Intended for healthcare professionals

Letters

Resolution of peanut allergy

BMJ 1998; 317 doi: https://doi.org/10.1136/bmj.317.7168.1317 (Published 07 November 1998) Cite this as: BMJ 1998;317:1317

Patients have not been proved to grow out of peanut allergy

  1. Tim David, Professor of child health
  1. Booth Hall Children's Hospital, Manchester M9 7AA
  2. Institute of Child Health, London WC1N 1EH
  3. South Manchester University Hospitals, Withington Hospital, Manchester M20 2LR
  4. University of Southampton, Southampton University Hospitals NHS Trust, Southampton SO16 6YD

    EDITOR—Histories of allergy are known to be unreliable, and this is particularly the case for parents' reports of a child's reactions to foods. When double blind food challenges were used, parents' reports could be confirmed in only 37 (28%) of 133 children with reported food intolerance; in another study only 27 (33%) of 81 reports of food intolerance in children could be confirmed. 1 2

    Hourihane et al performed an open food challenge in children with a history of possible peanut allergy.3 Serious doubt that some of these children had genuine peanut allergy arose either because the result of a skin test was negative (this is rare in subjects who have allergic reactions to peanut) or because the child was reported to have eaten peanut without problems (which suggested that peanut allergy was not present). When challenged, some patients had no reaction. The authors concluded, reasonably, that one should be prepared to challenge preschool children with reported peanut allergy because some of them will turn out to be tolerant.

    The unresolved question is whether children who fail to react to a challenge ever had peanut allergy in the first place. This paper contains no proof that those with “resolving” peanut allergy ever had peanut allergy, so caution is needed about the suggestion that some patients with peanut allergy grow out of the problem. Close examination of other claims of patients growing out of peanut allergy casts doubt on the original diagnosis. To prove that a patient has grown out of a food allergy one has to prove that he or she had the allergy in the first place.

    Some of the “resolvers” in Hourihane et al's study had had up to seven reported reactions to peanut, which suggests that they really did have peanut allergy. In my experience, however, it is remarkable how “definite” food allergy can evaporate once it is exposed to the test of a proper food challenge.

    References

    Authors' reply

    1. Jonathan O'B Hourihane, Lecturer in immunobiology,
    2. Stephen A Roberts, Consultant paediatrician,
    3. John O Warner, Professor of child health
    1. Booth Hall Children's Hospital, Manchester M9 7AA
    2. Institute of Child Health, London WC1N 1EH
    3. South Manchester University Hospitals, Withington Hospital, Manchester M20 2LR
    4. University of Southampton, Southampton University Hospitals NHS Trust, Southampton SO16 6YD

      EDITOR—We agree with David's statement about the rarity of peanut allergy in a person with a negative result of a skinprick test to peanut. Since our paper was published one of us (SAR) has identified a child in whom a challenge gave a negative result after having given a positive one previously. In a previous study of adult subjects we showed that a history of peanut allergy was 100% sensitive and 86% specific (60/69 histories supported by positive challenge).1 Similarly, a positive result of a skinprick test to peanut was 100% sensitive and 96% specific (2/62 subjects with a positive result were negative to peanut on challenge).

      David's “unresolved question” is discussed at length in our paper. We stated that our results suggest with some caution that some children grow out of peanut allergy, and we accept that absolute proof of resolution is absent.

      Allergic reactions to peanut are usually stereotyped, and the absence of typical features predicts absence of peanut allergy. Our patients were reported to have typical features of peanut allergy. We could find no other explanation for their symptoms. Some had negative results to skinprick tests, and after food challenge testing we concluded that clinical reactivity was absent. Other children with positive results to skinprick tests were found to be negative on challenge. Again no historical feature distinguished them from controls with persisting peanut allergy. The size of the response to the skinprick test distinguished the groups.

      If David's main point is that no child should be diagnosed as allergic to peanuts in the absence of a positive result on challenge testing he is wrong. Many children are too severely affected for challenge to be justifiable. Our experience suggests, however, that there is another group of young children who, in all good faith, have been treated as allergic to peanuts but do not have persisting disease. To identify such children without a challenge is impossible. The opportunity should not be lost to remove the severe anxiety that families have because of suspected peanut allergy. We attempted to identify historical features that may help paediatricians “flag” these children for challenge.

      We deplore the paucity of nationwide facilities for the appropriate management of children with acute and life threatening allergies; unfortunately this means that the wait for a food challenge test is unacceptably long.

      References

      Log in

      Log in through your institution

      Subscribe

      * For online subscription