Perinatal Characteristics and the Sensitization to Cow Milk, Egg Whites and Wheat in Children up to 3 Years of Age

Food sensitization in early life identifies children at risk of developing allergic diseases. We investigated the sensitization to cow milk (CM), egg whites, and wheat. Newborns and infants under 3 years of age with available specific immunoglobulin E (sIgE) data were identified. A retrospective survey was conducted using data from the Chang Gung Research Database. Perinatal characteristics, such as singleton or multiples in a single pregnancy, parity, meconium staining, maternal age, spontaneous delivery or cesarean section, meconium passage, weeks of gestation, birth length, body weight, head and chest circumferences, and season, were obtained. The data on sIgE were collected, and a logistic regression model was used to determine the odds of sensitization. Positive sIgE for CM and egg whites was more likely to occur in boys than in girls. Early-life egg white and wheat sensitization was associated with increased birth body length and weight. A multivariate analysis indicated an association between egg white sIgE positivity and logarithmic total IgE. Higher total IgE levels and younger age were associated with egg white sensitization, and elevated weight and length at birth were linked to food sensitization, particularly to egg whites and wheat.


Introduction
Early-life food sensitization has identified children who are at high risk for subsequent allergic diseases [1]. In particular, milk, eggs, and wheat were the most common food hypersensitivities encountered during the first 3 years of life [2]. A prospective birth cohort study addressed prenatal and perinatal issues and identified risk factors for food allergy in Taiwanese young children [3]. The survey demonstrated that eggs (26/95, 27.3%) and milk (22/95, 23.2%) were the most common food allergens.
Food allergies are complex diseases characterized by genetic and environmental factors [4,5]. Cow milk (CM) was often the first food introduced into a diet after breast milk. In a study on CM allergy (CMA), a European birth cohort of 12,049 children showed 0.54% challenge-proven CMA and 23.6% non-immunoglobulin (Ig) E-mediated CMA [6]. Children showed the maximum response of positive specific (s)IgE to CM (80%) [7]. One factor associated with the development of child-onset food allergies was a higher level of total IgE than that in adult-onset allergies [7]. Evidence from a randomized clinical trial of 330 enrolled neonates at risk of atopy suggested that sensitization to CM and food allergies, including CMA and anaphylaxis, risk of asthma, and recurrent wheezing in young children, especially among those with high total IgE levels, was primarily prevented by avoiding CM formula supplementation at least for the first 3 days of life [8,9]. However, the introduction of small amounts of CM between 1 and 2 months of life was associated with a lower likelihood of CMA [10]. Of all children with confirmed IgE-mediated CMA, fewer tolerated CM 1 year after diagnosis than infants with negative sIgE. Boer et al. identified cut-offs for milk sIgE in relation to baked milk tolerance that could support timely decisions regarding referrals to an oral food challenge [11]. Laboratory tests assisted in diagnosing IgE-mediated CMA. Early diagnosis of CMA is important because delayed diagnosis may lead to nutritional disorders and a subsequent increased risk of impaired growth [12]. Egg white allergies are another common food allergy that frequently occurs in the first year of life. These allergies could be immediate (IgE-mediated) or delayed (non-IgE-mediated) and could involve many organ systems [13]. A prospective database revealed that nearly one-third of the infants were allergic to multiple foods, including the three main allergens in Japan: egg, milk, and wheat [14,15]. However, some of the evidence about the perinatal factors for food sensitization was conflicting [16]. This study aimed to provide a better understanding of the perinatal characteristics associated with the odds of sensitization to CM, egg whites, and wheat.

