The Rehearsal That Ended in Silence
Australia now runs a national experiment that no other country has attempted at this scale. Since July 2024, every infant diagnosed with peanut allergy before twelve months of age has been offered free treatment across ten public children’s hospitals. The program carries the name ADAPT. Its stated ambition is to bend the arc of a life: from allergy development to an accelerated pathway to tolerance.
Before any infant enters treatment, the child must pass through a gate. That gate is the oral food challenge. Small, escalating doses of peanut flour are fed on a single day under medical supervision. The cumulative total reaches 2,435 milligrams of peanut protein, delivered in teaspoon measures of peanut flour. Brettig, the program’s medical lead, frames the dose as deliberately generous: it sits above the threshold where every child of this age with a real allergy should react. If the child reacts, the allergy is confirmed and treatment begins. If the child sails through untouched, the label was wrong.
Then came the number nobody planned for. Of the first 778 babies with a prior peanut allergy diagnosis, 35 per cent reacted to no dose at all. [3] These were not children with vague histories. They had developed rashes, swelling, hives or other signs after eating peanut-containing food. Their skin prick or blood tests pointed the same way. And still, on challenge day, their bodies stayed quiet. “This was a surprising result for us,” Brettig says. [3] “We expected a small proportion might pass the food challenge, but not 35 per cent.” [3] For these infants, the allergy that doctors had written into their records simply was not there anymore — or perhaps had never been as solid as the tests implied. The average gap between the original diagnosis and the challenge was six months. In half a year, a third of the diagnoses dissolved. Brettig offers one possible reading: a transient allergy, where the immune system fires an aberrant response to a food early on and then lets it go.
What the Tests Cannot Tell You
The scale of the gap between diagnosis and reality exposes a structural weakness in how peanut allergy is normally identified. Most diagnoses rest on clinical history plus skin prick or blood results. The oral food challenge — the test that shows what a child’s body actually does when peanut arrives — is usually skipped. John Ziegler at the University of New South Wales, Sydney, draws the consequence plainly: routine challenges would separate the children who need treatment from those who do not, and that separation would pay for itself. “The study confirms the importance of confirmation of the diagnosis of peanut allergy before embarking on oral immunotherapy,” he says.
There is a second, quieter limit here. Nobody knows why a third of these infants stopped reacting. Brettig does not pretend otherwise. The mechanism behind the vanishing allergy remains unmeasured, and the six-month window is too short and too thinly observed to explain it. What the program can say is what it saw: a large share of infant peanut allergies did not survive their first year.
The stakes of a wrong label are not abstract. A child tagged with peanut allergy carries restricted food choices, anxiety about accidental exposure, and emergency medication for years. Removing that label, Brettig says, will “make a huge difference.” So an unnecessary label may not merely burden a childhood — Brettig suggests it may also raise the risk of a true peanut allergy later, since avoiding peanut in early life is known to increase that risk.
The Same Third, Seen Over a Decade
The finding lands on ground already prepared by a longer study. In the HealthNuts cohort, 156 one-year-olds with challenge-confirmed peanut allergy were followed at ages four, six and ten. [1] By ten years of age, 33.9 per cent had outgrown the allergy. [1] Nearly all of those resolutions — 97.4 per cent — had already happened by age six. [1] One third of challenge-confirmed infant peanut allergy resolved by age ten, most of it by age six; one third of diagnosed infants in the Australian program turned out never to have had it at the moment of testing. The same fraction appears twice, from opposite directions: one measured over years, the other caught in a single afternoon.
The HealthNuts work also mapped the biology behind natural resolution. Falling levels of IgE against peanut and against Ara h 2, a peanut protein, tracked with the allergy disappearing. [2] Rising levels of IgG4 against the same targets did too, as did the ratios between the two. Yet no single biomarker threshold at diagnosis separated the children who would resolve from those who would not. Peanut-specific IgE measured at age one carried the strongest prognostic signal, with an area under the curve of 0.75 — useful, not decisive. [2]
That is the bridge between the two lines of research. One shows that the immune system can let go of peanut over years, and names the antibody shifts that accompany the release. The other shows that the letting-go can already be complete by the time a six-month-old diagnosis reaches a hospital door.