Food for Thought on Prevention and Treatment of Atopic Disease Through Diet

A topic disease appears to be increasing, particularly in “westernized” regions.1, 2 Consequently, the role of diet in preventing and treating atopic disease is an increasing focus of research. This is addressed by 3 studies in this issue of The Journal.3, 4, 5 The influence of infant nutrition on atopy has been debated for decades; in 2000, the American Academy of Pediatrics (AAP) Committee on Nutrition presented suggestions aimed toward high-risk infants (biparental or parent-sibling with a family history of allergy).6 The AAP Committee determined that conclusive studies were not available and that they could not make any definite recommendation, but they did advise exclusive breast-feeding until age 6 months (otherwise a hypoallergenic or possibly a partial hydrolysate formula), with continued breast-feeding to at least age 1 year and avoidance of dairy products to age 1 year, eggs to age 2 years, and peanuts, nuts, and fish to age 3 years. Furthermore, mothers were advised to consider avoiding peanuts during pregnancy and to avoid eggs, cow’s milk, peanuts, fish, and perhaps other foods while nursing.

European expert panels, also considering atopy prevention directed toward at-risk infants, did not recommend any particular maternal diet or specific foods to avoid.7, 8 In 2004, a European expert panel7 recommended that all infants be breast-fed preferably to age 6 months (at least to age 4 months), and in situations where the infant is high risk for atopy and breast-feeding is not possible, an extensive (or perhaps partial) hydrolysate with a documented effect on allergy prevention be given in the first 4 months. The UK Department of Health recommended that families with allergy risks avoid peanuts during pregnancy and lactation, and not introduce peanuts until the child is 3 years old.

Virtually every aspect of the aforementioned recommendations is controversial, because of a rather shaky evidence base. Although there is parity among expert opinion and data from meta-analyses regarding a relative prevention or delay of atopic dermatitis, cow’s milk allergy, and possibly asthma by breast-feeding relative to feeding with whole protein-based (milk or soy) infant formula,6, 7, 8, 9, 10, 11, 12 some individual studies implicate breast-feeding as a risk.11 Human breast milk is a complex mixture of dietary nutrients, proteins, and immune factors that can vary based on maternal diet or genetics and influence atopy outcomes. Maternal factors identified as possible risks for increasing atopic disease for breast-fed infants include maternal atopic status13 and asthma.14

In this issue of The Journal, Snijders et al3 report a study evaluating maternal asthma and allergic status as effect modifiers on the duration of breast-feeding with regard to atopy outcomes in a Dutch birth cohort. They found that longer duration of breast-feeding was significantly (P = .01) associated with a lower risk of eczema (by questionnaire) in mothers without allergy (without elevated total or allergen-specific IgE) or asthma, trended so (P = .14) in mothers with allergy but without asthma, and was not associated (P = .87) in mothers with asthma. In addition, longer duration of breast-feeding was more protective of recurrent wheezing, unrelated to maternal asthma or atopy status (attributed by the authors to reduced respiratory infection) and, unlike in a previous study,13 no effect modification of maternal allergy/asthma on infant IgE or allergen sensitization was seen. The authors found a weaker protective effect for atopic families, in contrast to the stronger protective effect noted in meta-analyses9, 10 and reflected in recommendations6, 7 aimed toward infants considered at risk for atopy (although these generally focused on studies in which the independent variable was breast-feeding of at least 3 months’ duration). Among several caveats to the study, there was no effect on atopic dermatitis (using a stricter definition); only several subgroups presented adjusted odds ratios (ORs) in which 95% confidence intervals (CIs) did not straddle 1.0; for example, for effect on infant eczema, only breast-feeding for more than 9 months was effective (OR = 0.51; 95% CI = 0.29 to 0.89), and there were potential confounders (eg, maternal/infant diet) with possible subtle interacting effects. Nonetheless, the overall findings remain in favor of breast-feeding, with notation that studies need to consider effect modifiers.

Regarding the influence of timing of introduction of solid foods on atopy, Filipiak et al4 in this issue of The Journal, present data from the German Infant Nutritional Intervention Program (GINI) supporting the notion that it is unnecessary to delay solid foods beyond the fourth month or allergenic solids beyond the sixth month to prevent eczema. The GINI study was initiated to compare the allergy-preventing effects of 3 hydrolyzed formulas with standard cow’s milk formula as the only substitutes for breast-feeding in the first 4 months of life in infants at risk for atopic disease. The study identified an intervention group (n = 1939) that was likely (based on family history) to have an atopic child and a nonintervention group (n = 2814) without a strong history of allergy, and included those not interested in participating in the intervention. The results generally showed no significant effect of timing or diversity of solid foods on eczema outcomes to age 4 years. The authors observed a few food-specific associations (eg, higher rate of physician-diagnosed eczema in the nonintervention group when egg was introduced after age 1 year; OR = 1.8), but advised caution in the interpretation of outcomes for specific foods because of possible disease-modifying interactions (eg, the use of soy early in children with eczema). In addition, the study could not control for reverse causation (ie, altering the diet because of observed disease), because initial data collection was done at age 1, after eczema likely would have developed. However, another German study that did adjust for reverse causation arrived at the same general conclusion.15 Recent reviews generally concur with the notion that early (eg, before age 4 months) introduction of solid foods is a risk factor for eczema and possibly milk allergy,16, 17 although the affect on other forms of atopy and the role of delaying foods that are associated with atopy (eg, eggs, milk, peanuts) remains uncertain.

