Updates in Food Allergy Prevention in Children

The Soriano et al study has highlighted that, although there is a role for early food introduction, feeding early alone may not be sufficient. In examining the sentinel studies on food allergy prevention, a key component of all study protocols was regularity of allergen ingestion, in addition to early food introduction. As 1 example, in the most effective egg prevention randomized controlled trial to date, the PETIT study, 147 infants with eczema were introduced to heated egg powder at 6 months of age (or avoidance until a year of age), but also were required to eat the heated egg powder at least daily, resulting in such a significant protective effect with early introduction (8% in the early introduction had egg allergy compared with 38% in the placebo group) that the trial was halted prematurely.³  Similarly, in the LEAP study, infants in the early introduction group ate peanut at least 3 times a week (6 g per week) until 5 years of age.¹  In keeping with this hypothesis, Soriano et al highlighted in their populational study that, although early peanut introduction had increased dramatically in Australia, <30% of infants were eating peanut 2 or more times per week and >20% had only eaten peanut <5 times.²¹  It is possible that the lack of change in prevalence of peanut allergy that was demonstrated may be partially explained by lack of regularity of peanut ingestion. However, no early introduction study has shown that a specific allergen quantity was necessary for successful early introduction.

Perhaps the best illustration of the potential importance of regularity of ingestion as a means of food allergy prevention stems from the cow’s milk allergy prevention literature. There have been several observational and randomized controlled trials that have consistently demonstrated that delayed ingestion and/or irregularity of ingestion increase the risk of cow’s milk allergy. In a 2010 prospective study of 13 019 general population infants, delayed (after 14 days) and/or irregular (<1 per day) cow’s milk ingestion significantly increased the risk of cow’s milk allergy compared with introduction in the first 14 days of life with regular daily exposure thereafter (odds ratio [OR] 19.3; 95% CI 6.0–62.1).²²  In a case control study of 51 patients with confirmed cow’s milk allergy compared with matched controls, as well as unmatched patients with egg allergy, there was a significantly increased risk of cow’s milk allergy among infants with delayed (>1 month after birth) and/or irregular (<1 per day) cow’s milk exposure (adjusted OR 23.74; 95% CI 5.39–104.52 compared with control, adjusted OR 10.16; 95% CI 2.48–41.64 compared with egg allergy group).²³  In a prospective study of 1992 general population (eg, “standard risk”) infants who were recruited on the basis of parental feeding preference to either exclusive breastfeeding or at least 1 meal of cow’s milk formula per day (with or without breastfeeding) for the first 2 months of life, there was a significantly reduced prevalence of cow’s milk allergy at a year of age among those infants who were regularly exposed to cow’s milk formula (relative risk 29.98, P < .001).²⁴  In a recent randomized controlled trial of early cow’s milk exposure, subgroup analysis of infants who ingested cow’s milk formula in the first 3 days of life found a significantly higher incidence of cow’s milk allergy among any infant in whom cow’s milk formula was discontinued (<1 month, 1–2 months, 3–5 months) compared with continuous ongoing ingestion until 6 months of age (P < .001 for all groups).²⁵ 

There is also emerging, although limited, evidence that the preventive effect of regular ingestion may persist into later childhood, in particular among at-risk children. In a follow-up study of 146 siblings of peanut allergic children (aged 3.4–7.5 years) who had tolerated peanut a median of 2.9 years earlier, the risk of peanut allergy was 0% (95% CI 0–6) among those patients eating peanut at least once monthly, 3% (95% CI 0.5–15) in patients eating peanut less than monthly, and 18% (95% CI 5–48) for children who had not eaten peanut at all.²⁶ 

The Canadian Society of Allergy and Clinical Immunology (CSACI) is the first allergy society internationally that has recently released a statement reiterating the importance of regular ingestion of common allergens, recommending that both early introduction and regular ingestion of age-appropriate amounts of allergens multiple times per month (with a goal of at least once weekly) are likely to be useful in food allergy prevention.²⁷  The CSACI further recommends that, once introduced, single or occasional exposures to an allergen could be detrimental and, if an allergen is not a common component of the diet (and hence regular ingestion not feasible), avoidance may be preferable to intermittent ingestion.

