Introduction of Complementary Foods to Infants

• There is level 1 evidence that early introduction of peanuts (from 4 to 11 months of age) reduces the prevalence of peanut allergy in infants at high risk of allergic disease (infants with severe eczema and/or egg allergy).

• The majority of current international guidelines recommend that complementary foods, including allergenic foods, can be introduced from 4 to 6 months of age irrespective of family history risk.

• As delayed peanut introduction may increase the risk of peanut allergy, interim guidelines state that healthcare providers should recommend introducing peanut-containing products into the diets of “high-risk” infants early on in life (between 4 and 11 months of age) in countries where peanut allergy is prevalent.

Over the past few decades, we have experienced a rising prevalence of Immunoglobulin E (IgE)-mediated food allergies in the pediatric setting, particularly in developed countries, although the prevalence also appears to be rising in developing countries [1]. Most reports are based on self-reported food allergy, however, and it has been repeatedly shown that self-reported data will overestimate the prevalence as compared with evaluation by an oral food challenge [2, 3, 4]. It is estimated that IgE-mediated food allergy affects approximately 6-8% of children in developed countries [1, 3, 4], thereby posing a significant burden on the afflicted children, their families, and the healthcare system. Globally, the by far most common IgE-mediated food allergies in childhood are allergies to cow's milk, hen's egg, soy, peanut, tree nuts, wheat, fish, and seafood [3, 4, 5]. Tolerance development is prevalent in milk and egg allergy; and the majority of milk-allergic children [6, 7] and about a half to two-thirds of egg-allergic children [8, 9] will outgrow their food allergy before school age. The rate of peanut allergy resolution is worse; when assessed by oral food challenges both at diagnosis and at follow-up in the Australian HealthNuts cohort study, only 22% of the children outgrew their peanut allergy by 4 years of age [10]. Collectively, a significant proportion of children will remain food allergic and are at risk of developing other comorbidities such as respiratory allergic disease. For instance, infant egg allergy, particularly when coexisting with eczema, has been reported to be a predictor of later respiratory allergies [11], and high levels of IgE antibodies to cow's milk, egg white, wheat, and soy are predictors of persistent food allergy [12].

Avoidance remains the only available treatment in established food allergy. Oral immune therapy, which includes a stepwise dose increase of the food allergen followed by a maintenance phase, is an emerging treatment option. Oral immune therapy has been demonstrated to induce desensitization, i.e., an increase in the amount of offending food that can be ingested as long as it is consumed regularly [13]. It is still undecided if permanent tolerance will develop, and oral immune therapy is not generally recommended unless within a clinical trials protocol. Adherence to an elimination diet is difficult, and there is still risk of accidental exposure and allergic reactions, including anaphylaxis [14]. Allergic children on elimination diets are also at risk of nutritional deficiencies [15, 16], impaired growth [16, 17, 18], and reduced quality of life [19]. Collectively, there is urgent need to develop effective strategies to promote tolerance development and prevent food allergy.

While earlier food allergy prevention strategies implemented food avoidance in early infancy, the current paradigm is shifting from avoidance to controlled exposure. The collective evidence from epidemiological studies reporting an association between delayed introduction of complementary foods and allergy risk, and animal models demonstrating that oral tolerance induction is driven by exposure to antigens and allergens [reviewed in [20, 21], led to the first randomized controlled trials (RCTs) to examine the role of early, regular exposure to “allergenic” foods for food allergy prevention. This review focuses on the outcome of these recently published RCTs and discusses the implementation of the results in clinical practice.

Both genetic and environmental factors will influence the risk of developing food allergy, and multifaceted changes in our modern environment are a likely driver. The hypotheses proposed to explain the epidemic rise in allergic disease include (a) the biodiversity hypothesis, which theorizes that reduced diversity and intensity of microbial exposures will impair normal development of immunoregulatory networks and increase allergy risk [22], (b) the vitamin D hypothesis that builds on epidemiological evidence that vitamin D deficiency is associated with an increased risk of allergic disease, and (c) the dual-barrier hypothesis [23, 24], which is discussed below. There are also data to suggest that food allergens, specific nutrients, lifestyle factors, and microbial exposures may influence the development of allergic disease through epigenetic mechanisms [25].

