Food allergies are commonly misunderstood, so please bear with me while I first explain what food allergies are and are not. Various foods can cause all sorts of unpleasant effects. Most of these are not allergies. Allergies are only reactions caused by a specific antibody (called IgE) that results in hives, trouble breathing, or a life-threatening condition called anaphylaxis. So, if yogurt gives you diarrhea, that’s not an allergy. It might be lactose intolerance. If coffee gives you palpitations, you’re not allergic to coffee; you’re having a side-effect from the caffeine. Ditto chocolate worsening your heartburn; not an allergy.

Of all foods that cause allergic reactions, peanut allergies are the leading cause of anaphylaxis and death, and the prevalence of peanut allergies in the US has grown fivefold in the last 13 years, from 0.4% in 1997 to more than 2% in 2010. This increasing prevalence of a potentially life-threatening allergy has caused some schools to ban peanut products and has caused some airlines to stop offering peanuts in their snacks.

Believing that repeated exposure in infancy of allergy-causing foods leads to allergies, health officials in the UK in 1998 and in the US in 2000 published guidelines recommending the exclusion of foods likely to cause allergies from the diets of infants at high risk of developing allergies. But subsequent studies failed to show that elimination prevented the development of allergies, so the recommendations were withdrawn in 2008. Since then, pediatricians have had no solid evidence on which to base recommendations, until now.

A study in the UK published this week in the New England Journal of Medicine (NEJM) enrolled 640 infants between the ages of 4 and 11 months who were considered to be at high risk for peanut allergy because they had severe eczema or egg allergies, or both. They were all given a skin-prick test to check for peanut sensitivity. The infants that had a severe reaction to the skin-prick test were excluded from the study. Infants who had no reaction or a mild reaction were enrolled and were randomized to two groups.

The parents of children in one group were told that their children should avoid peanut products. The parents of children in the second group were instructed to give their children at least two grams of peanut protein three times a week. (Their first exposure to peanut protein was done under medical supervision.)

The peanut source given to the infants in the study was Bamba, an extremely popular Israeli children’s snack made from puffed corn and peanut butter. If you’ve spent any time in Israel around kids you’ve seen Bamba. Hilariously, the authors admit that “it was not possible to administer a placebo for Bamba because of financial and logistic constraints.” I can imagine the researchers desperately trying to figure out how to make something that looked and tasted like Bamba but without peanuts, and then giving up when they realized that that this would be more expensive and take longer than the rest of the study. The authors tell us that smooth peanut butter was supplied for those infants who didn’t like Bamba, but intensive psychiatric testing would have been more appropriate, because Bamba is delicious.

The children were followed until they were five years old and then given a supervised oral challenge of peanut protein to test them for allergies.

The results were quite dramatic. Among the children who initially had no reaction to the peanut sensitivity skin-prick test, 13.7% (about 1 in 7) of the children who avoided peanuts became allergic, compared to 1.9% (about 1 in 50) of children who consumed peanuts. That means that for every 8 children who consumed peanuts one fewer child developed a peanut allergy.

The results in children who initially had a mild reaction to the skin-prick test were even more impressive. These children were at much higher risk of becoming allergic since their mild skin test result suggests that their immune system had already been partially sensitized to peanut protein. 35.3% (about 1 in 3) of the children who avoided peanuts became allergic, compared to 10.6% (about 1 in 10) of the children who consumed peanuts. That means that for every 4 children with a mildly positive skin test who consumed peanuts, one fewer child became allergic.

Recommendations will likely be updated to account for these findings. First, infants with no eczema or family history of peanut allergies are at low risk of developing allergies and should start eating peanut products as soon as they start eating solid foods. (Don’t feed whole peanuts to infants. They’re a choking hazard. Anyway, Bamba tastes better and now might be one of the most evidence-based snacks.) Infants who are at high risk for peanut allergy because of eczema, an egg allergy, or a family history of peanut allergy should have a skin test to check for sensitivity to peanut. Those who have a negative test can proceed with Bambafication. Those who have a positive test should have their first exposure to peanut product under the supervision of an allergist.

Learn more:

Exposing infants to peanuts causes big reduction in peanut allergy, study shows (The Washington Post)
Feeding Infants Peanut Products Could Prevent Allergies, Study Suggests (Well, New York Times health blog)
About-Face on Preventing Peanut Allergies (Wall Street Journal)
The LEAP Trial (NEJM Quick Take video)
Randomized Trial of Peanut Consumption in Infants at Risk for Peanut Allergy (NEJM article)
Preventing Peanut Allergy through Early Consumption — Ready for Prime Time? (NEJM editorial)