It should be appreciated that the duration of exclusive breast-feeding is inexorably tied to the timing of introduction of solid foods; importantly, formula made with whole protein cow’s milk or soy also represent exposure to a nonphysiological antigen, similar to solid foods. Indeed, the GINI study outcomes at age 3 years indicated that compared with cow’s milk formula, extensively hydrolyzed casein formula was associated with a significantly lower incidence of atopic dermatitis in intention-to-treat (ITT; n = 1363) and per protocol (PP; n = 904) analyses (ITT: population OR = 0.67, 95% CI = 0.45 to 0.99; PP: adjusted OR = 0.53, 95% CI = 0.32 to 0.88), as was a specific partially hydrolyzed whey formula in the PP analysis (ITT: OR = 0.76, 95% CI = 0.52 to 1.11; PP: adjusted OR = 0.60, 95% CI = 0.37 to 0.97).18 None of the formulas reduced the incidence of asthma. These findings indicate that there may be a window of immunologic immaturity during which whole protein in large amounts, whether solid or liquid, may induce atopic disease.

What about the diet in infants who already exhibit atopic dermatitis? In this issue of The Journal, Hill et al5 report that 90% of 51 consecutive infants presenting with moderate-severe atopic dermatitis to a dermatology clinic had test results demonstrating likely clinical allergy to eggs, milk, and/or peanuts (based on previous studies correlating test results with oral food challenge outcomes). The rate of food allergy in these infants was higher than the rate of 37% found in a similar study of older children.19 Most (94%) had been breast-fed, and 25 of the 29 who were exclusively breast-fed for longer than 4 months developed atopic dermatitis while breast-feeding. Of the 37 children with likely egg allergy, only 8 had been exposed, and all showed evidence of allergy: 2 with urticaria, 3 with eczema flare, 2 with vomiting, and 1 with eczema that cleared on an egg-free diet. Although the role of food allergy in atopic dermatitis remains a matter of debate despite extensive data demonstrating a relationship20, 21 (and further supported by the present study), the Hill et al study also raises questions about sensitization to allergens through breast-feeding or during pregnancy. Reviews of available studies on atopy prevention generally conclude that there is no strong evidence indicating that avoidance of major allergens during pregnancy or lactation has a protective effect,7, 22 although reduction of atopic dermatitis cannot be ruled out. Keep in mind that there are nutritional risks of multiple allergen avoidance, however. It should be appreciated that the infants in the Hill et al study already manifested atopic disease, and strategies that might be considered primary prevention were no longer applicable; rather, treatment (ie, medications, testing, and dietary management) was necessary.

The 3 studies reported in this issue of The Journal focus on diet, but many other variables can influence atopy outcomes as well. Even though US and UK agencies have recommended to avoid peanuts during pregnancy and lactation and in the atopy-prone child’s diet until age 3, the prevalence of peanut allergy has apparently doubled in recent decades.2, 23 Of course, how well these recommendations have been followed is unclear,24 and this observation does not imply cause and effect; however, it does indicate that food avoidance may not be the entire answer. The influence of allergens in the diet during pregnancy or lactation remains controversial.25, 26, 27 Noningestion exposure (ie, skin contact, inhalation) may be another source of sensitization.25 Numerous genetic, immunologic, and environmental influences aside from allergen content in the diet conspire to influence atopy outcomes, including: dietary components (eg, vitamins, fats), food additives, antibiotics, infection, antacids, and endotoxin exposure. There has been an apparent increase in sensitization to nonfood allergens in westernized countries in recent decades,1 and unless mothers are ingesting more cats, molds, and tree pollens during pregnancy and lactation, it is hard to blame their diet for this increase.

The 3 studies provide insight into and raise questions about atopy prevention and treatment. It may be envisioned that in the future, a genetic–immunologic profile may be needed to identify specific sensitizing or tolerizing dietary factors (ie, timing, dose, and frequency) for an individual child. Today, it seems reasonable to recommend exclusive breast-feeding for the first months of life to all mothers, with the knowledge that most studies indicate an atopy-protective effect compared with a diet of whole proteins. Studies of substitute infant formulas have focused primarily on families at risk for atopic disease and have shown a relatively protective effect compared with whole cow’s milk protein for extensively or partially hydrolyzed formulas in atopy-prone infants, an effect modulated in part by specific atopy risk factors.7, 18 It is unclear whether there is a preventive effect of altering the maternal diet to exclude allergens during breast-feeding or pregnancy, an approach typically considered for an at-risk population, but in the face of established significant atopic disease in infants, it appears prudent to consider the role of maternal diet for treatment purposes.5, 28

Beyond the apparent risk reduction associated with a period of exclusive breast-feeding or selection of specific infant formulas, the affect of the timing of introduction of whole proteins or the choice of proteins is an evolving story.3, 7, 15, 25, 29 For example, studies are underway in the UK to evaluate the role of early rather than delayed introduction of peanuts to prevent peanut allergy.

Given that we currently lack sufficient evidence-based advice on atopy prevention with regard to maternal diet during pregnancy and lactation and timing of whole proteins or specific allergenic proteins, we should reassure families with an atopic child that they need not feel guilty that they caused the allergy by following or not following the advice of various expert panels. In the meantime, studies such as those reviewed here continue to increase our appreciation of the complex relationship of diet and atopy.