A commonly used definition of allergy risk is based on a history of allergic disease in a first-degree relative [26] and is frequently used in both epidemiological studies and clinical trials. In some of the recently conducted RCTs, however, only infants with an already established allergic phenotype (eczema and/or manifest egg allergy) were included as they are at an even heightened risk [27, 28] (Table 1). For instance, it has been demonstrated that infant eczema is associated with an increased risk of percutaneous sensitization to environmental food allergens, facilitated by an impaired skin barrier [29]. Normally, a food allergen is introduced to and handled by the immune system in the gut to induce a tolerogenic response to the food protein [20, 30]. Accordingly, the dual-barrier hypothesis theorizes that avoidance of a specific food (such as egg or peanut) can increase the risk of developing food allergy if the infant is still exposed to the food allergen in the environment and is percutaneously sensitized [24].

Almost 2 decades ago, the American Academy of Pediatrics Committee on Nutrition launched guidelines suggesting a delayed introduction of dairy products in the first year of life in infants with a family history of allergic disease: egg until 2 years, peanuts, nuts, and fish until 2-3 years of age [31]. This recommendation also became integrated in infant feeding guidelines in many other countries at the time. Following the publication of more recent epidemiological studies across the globe, the guidelines were revised to reflect the lack of solid scientific evidence that delayed introduction of complementary foods beyond 4-6 months of age, or avoidance of “allergenic” foods such as cow's milk, egg, peanuts, tree nuts, fish, and seafood, reduce allergy risk [32, 33, 34, 35, 36]. Still, the “optimal” time for introduction of complementary food for allergy prevention is not known. There are data to suggest that starting complementary foods before 3-4 months of age may increase the risk of allergic disease [37, 38]. At that age, the gut is more permeable and gastrointestinal colonization is not yet well established, which might contribute to the observed risk increase [39, 40]. Consequently, many international infant feeding guidelines for allergy prevention recommend introduction of any solid food after 4 months of age [32, 33, 34, 35, 36].

In a cross-sectional study, du Toit et al. [41] found that the prevalence of peanut allergy was 10-fold higher in Jewish children in the UK compared with children in Israel. Interestingly, peanut consumption was initiated earlier and in larger quantities in Israel than in the UK. Based on these findings, the Learning Early About Peanut Allergy (LEAP) study was designed to examine if early, regular, controlled peanut consumption, compared with avoidance, could prevent peanut allergy in high-risk infants with severe eczema, egg allergy, or both [42] (Table 1). As part of the screening, a skin prick test (SPT) to peanut was performed, and infants with a wheal size ≥5 mm were excluded. The intervention was initiated between 4 and 11 months and continued until 5 years of age. The study results were pronounced; in the early-introduction group, peanut allergy was reduced with 86% in the group with a negative SPT to peanut at baseline, and with 70% in the group with SPT peanut 1-4 mm at baseline, compared with the avoidance group. Reassuringly, the investigators recently reported that 12-month peanut avoidance in the early-introduction group did not increase the prevalence of peanut allergy at the age of 6 years [43], suggesting that avoidance for a prolonged period will not break tolerance. Still, the long-term consequences of peanut avoidance beyond 12 months are unknown.