Methods
This retrospective study used data from the Chang Gung Research Database to investigate the association between perinatal characteristics and sIgE against CM, egg whites, and wheat from 2015 to 2019. The research protocol was approved by the Ethics Committee of the Chang Gung Memorial Hospital (202001016B0). As shown in Figure 1, we included children aged 0-3 years, selected from patients who had available measured serum-sIgE levels and who were healthy term neonates admitted to the baby room (n = 1468). Our inclusion criteria were babies with a single diagnosis of live-born infants (the International Classification of Disease, 9th Revision Clinical Modification, ICD-9-CM V30-37 or ICD-10-CM Z38) at the time of discharge from the baby room, normal spontaneous delivery, or cesarean section as per the payment category in the basic file of the hospitalization. They were considered discharged without any complications or comorbidities. We investigated CM (n = 1100), egg white (n = 1098), and wheat (n = 1089) sensitization. Children with allergic symptoms suspected of having food allergies were tested, and we identified the phenotypes of asthma/wheezing, dermatitis, and rhinitis using the codes provided by ICD-9-CM prior to 2016 and ICD-10 afterward. Data were extracted from the database of patients tested using the ImmunoCAP (Thermo Fisher Scientific Inc., Phadia AB, Uppsala, Sweden) or a multiple allergen simultaneous test system (MAST, OPTIGEN; Hitachi Chemical Diagnostics, Inc., Mountain View, CA, USA). Total IgE was detected using ImmunoCAP or nephelometry at the time of a specific IgE survey [17]. Categorical results were presented as numbers and continuous results as means ± standard deviation (SD) or median (range). We used chi-square or the Fisher's exact test to detect significant differences between sensitization to CM, egg whites, or wheat, and  Specific IgE between 0 and 0.34 kUA/L using ImmunoCAP or 0 and 26 LUs using MAST was considered negative [18,19]. For ImmunoCAP, positive sensitization (class 1-6) referred to values > = 0.35 kUA/L, and for MAST it referred to values > = 27 LUs [20,21]. Associated factors included singleton or multiple pregnancies, parity, meconium staining, maternal age, normal spontaneous delivery or cesarean section, birth length, body weight, head and chest circumferences, meconium passage, birth season, and weeks of gestation. Mothers who underwent emergency cesarean section for premature birth with tocolytic failure, babies with fetal distress, or severe maternal diseases such that newborns could not be admitted to the baby room were excluded from this study [22,23]. We performed a sensitivity analysis due to the increasing numbers of food allergy diagnoses after Learning Early About Peanut Allergy (LEAP) trial on evaluated allergens [24].
Categorical results were presented as numbers and continuous results as means ± standard deviation (SD) or median (range). We used chi-square or the Fisher's exact test to detect significant differences between sensitization to CM, egg whites, or wheat, and characteristics, such as sex, mode of delivery, parity, birth season, maternal age, and other perinatal factors. Student t-test was applied to explore the effect of gestational age, birth weight/length, and maternal ages divided into two groups (<40 years and 40 years). One-way analysis of variance was used to compare the three groups of maternal age ( 25 years, 25-40 years, and 40 years). Odds ratios (ORs) were adjusted for confounding variables and calculated using negative sensitization as the reference category. Adjustments with multivariable logistic regression were allowed for possible relevant confounders and factors showing a potential influence in univariate analysis, such as age, sex, logarithmic (log) total IgE, and birth body length, to identify the independent factors for allergen sensitization in infants. Analyses were performed using the Statistical Analysis System Package (SAS statistical software version 9.4; SAS Institute, Cary, NC, USA).

