- Email: [email protected]
Soy consumption is not a risk factor for peanut sensitization
Soy consumption is not a risk factor for peanut sensitization Jennifer Koplin, BSc (Hons),a,b Shyamali C. Dharmage, MBBS, MD, PhD,c Lyle Gurrin, PhD,c Nicholas Osborne, PhD,a,b Mimi L. K. Tang, FRACP, FRCPA, PhD,a,b,d Adrian J. Lowe, BBSc (Hons), MPH,c Cliff Hosking, MD, FRACP,e David Hill, FRACP,a and Katrina J. Allen, FRACP, PhDa,b,d Melbourne and Newcastle, Australia Background: A recent cohort study suggested that intake of soy milk or soy formula was associated with peanut allergy. If this finding is confirmed, it suggests an avenue for modification of diet as a peanut allergy prevention strategy. Objective: To investigate the relationship between soy consumption and peanut sensitization in a prospective cohort study of children. Methods: A total of 620 babies with a family history of allergic disease were recruited. Dietary information was obtained from telephone interviews every 4 weeks from birth until 15 months and then again at 18 months and 2 years. Skin prick tests to peanut, milk, and egg were performed at 6, 12, and 24 months. A wheal size $3 mm was considered positive for sensitization. Results: Children whose parents elected to introduce soy formula or soy milk into their children’s diet were more likely to be sensitized to peanuts at 2 years (odds ratio, 2.02; 95% CI, 1.04-3.92; P 5 .039). However, this relationship was explained by feeding of soy to children who had siblings with milk allergy or were themselves sensitized to milk. After adjusting for these factors, there was no evidence of an association between soy consumption and peanut sensitization (odds ratio, 1.34; 95% CI, 0.64-2.79; P 5 .434). Conclusion: The association between soy consumption and peanut sensitization is not causal but merely a result of preferential use of soy milk in infants with a personal or family From athe Murdoch Children’s Research Institute, Melbourne; bthe Department of Pediatrics and cthe Center for Molecular, Environmental, Genetic, and Analytic Epidemiology, University of Melbourne; dthe Department of Allergy and Immunology, Royal Children’s Hospital, Melbourne; and eJohn Hunter Children’s Hospital, Newcastle. Funding for the initial development of the Melbourne Atopy Cohort Study was provided by Nestle´ Australia. J.K. is a recipient of an Australian Postgraduate Award scholarship. K.J.A., S.D., and L.G. are recipients of National Health Medical Research Council Career Development Awards. A.L. is supported by Dairy Australia, Cooperative Research Center for Asthma, and VicHealth. Funding for analysis was also provided by the Ilhan Food Allergy Research Foundation. Disclosure of potential conflict of interest: S. Dharmage has received research support from the National Health Medical Research Council, Australia Research Council, and Ilhan Food Allergy Foundation. J. Koplin has received research support from the Ilhan Food Allergy Foundation. L. Gurrin has received research support from National Health Medical Research Council. N. Osborne has received research support from the Ilhan Food Allergy Foundation. M. L. K. Tang was on the advisory board for Nestle´ Nutrition Institute, made a presentation at a seminar for Wyeth, and has received research support from Numico, Diroform, and Commonwealth Serum Laboratories. C. Hosking is employed by Melbourne University. D. Hill is on the speakers’ bureau for Nutricia International. K. J. Allen has received research support from Ilhan Food Allergy Foundation and the National Health Medical Research Council. The rest of the authors have declared that they have no conflict of interest. Received for publication December 19, 2007; revised February 14, 2008; accepted for publication March 20, 2008. Available online April 25, 2008. Reprint requests: Shyamali C. Dharmage, MBBS, MD, PhD, Center for Molecular, Environmental, Genetic and Analytic Epidemiology, School of Population Health, Faculty of Medicine, Dentistry and Health Sciences, University of Melbourne, Level 2, 723 Swanston Street, Carlton VIC 3053 Australia. E-mail: [email protected]. 0091-6749/$34.00 Ó 2008 American Academy of Allergy, Asthma & Immunology doi:10.1016/j.jaci.2008.03.017
history of cow’s milk allergy. Future studies should take the confounding effects related to dietary modifications by parents into account when investigating the association between diet and childhood allergic diseases. (J Allergy Clin Immunol 2008;121:1455-9.) Key words: Skin prick test, peanut allergy, soy formula, soy milk, risk factors, sensitization
The potential severity of peanut allergy clinical reactions in combination with the apparent rise in prevalence1,2 make it an important public health issue. Research into potential environmental risk factors for peanut allergy aims to identify modifiable risk factors to help inform guidelines on preventative strategies. A recent large cohort study by Lack et al3 found that intake of soy milk or soy formula in the first 2 years of life was associated with peanut allergy. This is an important finding that is frequently referenced,4-6 although there are as yet no further confirmatory studies. The authors suggested that ingestion of soy could cause peanut allergy because of cross-sensitization through common epitopes shared by peanuts and soy.3 However, this appears unlikely because of the low prevalence of clinical reactivity to soy in children with peanut allergy.4 We hypothesized that the relationship between soy consumption and peanut allergy observed by Lack et al3 could be a result of dietary modifications by parents in response to early signs of atopic disease in infants. The objective of the current study was to examine whether the association between soy consumption and peanut allergy can be explained by selective introduction of soy to children who are at higher risk of peanut sensitization because of a family or personal history of cow’s milk allergy. To address this question, we have used data from the Melbourne Atopy Cohort Study (MACS). This study monitored the development of sensitization and symptoms of allergic disease in 620 infants randomly allocated to receive either soy formula, standard cow’s milk formula, or a partially hydrolyzed cow’s milk–based formula after cessation of exclusive breastfeeding. As part of the MACS, signs of allergic disease and sensitization to peanut, egg, and cow’s milk were recorded at frequent intervals during the first 2 years of life. Furthermore, infant dietary guidelines for the first year of life were clearly and prospectively defined and history of dietary intake recorded on a regular basis for the first 2 years of life. For these reasons, the MACS is particularly well suited to assess the association between soy consumption and the development of peanut sensitization.