For some common allergens such as egg, the degree of allergenicity can vary with the method of preparation (Fig 1). Egg and milk are heat-labile allergens, where the proteins creating the allergen are mainly the result of 3-dimensional protein folding and can be denatured with increasing temperature (eg, conformational epitopes), whereas with peanut, tree nut, and seed, the allergenic proteins are the result of contiguous linear areas (eg, linear epitopes) and are not heat-labile. There is some evidence that the form in which egg is introduced (eg, baked versus cooked versus raw) may influence its tolerability and effectiveness at food allergy prevention, given that a higher cooking temperature can denature the primarily conformational epitopes, and reduce the allergenicity.¹¹  There are 5 randomized controlled trials on early egg introduction as a means of egg allergy prevention which had very discrepant results with respect to safety and effectiveness, and it has been hypothesized that this is related to the form in which egg is introduced in these studies.¹¹  The most effective study, the previously described PETIT study, used gently heated egg (eg, poached) as its study protocol, a form subjected to a mild degree of heat denaturing. In contrast, the other 4 randomized controlled trials used raw pasteurized powdered egg and demonstrated either no significant protective effect with early egg introduction^28–31  and/or significant safety concerns.^30,31  Pragmatically, it is unlikely that raw egg from a culinary standpoint would be introduced outside a study protocol, and the choice of a raw egg was because of ease of use in a study protocol (crystallized form), but this trend in the literature suggests that the form of allergen itself may influence the effectiveness of early introduction.

There may be a further protective effect based on the way egg is heated (cooked versus baked), although evidence is limited to 1 study. A 2010 population-based cross-sectional study of 2589 infants demonstrated that, among infants with diagnosed egg allergy, in addition to a protective effect with early introduction, first exposure to cooked egg (egg cooked on a stove) reduced the risk of egg allergy compared with first exposure to egg in baked goods (egg baked into goods in the oven) (OR 0.2; 95% CI 0.06–0.71).³²  Joint guidance on prevention through the American Academy of Allergy, Asthma, and Immunology (AAAAI), the American College of Allergy, Asthma, and Immunology (ACAAI), and the CSACI recommends that egg be introduced in cooked forms only, avoiding any raw, pasteurized egg-containing products where possible.¹¹  The British Society of Allergy and Clinical Immunology specifies that, when egg is introduced at ∼6 months of age, it should be introduced in a cooked form (scrambled egg, omelet, soft- or hard-boiled egg).⁷  Further studies on this topic are needed, and it is not known to what degree this applies to the literature regarding other allergens such as cow’s milk.

Largely as a result of a priori decisions made in the LEAP study, the National Institute for Allergy and Infectious Diseases (NIAID) released an addendum guideline in 2017 for the prevention of peanut allergy in the United States, which recommended that infants with LEAP risk criteria (egg allergy and/or severe eczema) be strongly considered for preemptive testing before peanut introduction.¹⁰  The American Academy of Pediatrics (AAP) supports this recommendation, although notes that “it is hoped that the screening process for the infants at highest risk will not be a deterrent or generate ‘screening creep’ for infants not in the high-risk category. Furthermore, these guidelines may be difficult to follow in communities where there is no access to the medical care needed for their implementation.”⁹ 

Although targeted screening was supported by the NIAID in 2017, these recommendations are not in keeping with other international guidelines published since the LEAP study such as the Australasian Society of Clinical Immunology and Allergy, the British Society of Allergy and Clinical Immunology, the AAAAI, the ACAAI, or the CSACI. None of these societies recommend routine preemptive screening in infancy before allergenic solid food introduction (Fig 2).^7,11,33  The CSACI has strongly advocated against screening, noting that screening testing in infants is not recommended, irrespective of level of risk.¹²  Health economic modeling has shown that screening is most likely to overestimate the rate of allergy, leading to cost accumulation because of false-positive testing being considered as a surrogate for allergy.³⁴ 

There are several limitations to screening on a population level before peanut introduction (Table 1). Firstly, although all allergy testing, whether skin prick testing or peanut-specific immunoglobulin E (IgE) testing, is safe, sensitive, and widely available, it is poorly specific and will result in overdiagnosis of peanut allergy.¹⁶  The specificity of skin prick testing and peanut-specific IgE testing is <50%, and most infants with positive allergy tests can tolerate the food of concern when such IgE is identified, meaning the presence of the antibody is not pathognomonic for disease.^35–37  For example, a retrospective chart review of 125 children, of whom 96% had eczema, noted that 80% to 100% of foods which were avoided because of positive allergy testing could be reintroduced into the diet after an oral food challenge.³⁸  Secondly, preemptive screening on a populational level is not feasible, or actually necessary to promote safe early introduction. HealthNuts, an Australian prospective population-based cohort study, demonstrated that screening all infants with early onset eczema and/or egg allergy would require screening 16% of the population, and would still miss 23% of cases of peanut allergy.³⁹  In addition, in this study, 29% of infants would require follow-up because of positive testing. The resource limitations associated with a screening approach, resulting in delays in infant ingestion of peanut pending allergy assessment, could inadvertently negate the benefits of early peanut ingestion supported by the LEAP study, with infants missing the window of opportunity for allergy prevention with early ingestion of peanut. Thirdly, although the AAP cautioned against a potential screening creep, the reality of the increased nondiscriminant testing has emerged as a major concern. In 1 recent real-life study, only 48% of patients screened for peanut sensitization fit the NIAID criteria.⁴⁰  Another post-NIAID report demonstrated a significant increase in the number of nonhigh-risk infants that were inappropriately screened, receiving testing for a median number of 10 foods.⁴¹  Finally, peanut screening has been shown to be poorly cost-effective, and the Soriano et al study has clearly demonstrated that there can be uptake of early peanut ingestion on a population level in the absence of screening.^13,42 

