What the evidence actually says.

What was measured

The LEAP trial randomly assigned 640 infants aged four to eleven months, all of whom already had severe eczema, egg allergy, or both, either to eat peanut regularly or to avoid it entirely until they turned five. Among the infants whose skin-prick test was negative at the start, peanut allergy at age five affected 13.7 percent of the avoidance group and 1.9 percent of the eating group. Among those who already showed mild sensitisation, it was 35.3 percent against 10.6 percent.

The same children were followed to adolescence. At twelve years old, peanut allergy was present in 4.4 percent of the early-eating group and 15.4 percent of the avoidance group. The protection held even though many of these children had gone years at a stretch without eating peanut at all.

PreventADALL asked a version of the same question in 2,397 ordinary newborns rather than high-risk ones. Introducing peanut, milk, wheat and egg from three months of age reduced food allergy at three years from 2.3 percent to 0.9 percent. No serious adverse events were observed.

What was measured

PreventADALL tested this directly. Infants were randomly assigned to bath additives and facial cream at least four times a week from two weeks of age until nine months. Food allergy at three years occurred in 3.0 percent of the emollient group against 2.3 percent in the group that received no intervention.

In the same trial, run at the same time in the same children, the early-feeding arm did work, cutting food allergy to 0.9 percent. So this is not a study that failed to detect anything. It detected one effect and not the other.

What was measured

The scale is not in doubt. In 2021, roughly one in four children in the United States was reported to have a seasonal allergy, eczema or food allergy.

The epithelial barrier hypothesis argues that detergents, surfactants, microplastics, pollutants and heavily processed food damage the linings of the skin, gut and airway, and that this damage is upstream of the rise in allergic and autoimmune disease.

What was measured

Children who went on to develop food allergy tended to have fewer Bifidobacterium and Clostridia species and a less mature infant gut microbiome. The protective effect is thought to run through anti-inflammatory metabolites and regulatory T cells.

The practical point is timing. If the window really is roughly conception to age two, prevention becomes plausible and cheap in a way that treating established disease never is.

What was measured

The vocabulary matters. Desensitisation means a child is protected while they stay on treatment. Remission means the tolerance survives stopping it. The IMPACT trial was designed specifically to separate the two, measuring protection at 134 weeks on treatment and then again after 26 weeks off it.

An anti-IgE antibody given for 16 to 20 weeks allowed 67 percent of treated children to tolerate 600 mg of peanut protein without dose-limiting symptoms, against 33 percent on placebo, and it protected against several foods at once rather than one.

A wearable peanut patch worn daily for twelve months raised the reaction threshold in 67 percent of one-to-three-year-olds against 33.5 percent on placebo. Adverse events occurred in essentially every child in both groups, most of them local skin reactions.

What was measured

Researchers applied a harmless skin bacterium, Roseomonas mucosa, to fifteen children with eczema. They reported reduced severity, better barrier function, less Staphylococcus aureus on the skin and lower steroid use, with improvement persisting up to eight months after treatment stopped.

The shape of that result is exactly what we look for: something grown cheaply, with an effect that outlasts the dosing. The authors are careful to call it a signal rather than proof, and they are right to.