METHODS Subjects The subjects are members of the MACS, a randomized controlled trial of the effect of 3 infant formulas on allergy development in a cohort of children 1455
1456 KOPLIN ET AL
Abbreviations used MACS: Melbourne Atopy Cohort Study OR: Odds ratio SPT: Skin prick test
with a family history of allergy. The MACS involved 620 infants born between February 1990 and November 1994 whose mothers were recruited during pregnancy. Infants were eligible for recruitment if their mother attended the Mercy Hospital Antenatal Clinic and at least 1 parent or sibling had atopic dermatitis, asthma, hay fever, or food allergy. The recruitment process has been described in more detail previously.7,8 The MACS was approved by the Human Research Ethics Committee of the Mercy Maternity Hospital, Melbourne.
Randomly allocated study formulas Infants were randomly allocated to 1 of 3 study formulas at study entry, and mothers were instructed to use these formulas on cessation of exclusive breastfeeding. The 3 formulas were a standard cow’s milk–based formula (NAN; Nestle´, Sydney, Australia), a partially hydrolyzed cow’s milk–based formula (NAN HA; Nestle´), and a standard soy-based formula (ProSobee; Mead Johnson, Evansville, Ind).
J ALLERGY CLIN IMMUNOL JUNE 2008
TABLE I. Baseline demographic and clinical characteristics of study cohort (n 5 449)*
Age (y), mean (SD) Mean years of education (SD) Married Owner-occupied residence Current smoker Born in Australia/New Zealand
Maternal (n 5 449)
Paternal (n 5 447)
31.6 (4.2) 13.6 (1.9) 91.5% 83.1% 7.4% 88.0%
33.7 (5.2) 13.5 (2.0) 91.7% 83.6% 17.4% 82.3%
Baseline family history of atopic disease
Asthma Eczema Hay fever Food allergy Milk Egg Other
Maternal (n 5 449)
Paternal (n 5 445)
Any older sibling (n 5 265)
42.9%à 39.7%à 59.6%à
27.0% 21.1% 45.6%
56.6% 63.4% 34.3%
20.3%à 8.0%à 30.1%
9.9% 4.7% 15.5%
42.3% 26.8% 51.7%
*Excludes 171 infants who did not have a SPT at 2 years of age. Definitions were based on self-report by mothers at recruitment into the study during pregnancy. àn 5 448.
Feeding advice Mothers were encouraged to breast-feed and to delay the introduction of solids until after 6 months of age, after which low-allergen solids (ie, avoidance of peanuts, tree nuts, fish, and egg) were recommended until 12 months of age. At 12 months, an appropriate diet for age was recommended, which included the careful introduction of all other foods as tolerated.
Exposure to soy Infants could be exposed to soy protein through the randomly allocated soy-based study formula. In addition, parents of infants in any of the 3 randomly allocated formula groups could elect to give their child a nonrandomly allocated soy formula or soy milk at any stage during the first 2 years of life. These nonrandomly allocated soy formulas and soy milk have been collectively labeled as ‘‘parent-selected’’ soy products.
Questionnaire data The ages of the child’s mother and father and their birthplace, marital status, education, and current residence were recorded by using an interviewer-administered questionnaire during pregnancy. Mothers were also asked whether they or the child’s biological father or any previous children had asthma, hay fever, eczema, milk allergy, egg allergy, or other food allergy. Data are not available on the type of other food allergy, nor were specific questions asked about the presence of peanut allergy. Detailed dietary information, including age at which foods were first introduced and whether the food was tolerated, was obtained from surveys administered by telephone by an allergy-trained nurse every 4 weeks from birth until the age of 15 months and then at 18 months and 2 years of age. These surveys also collected details of any history of allergic disorders or other illnesses experienced by the child.
Definitions of sensitization to food allergens, eczema, and cow’s milk allergy Sensitization to peanut, milk, and egg was determined by skin prick tests (SPTs) at 6 months, 1 year, and 2 years of age. Children attended the 2-year SPT session at a median age of 2 years (range, 1.9-2.6 years). Sensitization was defined as a SPT wheal size 3 mm. Eczema in the first 2 years of life was defined as a parental report of either a doctor diagnosis of eczema or any rash
(excluding rash that affected only the nappy region or scalp) that was treated with topical steroid-based preparations. Parent-reported cow’s milk allergy was defined as a parent reporting that the child developed symptoms either within 2 hours of ingesting cow’s milk or more than 2 hours after ingestion on 2 or more occasions.
Statistical analyses The association between soy consumption and peanut sensitization was examined by using logistic regression. Variables were identified as possible confounding factors if they were associated with both the outcome variable (peanut sensitization) and the exposure variable (soy consumption) with a P value less than .05. Multiple logistic regression was used to obtain odds ratios (ORs) and 95% CIs adjusted for the possible influence of confounding factors. Stata software (release 10; Stata Corp, College Station, Tex) was used for all analyses.
RESULTS Study population Of the initial cohort of 620 children, SPT results were available for 449 (72.4%) at 2 years of age. Only those children who had SPT results available at 2 years of age were included in the analysis. Parents of the children included in this analysis were predominantly Australian-born, married, and owned their own home (Table I). Hay fever was the most commonly reported allergic disease in parents, whereas older siblings were most often reported to have eczema (Table I). The majority of children (72.6%) had more than 1 immediate family member reported as having an allergic disease. Demographic and clinical characteristics of the entire study cohort have been published previously.9 The children who had a SPT at 2 years of age did not differ significantly from those without a SPT at 2 years in terms of parental marital status, home ownership, smoking status, mean years of education, and history of allergic disease or sibling history of allergic disease (data not shown). There were also no significant differences
KOPLIN ET AL 1457
J ALLERGY CLIN IMMUNOL VOLUME 121, NUMBER 6
TABLE II. Percentage of children in each of the allocated study formula groups who received the allocated formula and any soy products in the first 2 years of life Allocated study formula
Received Received Received Received
allocated formula other soy formula soy milk any form of soy formula or soy milk
Soy-based (n 5 145)
Cow’s milk–based (n 5 158)
Partially hydrolyzed cow’s milk–based (n 5 146)
Whole cohort (n 5 449)
84.8% 35.9% 28.3% 89.7%
86.7% 22.8% 19.0% 24.1%
84.9% 23.3% 23.3% 27.4%
85.5% 27.2% 23.4% 46.4%
between these 2 groups in allocation to the soy-based study formula, consumption of other soy formulas or soy milk, presence of siblings, and sensitization to any food allergens at 6 months or 1 year of age (data not shown). The majority of children received the study formula to which they had been randomly allocated at least once in the first 2 years of life (Table II). There were no significant differences in the proportion of children in each of the 3 allocated formula groups who received the study formula. A significant proportion of children also consumed nonrandomly allocated (parent-selected) soy products, with 27.2% of children consuming a parent-selected soy formula and 23.4% consuming soy milk (Table II). In total, 133 children (29.6%) received a parent-selected soy product. Children who were allocated to the soy-based study formula were more likely also to consume parent-selected soy formulas or soy milk, with 38% of these children reporting intake of an additional soy product compared with 26% of the children allocated to either the cow’s milk–based or the hypoallergenic cow’s milk–based study formula who reported intake of at least 1 soy product (P 5 .008). Overall, 208 children (46.4%) were exposed to soy in the first 2 years of life through either the randomly allocated soy formula or parent-selected soy formulas or soy milk (Table II).