It is also important to highlight that feeding infants common allergens such as peanut (in an age-appropriate way) is safe, and the process to do so should not be overmedicalized. There has never been a fatality on first ingestion of a food in infancy, even in infants at high risk for food allergy.^43,44  In the LEAP study, for example, of those infants randomized to the early introduction group, only 2.2% had a positive oral food challenge (observed ingestion) at baseline; none required epinephrine and symptoms were predominantly cutaneous.¹  Australian data at a population level have noted that <5% of early peanut introduction has resulted in a severe reaction.³⁹ 

There is always a role for shared decision-making, especially in the context of a family who is not comfortable feeding an allergen such as peanut in the absence of screening testing.¹¹  However, the ultimate goal in such high-risk infants is early peanut ingestion, and such early introduction is the only identified measure that reduces the risk of peanut allergy. The associated negative impact peanut allergy has on long-term quality of life can be devastating for some families. Fundamentally, LEAP demonstrated the significant protective effect of early peanut introduction in a screened population, but not that screening was preventive or necessary for safe implementation. Furthermore, screening, because of access to care issues, may result in prolonged delays pending timely assessment, which also may paradoxically increase the burden associated with peanut allergy.

Although the LEAP study and subsequent guidelines have helped forward the narrative that early introduction of allergenic solids is safe and effective at preventing certain food allergies, how to optimally advise and implement this is still uncertain in many areas. These areas include which allergens to focus on as a priority for early introduction, how often to advise allergenic solid foods are fed, how much quantity of allergen to feed, and how long feeding is required for a full preventive effect.

The bulk of the evidence for benefit of early allergenic solid introduction exists for peanut,^1,5  egg,^5,29–32  and cow’s milk.^22–24,45  There are no randomized controlled trials focused exclusively on tree nut, soy, grains, seeds, legumes, finfish, or shellfish (although 2 randomized controlled trials have examined multifood ingestion early in life with discrepant results).^4,46  Some of the lack of data for other allergens may be a pragmatic limitation; given that some investigators consider the effects of early introduction to likely generalize across allergens, further trials with groups randomized to avoidance or delayed introduction may no longer be ethical. There is no evidence of harm from early feeding of other common allergens (in an age-appropriate way), and the mechanism of sensitization is thought to be similar for all common allergens.¹²  There is also some observational evidence that dietary diversity early in life may help in the prevention of food allergy.^11,47,48  Guidance on the prevention of food allergy endorsed by the AAAAI, ACAAI, and CSACI recommends specifically egg and peanut introduction at ∼6 but not before 4 months of life, but notes no evidence of harm with introduction of other allergens in this time interval and recommends no “deliberate delay” for the introduction of other potentially allergenic complementary foods.¹¹  The AAP focuses specifically on early peanut introduction because the most conclusive data were available for peanut, but notes no evidence that delaying introduction of other common allergens prevents atopic disease.⁹ 

In terms of frequency and duration of feeding, there is little evidence to inform any concrete recommendations other than that regularity of ingestion appears to play some role, but it may not be the sole factor. A per-protocol secondary analysis of the Inquiring About Tolerance study, a randomized controlled trial of early (3 months) versus standard (6 months) introduction of 6 common allergens, suggested that a dose of ∼2 g of peanut protein and egg white protein per week (∼1 boiled egg and 1.5 tsp of peanut butter) was sufficient for maintenance of tolerance, although further studies are required.^3,46  However, a recently published, multicenter, cluster-randomized trial of early (3 months) versus standard introduction of milk, egg, wheat, and peanut found a significant protective effect with early introduction, with no specific dosing requirements in the study (pragmatic design).⁴  AAAAI, ACAAI, and CSACI guidance notes “insufficient evidence to support a precise dose and frequency necessary to support tolerance,” recommending feeding amounts and types of allergens that the child enjoys in an age-appropriate way with some regular frequency. Similarly, the duration of ingestion required to maintain tolerance is unknown, although a follow-up to the LEAP study, the LEAP-On study, demonstrated that ongoing regular ingestion until 5 years of age was protective against development of peanut allergy in children who then underwent a full year of subsequent avoidance.⁴⁹  Although further studies are required, this study does suggest that regular ingestion through toddlerhood can help augment long-term protection, at least for peanut, though it remains unclear if such augmentation is truly necessary.