There is also epidemiological evidence to support that delayed introduction of egg to the infant diet increases allergy risk. In the HealthNuts cohort study, delayed introduction of egg at 10-12 months or after 12 months of age was associated with an increased risk of egg allergy compared with egg introduction at 4-6 months of age [44]. To date, 5 RCTs have examined if early versus late introduction of egg can reduce the risk of egg allergy [45, 46, 47, 48, 49] (Table 1). In the Solids Timing for Allergy Reduction (STAR) study, high-risk infants with moderate-to-severe eczema were randomized to intake of pasteurized raw whole egg powder or rice powder (placebo) from 4 to 8 months of age [45]. At 12 months of age, 33% in the active group versus 51% in the placebo group had developed egg allergy (relative risk 0.65, 95% CI 0.38-1.11, p = 0.11). In the Starting Time of Egg Protein (STEP) study, high-risk infants (based on maternal atopy but no allergic manifestation in the infant at baseline) were randomized to intake of pasteurized raw whole egg powder or rice powder (placebo) from 4 to 10 months of age [46]. At 12 months of age, 7% in the active group versus 10.3% in the placebo group had egg allergy (adjusted relative risk 0.75, 95% CI 0.48-1.17, p = 0.20). The Beating Egg Allergy Trial (BEAT) also included high-risk infants (based on allergic disease in any or both parents) [47]. Infants were randomized to pasteurized raw whole egg or rice powder (placebo) from 4 to 8 months of age. The primary outcome was egg sensitization at 12 months of age, and 11% in the active group versus 20% in the placebo group were sensitized (odds ratio 0.46, 95% CI 0.22-0.95, p = 0.03). Twenty-one infants were classified as having probable egg allergy. Of these, 6.2% were in the active group and 10.5% were in the placebo group (p = 0.20).

In the Prevention of Egg Allergy with Tiny Amount Intake (PETIT) study, high-risk infants with established eczema were randomized to intake of either heated egg powder or squash powder (placebo) from 6 to 12 months of age, with an increased dose of egg protein from 9 months [48]. There was a marked effect of the intervention with egg allergy diagnosed at 12 months in 9% in the active group versus 38% in the placebo group (risk ratio 0.221, 95% CI 0.09-0.543, p = 0.0001). In fact, the striking effect of the intervention in the preplanned interim analyses led the investigators to terminate the trial prematurely. As discussed by the investigators [48] the difference might be biased, leading to a bigger difference between the active and placebo groups than if the study had not been closed.

In contrast to the above-mentioned studies that included high-risk infants, the Hen's Egg Allergy Prevention (HEAP) study, randomized infants with normal risk (from the general population) to intake of pasteurized egg white powder or rice powder from 4 to 6 months until 12 months of age [49]. Infants were screened for egg sensitization, and all included infants had specific IgE to egg <0.35 kU/L at baseline. As in the BEAT study [47], the primary outcome was egg sensitization at 12 months. 5.6% in the active group were sensitized to egg versus 2.6% in the placebo group (relative risk 2.20, 95% CI 0.68-7.14, p = 0.24). At that age, 2.1% in the active group had egg allergy versus 0.6% in the placebo group (relative risk 3.30, 95% CI 0.35-31.32, p = 0.35).

Collectively, 4 out of 5 conducted RCTs designed for egg allergy prevention were negative (Table 1), although 3 of these studies [45, 46, 47] had nonsignificant results that might suggest a benefit of early egg introduction.

Observational studies have also reported an association between low food diversity in early life and both sensitization [50] and allergic manifestations [51]. In the Enquiring About Tolerance (EAT) study [52] (Table 1), 3-month-old breastfed infants from the general population were randomized to continued breastfeeding with introduction of cow's milk, peanut, hard-boiled egg, sesame, white fish, and wheat in a sequential order from 3 months of age or to continued exclusive breastfeeding for the first 6 months of life [52]. In the intention-to-treat analysis, 5.6% of the children in the early-introduction group had developed food allergy at 3 years of age compared to 7.1% in the group that introduced solid foods from the age of 6 months (relative risk 0.80, 95% CI 0.51-1.25, p = 0.32). Of note, only 42% in the early-introduction group were able to adhere to the food introduction regimen, demonstrating that it can be difficult to introduce multiple foods as compared to a single food item. In the per protocol analysis, however, the prevalence of “any” food allergy was 2.4% in the early-introduction group compared with 7.3% in the standard-introduction group (p = 0.01). The prevalence of peanut and egg allergy was also reduced in the early-introduction group (0 vs. 2.5%, p = 0.003, and 1.4 vs. 5.5%, p = 0.009, respectively). There was no difference between the 2 groups in the prevalence of allergy to milk, sesame, fish, or wheat.