Unadjusted relationship between soy consumption and sensitization to peanuts At 2 years of age, 40 children (9.4%) were sensitized to peanut. All children who had been exposed to soy in the first 2 years of life through consumption of the randomly allocated soy formula, a parent-selected soy formula, or soy milk (n 5 208) were initially grouped together. There was no evidence that these children had an increased risk of peanut sensitization compared with children who had never consumed soy formula or soy milk (OR, 1.64; 95% CI, 0.85-3.16; P 5 .141). Exposure to the randomly allocated soy formula was then considered separately from exposure to parent-selected soy products. A logistic regression model was fitted that simultaneously allowed for separate effects of peanut sensitization caused by consumption of the randomly allocated soy formula and by consumption of any parent-selected soy formula or soy milk. There was no evidence of an association between consumption of the randomized soy formula and peanut sensitization (OR, 1.35; 95% CI, 0.67-2.703; P 5.399). By contrast, consumption of a parent-selected soy formula or soy milk was associated with sensitization to peanut at 2 years of age compared with those who had not ingested soy (OR, 2.02; 95% CI, 1.04-3.92; P 5 .039). The relationship between consumption of parent-selected soy products and peanut sensitization was explored further in the following sections.
TABLE III. Factors associated with consumption of parentselected soy formula or soy milk Unadjusted OR (95% CI)
P value
(20.3) (9.8) (24.9)
2.33 (1.55-3.50) 1.45 (0.84-2.51) 2.79 (1.92-4.06)
.001 .183 <.001
(47.4)
0.99 (0.70-1.39)
.939
(4.5) (11.6) (4.7)
5.50 (2.22-13.64) 1.04 (0.59-1.83) 4.39 (2.11-9.13)
No. (%)
Family history Maternal milk allergy* 91 Paternal milk allergy 44 Sibling milk allergy 112 Allergic symptoms/signs in the child Eczema by 2 y 213 Positive SPT at 6 moà Milk 19 Egg 49 Parent-reported cow’s milk allergy 21
<.001 .892 <0.001
*n 5 448. n 5 447. àn 5 424.
Factors associated with intake of parent-selected soy products Children who were sensitized to milk at 6 months of age or experienced parent-reported symptoms of cow’s milk allergy were more likely to be given a parent-selected soy formula or soy milk (Table III). By contrast, there was no evidence of an association between sensitization to egg at 6 months or eczema in the first 2 years of life and soy consumption (Table III). Milk allergy in the child’s mother or siblings was also strongly associated with soy consumption (Table II). Factors associated with sensitization to peanuts Children who had siblings with milk allergy or were sensitized to milk or egg at 6 months of age or had eczema by age 2 years were more likely to be sensitized to peanuts, and these associations remained significant after controlling for sex, parental allergies, maternal and paternal smoking, and duration of breast-feeding (Table IV). Although there was only weak evidence of an association with parent-reported cow’s milk allergy before adjustment, controlling for the factors listed appeared to strengthen the association (Table IV). Maternal milk allergy showed no association with peanut sensitization (Table IV). Adjusted relationship between consumption of parent-selected soy products and sensitization to peanuts Only sensitization to cow’s milk at 6 months of age, parentreported symptoms of cow’s milk allergy, and history of milk
1458 KOPLIN ET AL
J ALLERGY CLIN IMMUNOL JUNE 2008
TABLE IV. Factors associated with peanut sensitization (SPT $ 3 mm) in probands Unadjusted
Family history Maternal milk allergy* Paternal milk allergy Sibling milk allergy Allergic symptoms/signs in the child Eczema by 2 years Positive SPT at 6 monthsà Milk Egg Parent-reported cow’s milk allergy
Adjusted§
No. (%)
Unadjusted OR (95% CI)
91 (20.3) 44 (9.8) 112 (24.9)
0.98 (0.43-2.20) 0.45 (0.11-1.95) 2.74 (1.41-5.33)
.959 .289 .003
213 (47.4)
4.30 (1.99-9.25)
<.001
4.78 (2.15-10.62)
<.001
19 (4.5) 49 (11.6) 21 (4.7)
6.58 (2.42-17.84) 10.68 (5.18-22.02) 2.56 (0.82-8.02)
<.001 <.001 .106
6.44 (2.27-18.25) 10.65 (4.90-23.17) 3.69 (1.05-13.01)
<.001 <.001 .042
P value
Adjusted OR (95% CI)
1.20 (0.51-2.83) 0.31 (0.07-1.41) 2.80 (1.40-5.61)
P value
.684 .129 .004
*n 5 448. n 5 447. àn 5 424. §The variables were adjusted for the infant’s sex, parental history of allergy (asthma, eczema, hay fever and food allergy treated separately), maternal and paternal smoking, and duration of breast-feeding (<6 mo vs 6 mo).
TABLE V. Results of multiple logistic regression for peanut sensitization Peanut sensitization (SPT $ 3 mm) Variable
Consumption of parent-selected soy formula or soy milk Consumption of randomly allocated soy formula Milk allergy in siblings Positive SPT to milk at 6 months Parent-reported cow’s milk allergy
Adjusted OR (95% CI)
P value
1.34 (0.64-2.79)
.434
1.10 (0.52-2.35)
.800
2.25 (1.09-4.64) 4.78 (1.63-13.96) 1.94 (0.56-6.72)
.028 .004 .297
allergy in siblings were associated with both soy consumption and peanut sensitization and emerged as potential confounders of the relationship between soy intake and sensitization to peanuts. These variables were entered into a multiple logistic regression analysis (Table V). After adjustment, the OR for peanut sensitization at 2 years of age in infants who had consumed parent-selected soy products compared with those who had not was 1.34 (95% CI, 0.64-2.79), which represents a 34% reduction from the unadjusted OR. The corresponding P value of .434 and the 95% CI indicate that there is no evidence that soy consumption was associated with sensitization to peanuts after controlling for sensitization to cow’s milk and history of cow’s milk allergy in siblings.