Is There a Role of Perinatal Characteristics for sIgE of CM?
Among the 1,100 children with available CM-sIgE, 307 (27.9%) tested positive for the sIgEs against CM. In total, 32.1% and 8.2% of the children tested positive for egg whitesand wheat-sIgE, respectively (Table 1). A total of 27 children were sensitized to eggs and wheat, 147 to CM and eggs, 2 to CM and wheat, and 54 to all three foods. The mean age at diagnosis of positive sIgE to CM was 2.0 years (SD, 0.6 years). Positive CM-sIgE levels were more likely to occur in boys (p = 0.012; Table 1). No significant difference in age was observed in the negative and positive groups (2.0 ± 0.7 and 2.0 ± 0.6 years, p = 0.286). The children with positive sIgE to CM had higher total serum IgE levels than the children with negative sIgE (median 178 U/mL (range 11.9-3749.0) in the group with positive sIgE; 59.2 U/mL (range 2.1-4349.0) in the group with negative sIgE; p < 0.001 of both log and arithmetic calculations). We found significantly higher log total IgE levels in boys when we divided our data into groups with negative and positive CM-sIgE (log total IgE 1.9 ± 0.6 in the boys' negative group; 1.7 ± 0.6 in the girls' negative group, p < 0.001; 2.3 ± 0.5 in the boys' positive group, 2.2 ± 0.5 in the girls' positive group, p = 0.046) [25]. The mean gestational age of the patients in the negative sensitization group was 38.5 (SD ± 1.6) weeks and that in the positive group was 38.5 (SD ± 1.3) weeks (p = 0.613). We hypothesized that there was an association between perinatal characteristics and the odds of sensitization to CM and found that CM-positive sIgE was not associated with cesarean delivery, number of fetuses, parity, meconium staining, maternal age, birth length and weight, head and chest circumferences, meconium passage, or birth season (p > 0.05). Similar results were observed when the patients were grouped according to sex. Independent factors for sIgE positive to CM were elevated total IgE or log total IgE values (adjusted odds ratio (aOR) 1.001, p < 0.001; aOR 3.820, p < 0.001) and not age, birth body length, or sex ( Figure 2). served when the patients were grouped according to sex. Independent factors for sIgE positive to CM were elevated total IgE or log total IgE values (adjusted odds ratio (aOR) 1.001, p < 0.001; aOR 3.820, p < 0.001) and not age, birth body length, or sex ( Figure 2).

Is There an Effect of Perinatal Characteristics on sIgE of Egg White and Wheat?
Our findings indicate that a greater proportion of boys (36.3%) than girls (26.4%) aged 0-3 years had egg white sensitization (p = 0.001, Table 1). Age and gestational age were similar in the two groups with negative and positive egg white-sIgE. The mean age of allergic children with negative egg white-sIgE was 2.0 years (SD 0.7), and that of children with positive sIgE was 2.0 years (SD 0.7, p = 0.207). The mean levels (SD) of gestational age in children with negative and positive egg white-sIgE were 38.5 ± 1.6 weeks and 38.5 ± 1.3 weeks (p = 0.629), respectively. Two hundred and seventy-six (25.1%) children were born by a cesarean section (p = 0.796 between the groups with negative and positive egg white-sIgE). Furthermore, greater birth body lengths and weights were associated with significantly higher odds of egg white and wheat IgE positivity (p < 0.05, Tables 2A-2C). This association between birth body length and egg white-sIgE persisted even after controlling for age, sex, and log total IgE (adjusted OR (aOR), 1.067, 95% confidence interval (CI), 1.007-1.130, p = 0.029, Figure 2). The independent factors for positive sIgE to egg whites were younger age and higher log total IgE (aOR 0.713, 95% CI, 0.568-0.896, p = 0.004; aOR 5.556, 95% CI, 4.151-7.437, p < 0.001) but not sex (aOR 1.076, 95% CI, 0.791-1.465, p = 0.641).

Perinatal Characteristics and the Odds of Sensitization
Singleton or multiple pregnancies, parity, meconium staining, maternal age, normal spontaneous delivery or cesarean section, birth head and chest circumferences, meconium passage, birth season, and weeks of gestation were not associated with egg white sensitization. A univariate analysis showed that the developments of sIgE to CM, egg whites, or wheat were not related to birth season (p > 0.05). Maternal age was similar in the negative and positive groups whether we divided it into two groups (<40 years and 40 years, Tables 2A-2C) or three groups ( 25 years, 25-40 years, and 40 years; p > 0.05).

Association between Allergic Diseases and Sensitization to Allergens
In this association analysis, 53.4% of the patients with positive CM-sIgE had significantly higher rates of asthma or wheezing than patients with negative CM-sIgE (38.0%, p = 0.006 after the Bonferroni correction for multiple comparisons) but not wheat (p = 1.000; Table 3). Children with positive CM sensitization had a significantly increased coexistence of asthma and rhinitis (p = 0.014). Children with wheat sensitization had similar odds of developing the above-mentioned allergic diseases compared to those without wheat sensitization (p > 0.05) ( Table 3).

