---|---|---|
| High-risk infants, not yet sensitized (LEAP) | Peanut allergy at 60 months | 137 in 1,000 | 19 in 1,000 | 118 fewer per 1,000 [1] |
| High-risk infants, mildly sensitized (LEAP) | Peanut allergy at 60 months | 353 in 1,000 | 106 in 1,000 | 247 fewer per 1,000 [1] |
| High-risk infants (LEAP-On) | Peanut allergy at 72 months, after 1 year of avoidance | 186 in 1,000 | 48 in 1,000 | 138 fewer per 1,000 [2] |
| General population (EAT, primary result) | Any food allergy at 36 months | 71 in 1,000 | 56 in 1,000 | Not significant [3] |
| Infants with eczema, not yet egg-allergic (PETIT) | Egg allergy | 38 in 100 | 8 in 100 | Treatment arm had 6 hospital admissions vs 0 in placebo [6] |
| Serious adverse events (LEAP) | — | No significant difference between groups | — | [1] |
The number needed to treat in LEAP's non-sensitized high-risk group was about 9 (1,000 ÷ 118) — an unusually large effect for a dietary intervention. In the general population, no reliable NNT can be calculated because the primary result was inconclusive.