DISCUSSION Our study found no evidence that soy consumption leads to peanut sensitization. Although we observed a positive association between consumption of parent-selected soy products and peanut sensitization, this relationship was explained by the fact that children who had siblings with cow’s milk allergy or who were themselves sensitized to cow’s milk at 6 months of age were more likely to both consume soy and be sensitized to peanuts. Dietary information for our study was collected prospectively at frequent intervals throughout the first 2 years of life, minimizing the chance of recall bias based on allergic status and allowing us to examine the effects of reverse causality. Furthermore, the
accuracy of the data collected was likely to be high because the period for which parents were asked to recall information was short. We were unable to examine clinical reactions to peanuts because introduction of peanut products was not recommended until after 12 months in all infants, and those with positive peanut SPT results at 12 months were told to avoid peanut until 2 years of age. As a result, the majority of children with a positive SPT result at 2 years of age had not ingested peanuts in the first 2 years of life (data not shown). Although not all children who are sensitized to peanut will experience clinical symptoms, these children are at higher risk of IgE-mediated peanut allergy.10 In addition, children with a SPT wheal size <3 mm are highly unlikely to have IgEmediated peanut allergy.10 Because all children in the study had a family history of atopy, we could not assess the affect of soy consumption on peanut sensitization in children without a family history of allergy. These results therefore only pertain to individuals with a family history of allergy. A potential limitation of our study is that parental and sibling food allergies were listed as one of the possible inclusion criteria. There is a well known overreporting of food allergy in the community.11 It is therefore possible that children who were recruited purely on the basis of self-reported food allergies in family members might not have had a truly atopic family history. However, only 4 of the 449 children with SPT results at 2 years (0.9%) were included in the cohort purely on the basis of having a family member with food allergy. All of the other children who had a parent or sibling with food allergy also had an immediate family member with asthma, eczema, or hay fever. Therefore, overreporting of food allergy is unlikely to have biased the selection criteria in this cohort. A second potential limitation of the study is that only 72% of the initial cohort had a SPT performed at 2 years of age. However, this is unlikely to have affected our results because we observed no differences in sensitization to food allergens at 6 or 12 months of age, intake of soy, or sibling history of milk allergy between those who had a SPT and those who did not have a 2-year SPT. The randomized controlled trial design of the MACS allowed us to compare the ways in which a randomly allocated soy formula affected the likelihood of peanut sensitization compared with parent-selected soy products. The randomly allocated
J ALLERGY CLIN IMMUNOL VOLUME 121, NUMBER 6
soy-based formula was not a hypoallergenic or hydrolyzed formula; thus, our finding that the association between soy consumption and peanut sensitization occurred only for children given parent-selected soy products strengthens our conclusion that soy consumption itself does not lead to peanut sensitization. The study design is unlikely to have affected our other findings because formula allocation did not affect SPT results at 6, 12, or 24 months or childhood eczema (A. Lowe, unpublished data, December 2007). A subgroup analysis in a large cohort study by Lack et al3 found that ingestion of soy milk or soy formula in the first 2 years of life was associated with peanut allergy. We identified 2 potential confounders of the relationship between soy consumption and peanut sensitization that have not been previously well addressed. Lack et al3 did not control for sibling milk allergy, although they did adjust for maternal atopy as defined by responses to a questionnaire administered before delivery. In their study, parent-reported symptoms of cow’s milk allergy at age 6 months did not affect the association between soy consumption and peanut allergy. At 6 months of age, a significant proportion (74% in our study, 30% in the study by Lack et al3) of infants were still breast-fed. These infants may not have been exposed to cow’s milk formula; therefore, clinical reactions to cow’s milk might not have become apparent at this age. If this is the case, skin prick testing at 6 months is likely to be a more sensitive method than clinical reactions for defining IgE-mediated cow’s milk allergy. Our study is the first to show that the previously reported relationship between soy ingestion and peanut allergy can be explained by selective introduction of soy to infants with a family or personal history of cow’s milk allergy. This has significant methodologic implications because it highlights the importance of considering the confounding effects related to dietary modifications by parents when investigating the association between diet and childhood allergic diseases. Our finding that consumption of a randomly allocated soy formula does not cause peanut sensitization confirms results from a previous study, which compared outcomes of random allocation of soy versus extensively hydrolyzed cow’s milk formula during the first 2 years of life in a cohort of infants with cow’s milk allergy.6 Our study identified a positive association between other forms of allergy including egg and cow’s milk sensitization and eczema and the presence of peanut sensitization, suggesting that the presence of these conditions might identify infants at high risk of peanut allergy. These conditions might simply aggregate in genetically at-risk individuals, or, alternatively, common mechanisms of disease might underlie these associations. For example, cow’s milk allergy might lead to gastrointestinal mucosal damage, breakdown of the gastrointestinal antigenic sampling mechanisms, and increased risk of translocation of antigenic proteins.12 Alternatively, egg allergy might lead to eczema, which may in
KOPLIN ET AL 1459
turn result in sensitization to peanut through application of peanut oil based creams as described by Lack et al.3 Information on use of these creams is not available for our cohort study. We did not find any evidence to support a causal relationship between the consumption of soy formula or soy milk and peanut sensitization in children with a family history of allergy. Although we observed a positive association between consumption of parent-selected soy formula or soy milk and peanut sensitization, our results suggest that this association is a confounding effect of an underlying increased risk of peanut sensitization related to either personal or family history of cow’s milk allergy. Our results show that avoidance of soy cannot be recommended as a peanut allergy prevention strategy. We thank Dr John Thorburn, FRACP, for assistance in patient recruitment and administrative assistance, Christine Axelrad for assistance with data collection, Anne Balloch for data management, and all of the MACS infants and parents for their participation.
