Maternal peanut consumption, early introduction protect against allergy

Matthew Greenhawt, MD, MSc

Is early peanut introduction the only factor that promotes immunologic tolerance to peanut?  Surely not. Although the LEAP study showed that early peanut introduction is protective against developing peanut allergy, many leading experts believe that a host of other factors also play a role besides the timing of introduction of potentially allergenic solids.  These include vitamin D, early diversity of the diet, maternal diet during pregnancy, and the microbiome among leading theories.  

In this month’s Journal of Allergy and Clinical Immunology, secondary analysis of nested data from a large, longitudinal Canadian birth cohort, the Canadian Asthma and Primary Prevention Study (CAPPS), took a look at this very issue.  Infants were enrolled starting in 1995, predating the 2000 AAP recommendations for delayed peanut introduction in at-risk infants (with a parental history of allergic disease).  The cohort had nested data from caregiver questionnaires related to breast-feeding, infant feeding, and allergic outcomes during the study, as well as had allergy skin testing to foods (but not actual food allergy) as part of the final outcome assessment.  Food-related outcomes were not a primary aim of this study, so complete data were available on 342/545 of the originally enrolled children.  

Investigators noted that only 23% of caregivers introduced peanut in the first year of life — with the rest of the cohort introducing peanut equally in the second year and the third year of life.  Nearly 58% of the mothers reported consuming peanut while breast-feeding (which per the year 2000 AAP guidelines was also not recommended). At the end of the study, when the children were aged 7 years, only 9.4% were sensitized to peanut (eg, have positive skin testing to peanut).  Athough the highest rates of peanut- sensitized children were from mothers consuming peanut while breast-feeding or who introduced peanut in the first year of life, the lowest incidence of sensitization was in the children where the mom both ate peanut while breast-feeding and introduced peanut in the first year of life.  It was unclear how much overlap there was between these two factors.

Although this finding is interesting, there are many problems with the study. First, this does not imply allergy, just peanut sensitization.  To be allergic, one needs both sensitization to the food and IgE-mediated symptoms upon ingestion of that food. A positive test alone does not mean allergy.  Second, there is considerable recall bias. There was poor reporting of the frequency/amounts of exposure of the peanut consumption while both breast-feeding and in the first year of life, as well as how precise the degree of overlap between maternal exposure and peanut introduction in the child may have been. Thus, these findings highlight a potential association of interest but do not implicate causality, so enthusiasm must be tempered.  It is important to note that in the U.S., Canada, the U.K., Australia and New Zealand, peanut is recommended to be given in the first year of life, and that no guidelines recommend any allergen avoidance while pregnant or while breast-feeding. Therefore, there would not be any implications to change current clinical practices based on this study.

I do caution that this association may be very difficult to prove definitively, even with a better designed study. Remember, the LEAP study looked only at timing of peanut introduction as a randomized intervention, and did not randomize any other outcome. Therefore, even similar nested data from the LEAP study (with a definitive outcome of who was peanut allergic or not) could do no more than reaffirm the potential association seen in the CAAPS data but could not show causality.  The issue is with maternal allergen transfer. Although many believe that allergens can pass through breast milk to the infant, this has not actually been conclusively demonstrated to happen consistently in previous studies. Therefore, considerable work would be needed first to definitively prove the hypothesis of passive allergen transfer through breast milk, that this can induce allergy in the infant, and then design a study that could control the amount/frequency the mother consumes as a variable to interact with an exact timing of overlap with active peanut introduction in the infant. This would be exceptionally difficult to do.

Nonetheless, the take-home message from this study is to inform that although the early introduction of peanut may be a key factor in preventing peanut allergy, it may not be the only such factor.