In Israel, there's a custom of feeding infants a popular peanut butter snack "as soon as they possibly can," he said. "And they found out that the infants in Israel had a remarkably lower incidence of peanut allergy than infants in the UK -- even Jewish infants in the UK whose parents did not follow the custom" of giving infants the peanut snack.

From this largely unscientific observation, Nepom and his colleagues constructed a scientific study.

"We designed the trial and ran the trial because there was a clear public health need and there was a lot of rationale for it," Nepom said.

More than 600 children participated in the study, and each was randomly assigned to one of two groups. One group consumed low-dose peanut-containing food three times a week, starting in the first year of life and continuing to the age of 5. The other group abstained from peanuts for the first five years of life.

Nepom stressed that the parents did not give whole peanuts to babies: "That would be very dangerous. There's a choking hazard there."

Instead, they had peanut butter or, more likely, Bamba, a peanut-flavored puff commercially produced in Israel.

All the children participating in the study were at high risk of peanut allergy due to family history or having eczema or egg allergy themselves, said Nepom.

At age 5, the children in both groups were given peanuts and observed, Nepom said: Eighteen percent of the children who had been avoiding peanuts had a peanut allergy at age 5, compared with only 1% of the children who had been introduced to peanut butter or Bamba early in life.

"This showed that early introduction of peanut flour had over 80% prevention effect," Nepom said.