Clinical implications: There is no evidence to suggest that avoidance of soy can be recommended as a strategy for peanut allergy prevention in infants of atopic families. REFERENCES 1. Grundy J, Matthews S, Bateman B, Dean T, Arshad SH. Rising prevalence of allergy to peanut in children: data from 2 sequential cohorts. J Allergy Clin Immunol 2002;110:784-9. 2. Sicherer SH, Munoz-Furlong A, Sampson HA. Prevalence of peanut and tree nut allergy in the United States determined by means of a random digit dial telephone survey: a 5-year follow-up study. J Allergy Clin Immunol 2003;112:1203-7. 3. Lack G, Fox D, Northstone K, Golding J. Factors associated with the development of peanut allergy in childhood. N Engl J Med 2003;348:977-85. 4. Sicherer SH, Sampson HA. Peanut allergy: emerging concepts and approaches for an apparent epidemic. J Allergy Clin Immunol 2007;120:491-503. 5. Khakoo GA, Lack G. Introduction of solids to the infant diet. Arch Dis Child 2004; 89:295. 6. Klemola T, Kalimo K, Poussa T, Juntunen-Backman K, Korpela R, Valovirta E, et al. Feeding a soy formula to children with cow’s milk allergy: the development of immunoglobulin E-mediated allergy to soy and peanuts. Pediatr Allergy Immunol 2005;16:641-6. 7. Hill DJ, Sporik R, Thorburn J, Hosking CS. The association of atopic dermatitis in infancy with immunoglobulin E food sensitization. J Pediatr 2000;137:475-9. 8. Hill DJ, Hosking CS. Food allergy and atopic dermatitis in infancy: an epidemiologic study. Pediatr Allergy Immunol 2004;15:421-7. 9. Lowe AJ, Carlin JB, Bennett CM, Abramson MJ, Hosking CS, Hill DJ, et al. Atopic disease and breast-feeding: cause or consequence? JAllergy Clin Immunol 2006;117:682-7. 10. Sporik R, Hill DJ, Hosking CS. Specificity of allergen skin testing in predicting positive open food challenges to milk, egg and peanut in children. Clin Exp Allergy 2000;30:1540-6. 11. Venter C, Pereira B, Grundy J, Clayton CB, Roberts G, Higgins B, et al. Incidence of parentally reported and clinically diagnosed food hypersensitivity in the first year of life. J Allergy Clin Immunol 2006;117:1118-24. 12. van Wijk F, Knippels L. Initiating mechanisms of food allergy: oral tolerance versus allergic sensitization. Biomed Pharmacother 2007;61:8-20.
history of cow’s milk allergy. Future studies should take the confounding effects related to dietary modifications by parents into account when investigating the association between diet and childhood allergic diseases. (J Allergy Clin Immunol 2008;121:1455-9.) Key words: Skin prick test, peanut allergy, soy formula, soy milk, risk factors, sensitization
The potential severity of peanut allergy clinical reactions in combination with the apparent rise in prevalence1,2 make it an important public health issue. Research into potential environmental risk factors for peanut allergy aims to identify modifiable risk factors to help inform guidelines on preventative strategies. A recent large cohort study by Lack et al3 found that intake of soy milk or soy formula in the first 2 years of life was associated with peanut allergy. This is an important finding that is frequently referenced,4-6 although there are as yet no further confirmatory studies. The authors suggested that ingestion of soy could cause peanut allergy because of cross-sensitization through common epitopes shared by peanuts and soy.3 However, this appears unlikely because of the low prevalence of clinical reactivity to soy in children with peanut allergy.4 We hypothesized that the relationship between soy consumption and peanut allergy observed by Lack et al3 could be a result of dietary modifications by parents in response to early signs of atopic disease in infants. The objective of the current study was to examine whether the association between soy consumption and peanut allergy can be explained by selective introduction of soy to children who are at higher risk of peanut sensitization because of a family or personal history of cow’s milk allergy. To address this question, we have used data from the Melbourne Atopy Cohort Study (MACS). This study monitored the development of sensitization and symptoms of allergic disease in 620 infants randomly allocated to receive either soy formula, standard cow’s milk formula, or a partially hydrolyzed cow’s milk–based formula after cessation of exclusive breastfeeding. As part of the MACS, signs of allergic disease and sensitization to peanut, egg, and cow’s milk were recorded at frequent intervals during the first 2 years of life. Furthermore, infant dietary guidelines for the first year of life were clearly and prospectively defined and history of dietary intake recorded on a regular basis for the first 2 years of life. For these reasons, the MACS is particularly well suited to assess the association between soy consumption and the development of peanut sensitization.
METHODS Subjects The subjects are members of the MACS, a randomized controlled trial of the effect of 3 infant formulas on allergy development in a cohort of children 1455
1456 KOPLIN ET AL
Abbreviations used MACS: Melbourne Atopy Cohort Study OR: Odds ratio SPT: Skin prick test
with a family history of allergy. The MACS involved 620 infants born between February 1990 and November 1994 whose mothers were recruited during pregnancy. Infants were eligible for recruitment if their mother attended the Mercy Hospital Antenatal Clinic and at least 1 parent or sibling had atopic dermatitis, asthma, hay fever, or food allergy. The recruitment process has been described in more detail previously.7,8 The MACS was approved by the Human Research Ethics Committee of the Mercy Maternity Hospital, Melbourne.
Randomly allocated study formulas Infants were randomly allocated to 1 of 3 study formulas at study entry, and mothers were instructed to use these formulas on cessation of exclusive breastfeeding. The 3 formulas were a standard cow’s milk–based formula (NAN; Nestle´, Sydney, Australia), a partially hydrolyzed cow’s milk–based formula (NAN HA; Nestle´), and a standard soy-based formula (ProSobee; Mead Johnson, Evansville, Ind).
J ALLERGY CLIN IMMUNOL JUNE 2008
TABLE I. Baseline demographic and clinical characteristics of study cohort (n 5 449)*
Age (y), mean (SD) Mean years of education (SD) Married Owner-occupied residence Current smoker Born in Australia/New Zealand
Maternal (n 5 449)
Paternal (n 5 447)
31.6 (4.2) 13.6 (1.9) 91.5% 83.1% 7.4% 88.0%
33.7 (5.2) 13.5 (2.0) 91.7% 83.6% 17.4% 82.3%
Baseline family history of atopic disease
Asthma Eczema Hay fever Food allergy Milk Egg Other
Maternal (n 5 449)
Paternal (n 5 445)
Any older sibling (n 5 265)
42.9%à 39.7%à 59.6%à
27.0% 21.1% 45.6%
56.6% 63.4% 34.3%
20.3%à 8.0%à 30.1%
9.9% 4.7% 15.5%
42.3% 26.8% 51.7%
*Excludes 171 infants who did not have a SPT at 2 years of age. Definitions were based on self-report by mothers at recruitment into the study during pregnancy. àn 5 448.