In a recent comprehensive systematic review and meta-analysis, Ierodiakonou et al. [53] included the trials discussed above that had assessed early versus late egg introduction for egg allergy prevention (5 trials, 1,915 participants) [45, 46, 47, 48, 49] (Table 1). They found with moderate certainty evidence that early versus late introduction of egg was associated with a reduced egg allergy risk (risk ratio 0.56, 95% CI 0.36-0.87, p = 0.009). They also identified the LEAP [43] and EAT [52] studies (Table 1) (2 trials, 1,550 participants) to be included in a meta-analysis of early versus late introduction of peanuts and reported that early introduction was associated with a reduced peanut allergy risk (risk ratio 0.29, 95% CI 0.11-0.74, p = 0.009). The authors underscored that the studies were few and that the certainty of the evidence was reduced due to imprecision, indirectness, and heterogeneity in interventions and study populations [53]. An interesting finding, however, was that there was no distinct difference among infants at “normal” versus high risk of allergy in their analyses.

A shared feature of the studies that included high-risk infants and used pasteurized raw egg powder is that a varying proportion (4.7-31%) of the participants in the early egg intake groups discontinued egg ingestion due to allergic reactions to the egg powder [45, 46, 47]. This has raised the question if screening for sensitization would be necessary before introducing egg to the infant diet. Reassuringly, in the STEP trial that included infants with familial predisposition but no eczema, there were no anaphylactic reactions to the egg powder [46]. The authors underscored that assessment of egg sensitization before the introduction of egg and egg-containing products to the infant diet is not necessary in the community setting [46]. In the PETIT trial [48], which included high-risk infants with eczema, a few participants reported mild to moderate allergic manifestations following ingestion of the study powder, but at a similar frequency in the active and placebo groups. No participant discontinued the intervention because of allergic reactions to the egg powder, and it has been argued that this could be a matter of reduced allergenicity of heated versus pasteurized (raw) egg powder [48]. The risk of adverse reactions to peanut was low in the LEAP study; 5% of infants randomized to early peanut intake reacted at the baseline peanut challenge. However, infants at a presumably higher risk (peanut wheal size ≥5 mm) were excluded.

Current international guidelines already state that introduction of allergenic foods, including egg and peanut, does not need to be postponed beyond 4-6 months of age [32, 33, 34, 35, 36]. With a few exceptions, these guidelines do not, however, advocate that allergenic foods should be actively introduced to the infant diet between 4-6 months of age. Based on level 1 evidence from the LEAP study [42], interim guidelines on peanut introduction for allergy prevention in high-risk infants were launched in 2015 (Table 2) [54]. In an opinion paper, Allen and Koplin [55] identified and discussed the challenges in translating the findings from the LEAP study to the general population level. Safety remains one issue, particularly in very high-risk infants, as the LEAP study excluded infants with an SPT to peanut ≥5 mm, cost-effectiveness another [55]. Very recently, addendum guidelines for penaut allergy prevention in the United States were launched [56]. In brief, the guideline panel suggests introducing peanuts at home to the majority of infants in the first year of life. Infants with severe eczema, egg allergy, or both should undergo medical assessment including assessment of sensitization to peanut before peanut introduction at 4-6 months of age [56]. If other allergenic foods, such as egg, should also be actively introduced to the infant diet from 4 to 6 months of age remains undetermined. Recent allergy prevention guidelines in Australia now suggest introducing cooked (but not raw) egg from 4 to 6 months of age irrespective of allergic heredity [36]. As underlined by Ierodiakonou et al. [53], the findings from their systematic review on early versus late introduction of complementary foods for allergy prevention cannot be directly translated to new guidelines. Collectively, the optimal timing, doses and form of egg, and if these regimens should be stratified according to the infant's allergy risk remain to be determined.

The level 1 evidence form the LEAP study [43] has resulted in interim guidelines recommending early introduction of peanut into the diets of “high-risk” infants [54]. Further studies should aim at optimizing infant feeding regimens. Supporting the most favorable “tolerogenic” microenvironment in the gut during the period of food allergen introduction is also likely to involve “optimal” colonization of the gastrointestinal tract, breastfeeding, and other dietary factors with immunomodulatory capacity [39, 40].

The author reports no conflict of interest in relation the topic. The writing of this article was supported by Nestlé Nutrition Institute.