To effectively implement widespread early introduction of allergenic foods to all infants on a population level, it will require buy-in from caregivers, primary care pediatricians, professional and advocacy organizations, and allergists/immunologists. Caregivers can lose confidence when guidelines change, and particularly when new recommendations contradict previous advice.⁵⁰  For almost 20 years before these new recommendations, parents were specifically told to avoid giving their infants any of the “top 8” allergenic foods.⁵¹  This was on the basis of expert opinion at the time and not evidence or studies demonstrating protection through avoidance. However, to now implement a paradigm shift that contradicts previous advice, clinicians and guidelines need to address why the advice has changed, why we can trust this new approach, and why it’s important to consider. This requires humility, proactive discussion, and time to address parental concerns (Tables 2 and 3). A survey of 2000 soon-to-be or current parents of infants was conducted 1 year after the NIAID addendum guidelines were published, and only 31% of respondents were willing to introduce peanut before 6 months of age.¹⁷  Although current parental attitudes toward early introduction have not been formally studied in recent years, this still warrants time and explanation to families during individual patient encounters. As addressed earlier, the current inconsistency in screening recommendations across various guidelines is only perpetuating the confusion regarding food allergy prevention.

Consistent positive messaging surrounding the safety and benefits of introducing allergenic foods during early infancy is important. This requires clinicians to understand the evidence, commit to proactively discussing during patient encounters, and incorporate this within existing time constraints on patient care. Some practical advice includes not rubbing food on the skin before feeding, reviewing time to onset of food reactions (1–2 hours), discussing other common childhood conditions that wax and wane with food introduction (such as constipation), and offering to be available for follow-up questions (Table 2). Little data exist on how well this is being done by clinicians, but surveys suggest ongoing hesitancy and need for further education.¹⁹  Incorporation into the electronic medical record is 1 method to help standardize and increase the consistency of these conversations. Australia adopted widespread public health messaging surrounding food allergy prevention and has demonstrated increased acceptance and introduction of peanut over the past few years.²¹ 

Soon after the NIAID addendum guidelines were published in 2017, various companies started producing commercial products containing multiple allergenic foods in palatable forms for infants, such as powders, puffs, cereals, and cookies.⁵²  These commercial products are marketed directly to consumers and also pediatricians in an effort to have them recommend to families. Aside from the cost associated with these products, there is significant inter- and intraproduct variability in regard to the amount of protein included for each allergen.^53,54  In addition, none of these products have evidence demonstrating that they can prevent food allergy development through their use. With these marketed across the world, caregivers may be led to believe that these commercial products are necessary to prevent food allergy, or that they are safer than giving actual food to their infants. This adds a layer of confusion and mixed messaging that parents have to navigate as they try to understand and incorporate food allergens into their infant’s diet.

It may seem like an insurmountable task, but these challenges can hopefully be overcome through dedicated and consistent effort across multiple levels.

Although early introduction has been demonstrated to be a highly effective intervention in the prevention of food allergy, it may not be enough. There may be a role for ongoing regularity of ingestion in the prevention of food allergy, and in fact, regularity of ingestion may play as significant a role as timing of introduction. For some allergens such as egg, the form in which the allergen is introduced may play a role. There is still much to learn about the practicalities of early feeding, although guidance largely applies to peanut, and potentially egg and cow’s milk at this time. There are significant harms to a preemptive screening approach for any common allergen, and in general, testing before food introduction is not recommended.

AAAAI

American Academy of Allergy, Asthma, and Immunology

AAP

American Academy of Pediatrics

ACAAI

American College of Allergy, Asthma, and Immunology

CI

confidence interval

CSACI

Canadian Society of Allergy and Clinical Immunology

IgE

immunoglobulin E

LEAP

Learning Early About Peanut Study

NIAID

National Institutes of Allergy and Infectious Diseases

OR

odds ratio