Feeding advice Mothers were encouraged to breast-feed and to delay the introduction of solids until after 6 months of age, after which low-allergen solids (ie, avoidance of peanuts, tree nuts, fish, and egg) were recommended until 12 months of age. At 12 months, an appropriate diet for age was recommended, which included the careful introduction of all other foods as tolerated.
Exposure to soy Infants could be exposed to soy protein through the randomly allocated soy-based study formula. In addition, parents of infants in any of the 3 randomly allocated formula groups could elect to give their child a nonrandomly allocated soy formula or soy milk at any stage during the first 2 years of life. These nonrandomly allocated soy formulas and soy milk have been collectively labeled as ‘‘parent-selected’’ soy products.
Questionnaire data The ages of the child’s mother and father and their birthplace, marital status, education, and current residence were recorded by using an interviewer-administered questionnaire during pregnancy. Mothers were also asked whether they or the child’s biological father or any previous children had asthma, hay fever, eczema, milk allergy, egg allergy, or other food allergy. Data are not available on the type of other food allergy, nor were specific questions asked about the presence of peanut allergy. Detailed dietary information, including age at which foods were first introduced and whether the food was tolerated, was obtained from surveys administered by telephone by an allergy-trained nurse every 4 weeks from birth until the age of 15 months and then at 18 months and 2 years of age. These surveys also collected details of any history of allergic disorders or other illnesses experienced by the child.
Definitions of sensitization to food allergens, eczema, and cow’s milk allergy Sensitization to peanut, milk, and egg was determined by skin prick tests (SPTs) at 6 months, 1 year, and 2 years of age. Children attended the 2-year SPT session at a median age of 2 years (range, 1.9-2.6 years). Sensitization was defined as a SPT wheal size 3 mm. Eczema in the first 2 years of life was defined as a parental report of either a doctor diagnosis of eczema or any rash
(excluding rash that affected only the nappy region or scalp) that was treated with topical steroid-based preparations. Parent-reported cow’s milk allergy was defined as a parent reporting that the child developed symptoms either within 2 hours of ingesting cow’s milk or more than 2 hours after ingestion on 2 or more occasions.
Statistical analyses The association between soy consumption and peanut sensitization was examined by using logistic regression. Variables were identified as possible confounding factors if they were associated with both the outcome variable (peanut sensitization) and the exposure variable (soy consumption) with a P value less than .05. Multiple logistic regression was used to obtain odds ratios (ORs) and 95% CIs adjusted for the possible influence of confounding factors. Stata software (release 10; Stata Corp, College Station, Tex) was used for all analyses.
RESULTS Study population Of the initial cohort of 620 children, SPT results were available for 449 (72.4%) at 2 years of age. Only those children who had SPT results available at 2 years of age were included in the analysis. Parents of the children included in this analysis were predominantly Australian-born, married, and owned their own home (Table I). Hay fever was the most commonly reported allergic disease in parents, whereas older siblings were most often reported to have eczema (Table I). The majority of children (72.6%) had more than 1 immediate family member reported as having an allergic disease. Demographic and clinical characteristics of the entire study cohort have been published previously.9 The children who had a SPT at 2 years of age did not differ significantly from those without a SPT at 2 years in terms of parental marital status, home ownership, smoking status, mean years of education, and history of allergic disease or sibling history of allergic disease (data not shown). There were also no significant differences
KOPLIN ET AL 1457
J ALLERGY CLIN IMMUNOL VOLUME 121, NUMBER 6
TABLE II. Percentage of children in each of the allocated study formula groups who received the allocated formula and any soy products in the first 2 years of life Allocated study formula
Received Received Received Received
allocated formula other soy formula soy milk any form of soy formula or soy milk
Soy-based (n 5 145)
Cow’s milk–based (n 5 158)
Partially hydrolyzed cow’s milk–based (n 5 146)
Whole cohort (n 5 449)
84.8% 35.9% 28.3% 89.7%
86.7% 22.8% 19.0% 24.1%
84.9% 23.3% 23.3% 27.4%
85.5% 27.2% 23.4% 46.4%
between these 2 groups in allocation to the soy-based study formula, consumption of other soy formulas or soy milk, presence of siblings, and sensitization to any food allergens at 6 months or 1 year of age (data not shown). The majority of children received the study formula to which they had been randomly allocated at least once in the first 2 years of life (Table II). There were no significant differences in the proportion of children in each of the 3 allocated formula groups who received the study formula. A significant proportion of children also consumed nonrandomly allocated (parent-selected) soy products, with 27.2% of children consuming a parent-selected soy formula and 23.4% consuming soy milk (Table II). In total, 133 children (29.6%) received a parent-selected soy product. Children who were allocated to the soy-based study formula were more likely also to consume parent-selected soy formulas or soy milk, with 38% of these children reporting intake of an additional soy product compared with 26% of the children allocated to either the cow’s milk–based or the hypoallergenic cow’s milk–based study formula who reported intake of at least 1 soy product (P 5 .008). Overall, 208 children (46.4%) were exposed to soy in the first 2 years of life through either the randomly allocated soy formula or parent-selected soy formulas or soy milk (Table II).
Unadjusted relationship between soy consumption and sensitization to peanuts At 2 years of age, 40 children (9.4%) were sensitized to peanut. All children who had been exposed to soy in the first 2 years of life through consumption of the randomly allocated soy formula, a parent-selected soy formula, or soy milk (n 5 208) were initially grouped together. There was no evidence that these children had an increased risk of peanut sensitization compared with children who had never consumed soy formula or soy milk (OR, 1.64; 95% CI, 0.85-3.16; P 5 .141). Exposure to the randomly allocated soy formula was then considered separately from exposure to parent-selected soy products. A logistic regression model was fitted that simultaneously allowed for separate effects of peanut sensitization caused by consumption of the randomly allocated soy formula and by consumption of any parent-selected soy formula or soy milk. There was no evidence of an association between consumption of the randomized soy formula and peanut sensitization (OR, 1.35; 95% CI, 0.67-2.703; P 5.399). By contrast, consumption of a parent-selected soy formula or soy milk was associated with sensitization to peanut at 2 years of age compared with those who had not ingested soy (OR, 2.02; 95% CI, 1.04-3.92; P 5 .039). The relationship between consumption of parent-selected soy products and peanut sensitization was explored further in the following sections.
