Does what a pregnant mother eat really influence the development of food allergy in her child? This is a highly debatable question, without a good answer or well-designed studies to provide an answer. Currently available data on this subject are conflicting. Expert opinion on the matter has often trended toward conservative recommendations for avoidance of high-risk foods such as peanut or tree nut, without much evidence to anchor that opinion. The AAP and the AAAAI/ACAAI have no present formal recommendations on this subject.
Frazier and colleagues have published a nested longitudinal study from the Nurses Health Study II. A total of 8,205 mothers reported their perinatal dietary intake actively during delivery, or upon recall shortly after delivery between 1990-1994. In 2006, this cohort was then examined for the prevalence of food allergy by questionnaire. Two physicians were responsible for assessing allergy, through record review, among the cohort. Of 8,205 children, 140 cases of “peanut/tree nut allergy were identified by the pannel (58 “probable” cases and 82 “possible” cases). Among non-food allergic mothers, higher rates of monthly peanut/tree nut consumption (4 quartiles ranging from >5 servings/week to <1 serving/month) were associated with significantly lower odds (69 times) of the reported development of peanut/tree nut allergy. When analyzing only the 58 probable cases, the results were unchanged. There was no effect of consumption among mothers with reported peanut/tree nut allergy (rating tree nut consumption if peanut allergic, and vice versa).
So, what can we take away from this study? This is a retrospective study, which cannot determine a cause-and-effect relationship, so no recommendations or changes should be made based on this study. This study is certainly innovative, and may be our best evidence to date that supports frequent, early exposure of high-risk allergen. However, there are several major limitations to the validity of the findings.
First, the study was not originally designed to evaluate the effect of peanut intake on the development of peanut allergy in the child. This was something crafted after the fact.
Second, the definition of food allergy used in the study is problematic. Though chart review of physician notes was used when available, it was not specified if these were from an allergist or non-allergist. There is no way to account for variability (or level of expertise) in those reports, or the extent of overreliance on allergen testing in making the diagnosis. Without using food challenge as a gold standard, there cannot be 100% certainty of a diagnosis of food allergy in any of the cases.
Third, there are significant recall biases. Less than 50% of the participants filled out their diet diaries during pregnancy, and only 76% within a year of delivery. Thus, it is difficult to have complete confidence in the accuracy and certainty of peanut or tree nut consumption per week/month. Moreover, information about food allergy was asked well into childhood, not on an ongoing regular basis, also introducing a source of recall bias.
Lastly, there was poor consideration of the effects of timing of peanut/tree nut into the diet. This was controlled for as a dichotomous variable in supplemental table 2 (introduction before or after a year), but that is a very arbitrary line to draw, without theoretical significance, accounting for the importance of the frequency of exposure thereafter, or if earlier or later times of introduction had potential significance.
I do not raise these issues to be over critical of a very thoughtful, interesting and innovative study. The authors actually are quite forthcoming of these limitations, some of which were unavoidable and are par for the course in large cohort studies followed over many years. This study is positioned as hypothesis generating, and the group makes no claims that increased consumption has a causal relationship with preventing the development of allergy.
I raise these concerns because this is a highly controversial and sensitive topic, and to make sure that all health care providers who read this study truly understand that the findings do not recommend that pregnant mothers should frequently eat peanut to prevent peanut allergy in their child. We just do not know if this is the best policy and the field needs others to follow the lead, and develop studies that can better answer the question, with as little bias as possible. These data certainly offer promise for the better, that perhaps the answer is not in blanket allergen avoidance during pregnancy. I know that would be a relief to many pregnant mothers.
Matthew J. Greenhawt, MD, MBA, MSc, FAAP
Infectious Diseases in Children Editorial Board
Disclosures: Greenhawt reports no relevant financial disclosures.