Introducing Allergens to a Baby With Eczema: Why Severity Matters

The NIAID addendum guidelines for peanut allergy prevention, written in 2017 by a panel for the US National Institute of Allergy and Infectious Diseases, sort babies into three groups, and eczema severity separates the first two. Guideline 1 covers "infants with severe eczema, egg allergy, or both" and recommends "introduction of age-appropriate peanut-containing food as early as 4 to 6 months of age to reduce the risk of peanut allergy." The same guideline recommends that peanut-specific IgE, skin prick tests, or both "be strongly considered before introduction of peanut to determine if peanut should be introduced and, if so, the preferred method of introduction."

Guideline 2 covers mild to moderate eczema: peanut "around 6 months of age", and those babies "may have dietary peanut introduced at home without an in-office evaluation." Guideline 3 covers babies with no eczema and no food allergy, for whom peanut is "freely introduced in the diet". Severe has a definition in the addendum too: eczema that persists or keeps coming back, "assessed as severe by a health care provider", and needing frequent prescription-strength treatment despite regular emollient use. That is a clinical judgement, so we left the label to our paediatrician.

The addendum rests largely on one trial. LEAP (Du Toit and colleagues, New England Journal of Medicine, 2015) randomly assigned 640 babies with severe eczema, egg allergy, or both, all at least 4 months old and younger than 11 months, to eat peanut regularly or avoid it until age five. A skin prick test at the start split them into two cohorts: babies with no measurable wheal, and babies with a small one of 1 to 4 mm.

Among babies who tested negative at the start, peanut allergy at 60 months was 13.7% in the avoidance group and 1.9% in the group that ate peanut. Among those who began with a small positive result, it was 35.3% against 10.6%. Babies with a larger skin response at screening were not enrolled, as the addendum notes. So the trial that showed early peanut helps was run in exactly the babies parents worry about most, and every one of them was tested before the first spoonful.

The 2026 ASCIA guideline on infant feeding for food allergy prevention comes at eczema from a different direction. On egg, ASCIA notes that infants with eczema in the first months of life, "particularly moderate to severe", are at higher risk of developing an egg allergy, and says the recommendation to introduce well-cooked egg once solids start at around 6 months "also applies to these infants." The next sentence is short: "Eczema should be well managed."

On peanut, ASCIA "recommends timely introduction of peanut for infants irrespective of their family or personal history of allergy", in a form such as smooth peanut butter, finely ground peanut or peanut flour. Whole peanuts are a choking hazard at this age. Our peanut walkthrough covers thinning the butter. The guideline also speaks to eczema around the mouth: a barrier cream before eating "can be used to reduce the risk of irritation from direct food contact", and a mild rash around the mouth with no other symptoms "may not be a sign of an allergic reaction."

We are not going to recommend a cream, and nothing here is treatment advice. What we leaned on was the NHS page on atopic eczema, which lists applying emollients "as often as possible (at least 2 times a day)" and carrying on even when the eczema improves, avoiding things that make it worse such as soap and washing detergent, and not changing diet "unless you are told to by a doctor", which the NHS says includes breastfeeding a baby with eczema.

The same page lists "certain foods" among the things that can make eczema worse, next to heat, pollen and house-dust mites. A flare on its own proves very little for that reason. For us the useful part was about the log. A new soap, a new washing powder or a hot week can show up on the skin much as a food can, so we kept the skin routine the same in any week we started a new allergen. When two things change on the same day, no record can pull them apart afterwards.

Ours was the slow kind. The patches sat on our baby's cheeks and behind the knees, and they flared a day or two after dairy, which we only saw once the meals and the skin were written down against the same dates. The cow's milk allergy post tells that part in full.

That gap is what the log was for. Hives, swelling around the lips or eyes, or vomiting within two hours of a food is the fast pattern, and it sits right next to the meal on the timeline. An eczema flare that starts the next day or the day after has a different shape, and it only means something if it keeps landing after the same food. We wrote down the time of each meal, then separately the time the skin changed and where on the body. We photographed the patches by a window in daylight, the same spot each time, because bedtime lamplight made every patch look equally red. Our post on eczema and delayed reactions goes further into reading that timeline. Trouble breathing, floppiness or a swelling face is an emergency call, not a log entry.