TABLE III. Factors associated with consumption of parentselected soy formula or soy milk Unadjusted OR (95% CI)
P value
(20.3) (9.8) (24.9)
2.33 (1.55-3.50) 1.45 (0.84-2.51) 2.79 (1.92-4.06)
.001 .183 <.001
(47.4)
0.99 (0.70-1.39)
.939
(4.5) (11.6) (4.7)
5.50 (2.22-13.64) 1.04 (0.59-1.83) 4.39 (2.11-9.13)
No. (%)
Family history Maternal milk allergy* 91 Paternal milk allergy 44 Sibling milk allergy 112 Allergic symptoms/signs in the child Eczema by 2 y 213 Positive SPT at 6 moà Milk 19 Egg 49 Parent-reported cow’s milk allergy 21
<.001 .892 <0.001
*n 5 448. n 5 447. àn 5 424.
Factors associated with intake of parent-selected soy products Children who were sensitized to milk at 6 months of age or experienced parent-reported symptoms of cow’s milk allergy were more likely to be given a parent-selected soy formula or soy milk (Table III). By contrast, there was no evidence of an association between sensitization to egg at 6 months or eczema in the first 2 years of life and soy consumption (Table III). Milk allergy in the child’s mother or siblings was also strongly associated with soy consumption (Table II). Factors associated with sensitization to peanuts Children who had siblings with milk allergy or were sensitized to milk or egg at 6 months of age or had eczema by age 2 years were more likely to be sensitized to peanuts, and these associations remained significant after controlling for sex, parental allergies, maternal and paternal smoking, and duration of breast-feeding (Table IV). Although there was only weak evidence of an association with parent-reported cow’s milk allergy before adjustment, controlling for the factors listed appeared to strengthen the association (Table IV). Maternal milk allergy showed no association with peanut sensitization (Table IV). Adjusted relationship between consumption of parent-selected soy products and sensitization to peanuts Only sensitization to cow’s milk at 6 months of age, parentreported symptoms of cow’s milk allergy, and history of milk
1458 KOPLIN ET AL
J ALLERGY CLIN IMMUNOL JUNE 2008
TABLE IV. Factors associated with peanut sensitization (SPT $ 3 mm) in probands Unadjusted
Family history Maternal milk allergy* Paternal milk allergy Sibling milk allergy Allergic symptoms/signs in the child Eczema by 2 years Positive SPT at 6 monthsà Milk Egg Parent-reported cow’s milk allergy
Adjusted§
No. (%)
Unadjusted OR (95% CI)
91 (20.3) 44 (9.8) 112 (24.9)
0.98 (0.43-2.20) 0.45 (0.11-1.95) 2.74 (1.41-5.33)
.959 .289 .003
213 (47.4)
4.30 (1.99-9.25)
<.001
4.78 (2.15-10.62)
<.001
19 (4.5) 49 (11.6) 21 (4.7)
6.58 (2.42-17.84) 10.68 (5.18-22.02) 2.56 (0.82-8.02)
<.001 <.001 .106
6.44 (2.27-18.25) 10.65 (4.90-23.17) 3.69 (1.05-13.01)
<.001 <.001 .042
P value
Adjusted OR (95% CI)
1.20 (0.51-2.83) 0.31 (0.07-1.41) 2.80 (1.40-5.61)
P value
.684 .129 .004
*n 5 448. n 5 447. àn 5 424. §The variables were adjusted for the infant’s sex, parental history of allergy (asthma, eczema, hay fever and food allergy treated separately), maternal and paternal smoking, and duration of breast-feeding (<6 mo vs 6 mo).
TABLE V. Results of multiple logistic regression for peanut sensitization Peanut sensitization (SPT $ 3 mm) Variable
Consumption of parent-selected soy formula or soy milk Consumption of randomly allocated soy formula Milk allergy in siblings Positive SPT to milk at 6 months Parent-reported cow’s milk allergy
Adjusted OR (95% CI)
P value
1.34 (0.64-2.79)
.434
1.10 (0.52-2.35)
.800
2.25 (1.09-4.64) 4.78 (1.63-13.96) 1.94 (0.56-6.72)
.028 .004 .297
allergy in siblings were associated with both soy consumption and peanut sensitization and emerged as potential confounders of the relationship between soy intake and sensitization to peanuts. These variables were entered into a multiple logistic regression analysis (Table V). After adjustment, the OR for peanut sensitization at 2 years of age in infants who had consumed parent-selected soy products compared with those who had not was 1.34 (95% CI, 0.64-2.79), which represents a 34% reduction from the unadjusted OR. The corresponding P value of .434 and the 95% CI indicate that there is no evidence that soy consumption was associated with sensitization to peanuts after controlling for sensitization to cow’s milk and history of cow’s milk allergy in siblings.
DISCUSSION Our study found no evidence that soy consumption leads to peanut sensitization. Although we observed a positive association between consumption of parent-selected soy products and peanut sensitization, this relationship was explained by the fact that children who had siblings with cow’s milk allergy or who were themselves sensitized to cow’s milk at 6 months of age were more likely to both consume soy and be sensitized to peanuts. Dietary information for our study was collected prospectively at frequent intervals throughout the first 2 years of life, minimizing the chance of recall bias based on allergic status and allowing us to examine the effects of reverse causality. Furthermore, the
accuracy of the data collected was likely to be high because the period for which parents were asked to recall information was short. We were unable to examine clinical reactions to peanuts because introduction of peanut products was not recommended until after 12 months in all infants, and those with positive peanut SPT results at 12 months were told to avoid peanut until 2 years of age. As a result, the majority of children with a positive SPT result at 2 years of age had not ingested peanuts in the first 2 years of life (data not shown). Although not all children who are sensitized to peanut will experience clinical symptoms, these children are at higher risk of IgE-mediated peanut allergy.10 In addition, children with a SPT wheal size <3 mm are highly unlikely to have IgEmediated peanut allergy.10 Because all children in the study had a family history of atopy, we could not assess the affect of soy consumption on peanut sensitization in children without a family history of allergy. These results therefore only pertain to individuals with a family history of allergy. A potential limitation of our study is that parental and sibling food allergies were listed as one of the possible inclusion criteria. There is a well known overreporting of food allergy in the community.11 It is therefore possible that children who were recruited purely on the basis of self-reported food allergies in family members might not have had a truly atopic family history. However, only 4 of the 449 children with SPT results at 2 years (0.9%) were included in the cohort purely on the basis of having a family member with food allergy. All of the other children who had a parent or sibling with food allergy also had an immediate family member with asthma, eczema, or hay fever. Therefore, overreporting of food allergy is unlikely to have biased the selection criteria in this cohort. A second potential limitation of the study is that only 72% of the initial cohort had a SPT performed at 2 years of age. However, this is unlikely to have affected our results because we observed no differences in sensitization to food allergens at 6 or 12 months of age, intake of soy, or sibling history of milk allergy between those who had a SPT and those who did not have a 2-year SPT. The randomized controlled trial design of the MACS allowed us to compare the ways in which a randomly allocated soy formula affected the likelihood of peanut sensitization compared with parent-selected soy products. The randomly allocated
J ALLERGY CLIN IMMUNOL VOLUME 121, NUMBER 6
soy-based formula was not a hypoallergenic or hydrolyzed formula; thus, our finding that the association between soy consumption and peanut sensitization occurred only for children given parent-selected soy products strengthens our conclusion that soy consumption itself does not lead to peanut sensitization. The study design is unlikely to have affected our other findings because formula allocation did not affect SPT results at 6, 12, or 24 months or childhood eczema (A. Lowe, unpublished data, December 2007). A subgroup analysis in a large cohort study by Lack et al3 found that ingestion of soy milk or soy formula in the first 2 years of life was associated with peanut allergy. We identified 2 potential confounders of the relationship between soy consumption and peanut sensitization that have not been previously well addressed. Lack et al3 did not control for sibling milk allergy, although they did adjust for maternal atopy as defined by responses to a questionnaire administered before delivery. In their study, parent-reported symptoms of cow’s milk allergy at age 6 months did not affect the association between soy consumption and peanut allergy. At 6 months of age, a significant proportion (74% in our study, 30% in the study by Lack et al3) of infants were still breast-fed. These infants may not have been exposed to cow’s milk formula; therefore, clinical reactions to cow’s milk might not have become apparent at this age. If this is the case, skin prick testing at 6 months is likely to be a more sensitive method than clinical reactions for defining IgE-mediated cow’s milk allergy. Our study is the first to show that the previously reported relationship between soy ingestion and peanut allergy can be explained by selective introduction of soy to infants with a family or personal history of cow’s milk allergy. This has significant methodologic implications because it highlights the importance of considering the confounding effects related to dietary modifications by parents when investigating the association between diet and childhood allergic diseases. Our finding that consumption of a randomly allocated soy formula does not cause peanut sensitization confirms results from a previous study, which compared outcomes of random allocation of soy versus extensively hydrolyzed cow’s milk formula during the first 2 years of life in a cohort of infants with cow’s milk allergy.6 Our study identified a positive association between other forms of allergy including egg and cow’s milk sensitization and eczema and the presence of peanut sensitization, suggesting that the presence of these conditions might identify infants at high risk of peanut allergy. These conditions might simply aggregate in genetically at-risk individuals, or, alternatively, common mechanisms of disease might underlie these associations. For example, cow’s milk allergy might lead to gastrointestinal mucosal damage, breakdown of the gastrointestinal antigenic sampling mechanisms, and increased risk of translocation of antigenic proteins.12 Alternatively, egg allergy might lead to eczema, which may in
KOPLIN ET AL 1459
turn result in sensitization to peanut through application of peanut oil based creams as described by Lack et al.3 Information on use of these creams is not available for our cohort study. We did not find any evidence to support a causal relationship between the consumption of soy formula or soy milk and peanut sensitization in children with a family history of allergy. Although we observed a positive association between consumption of parent-selected soy formula or soy milk and peanut sensitization, our results suggest that this association is a confounding effect of an underlying increased risk of peanut sensitization related to either personal or family history of cow’s milk allergy. Our results show that avoidance of soy cannot be recommended as a peanut allergy prevention strategy. We thank Dr John Thorburn, FRACP, for assistance in patient recruitment and administrative assistance, Christine Axelrad for assistance with data collection, Anne Balloch for data management, and all of the MACS infants and parents for their participation.
Clinical implications: There is no evidence to suggest that avoidance of soy can be recommended as a strategy for peanut allergy prevention in infants of atopic families. REFERENCES 1. Grundy J, Matthews S, Bateman B, Dean T, Arshad SH. Rising prevalence of allergy to peanut in children: data from 2 sequential cohorts. J Allergy Clin Immunol 2002;110:784-9. 2. Sicherer SH, Munoz-Furlong A, Sampson HA. Prevalence of peanut and tree nut allergy in the United States determined by means of a random digit dial telephone survey: a 5-year follow-up study. J Allergy Clin Immunol 2003;112:1203-7. 3. Lack G, Fox D, Northstone K, Golding J. Factors associated with the development of peanut allergy in childhood. N Engl J Med 2003;348:977-85. 4. Sicherer SH, Sampson HA. Peanut allergy: emerging concepts and approaches for an apparent epidemic. J Allergy Clin Immunol 2007;120:491-503. 5. Khakoo GA, Lack G. Introduction of solids to the infant diet. Arch Dis Child 2004; 89:295. 6. Klemola T, Kalimo K, Poussa T, Juntunen-Backman K, Korpela R, Valovirta E, et al. Feeding a soy formula to children with cow’s milk allergy: the development of immunoglobulin E-mediated allergy to soy and peanuts. Pediatr Allergy Immunol 2005;16:641-6. 7. Hill DJ, Sporik R, Thorburn J, Hosking CS. The association of atopic dermatitis in infancy with immunoglobulin E food sensitization. J Pediatr 2000;137:475-9. 8. Hill DJ, Hosking CS. Food allergy and atopic dermatitis in infancy: an epidemiologic study. Pediatr Allergy Immunol 2004;15:421-7. 9. Lowe AJ, Carlin JB, Bennett CM, Abramson MJ, Hosking CS, Hill DJ, et al. Atopic disease and breast-feeding: cause or consequence? JAllergy Clin Immunol 2006;117:682-7. 10. Sporik R, Hill DJ, Hosking CS. Specificity of allergen skin testing in predicting positive open food challenges to milk, egg and peanut in children. Clin Exp Allergy 2000;30:1540-6. 11. Venter C, Pereira B, Grundy J, Clayton CB, Roberts G, Higgins B, et al. Incidence of parentally reported and clinically diagnosed food hypersensitivity in the first year of life. J Allergy Clin Immunol 2006;117:1118-24. 12. van Wijk F, Knippels L. Initiating mechanisms of food allergy: oral tolerance versus allergic sensitization. Biomed Pharmacother 2007;61:8-20.