Discussion
This study investigated the association between perinatal factors and IgE sensitization to common food allergens. This retrospective survey, which involved a large database, enhanced the findings of previous studies. The results showed a relationship between sIgE against egg whites and log total IgE, younger age, and birth body length. These findings agreed with the results of previous studies that tested egg white tolerance and indicated that most egg allergies developed in the first year of life and that the overall prognosis for tolerance was good [14,26,27]. Healthy newborns with larger body sizes tended to develop food sensitization, especially for egg whites and wheat. In accordance with the findings of a previous longitudinal study, which found that a decreasing trend in infantile body mass index Z-scores was significantly associated with a higher prevalence of IgE sensitization in 1-year-olds, a previous cross-sectional study revealed that body height was negatively associated with egg white sensitization and positively associated with mite sensitization in a preschool subgroup of children with allergies [28,29]. An increased surface area of contact with the allergens was suspected. The study of atopic phenotypes and skin prick tests of inhaled allergens identified no association between birth body length and atopy. However, our study focused on food and not inhalant allergens [30]. Future research should investigate the longitudinal changes in body sizes, including body height and weight.
The landmark LEAP trial suggested that peanut allergies could be prevented through early peanut introduction after screening high-risk infants. These guidelines have been rewritten to provide strong evidence that early peanut crop introduction has a potentially dramatic benefit. Multiallergen testing in infants became more popular after the LEAP study [24,[31][32][33]. The findings agreed well with the subgroup results of our study (Tables 4A-4C).
The relevant literature on male predisposition to allergic diseases or allergens appears to be compatible with the specific results obtained in the present study [34,35]. Numerous studies in the literature have dealt with the role of sex in total IgE concentration and suggested significantly higher total IgE levels in male patients, for both children and adults [36]. One explanation for this finding was sex differences in the stress response systems to early-life stress and the association between increased prenatal maternal stress and early asthma phenotypes [37]. In contrast, certain food allergies, such as shrimp, demonstrated no sex-related differences [38]. In line with our results, these findings suggest that the effects of cesarean section, gestational age, multiparity, birth season, and maternal age were not pronounced among children with food allergies [39][40][41][42]. Regardless of food types, studies have indicated that children born in autumn and winter were more prone to food allergies, but Taiwan has plenty of sunshine throughout the year and thus, has little effect on the fluctuation of vitamin D [43,44].
A nationwide Swedish cohort study found that food allergies were more likely to occur in girls and were associated with cesarean delivery during a 13-year follow-up period. However, the above-mentioned study did not exclude newborns with complications, such as premature birth, nor did it analyze specific food allergies, and it was difficult to believe that perinatal factors still had an impact on food allergies 13 years later [39]. In a birth cohort study, cesarean sections were reported to cause a predisposition to food allergies [40].
Patients included in this analysis were highly atopic. These findings are in accordance with previous results that tested highly atopic infants with greater sensitization to CM, egg whites, and wheat, even though these studies used different measures of sIgE [45]. Although CM sensitization was frequently observed in admitted wheezing infants, positive CM-sIgE was not predictive of later childhood asthma induced by wheat, egg white, or inhalation allergens [46][47][48]. While fetal measurements are a biologically plausible index of respiratory function, whether positivity for egg white is significantly correlated with birth body length remains unclear [49].
The strengths of this study included the large database design and detailed perinatal stress factor analysis (i.e., previously identified vulnerable developmental windows for atopy risk and stress programming). Despite the large number of participants in the database study, the present study had some limitations. The first limitation was rooted in false negatives, which were non-IgE-mediated food allergies [39,50]. The second limitation was that children with allergy symptoms were more likely to be tested, and our patients were those who tended to have high total IgE, resulting in selection bias and a high positive rate. Although the current study used a large database design with detailed perinatal factors, some information influencing food sensitization could not be acquired from the database, including breastfeeding duration, maternal smoking, pet exposure, maternal supplementation of vitamin D or folate, timing of solid food introduction, day care attendance, and family history of allergy [51]. After validation, the Chang Gung Research Database included a population with high disease severity [52]. The adults did not visit medical centers for mild rhinitis or dermatitis [53]. We also found that the positive rate of food allergens in adults was much lower than that in children [54]. Therefore, maternal sIgE levels and reported maternal allergic diseases were not included in the analysis of this database.
However, the influence of birth factors on food sIgE levels remains controversial. In summary, higher total IgE levels and a younger age were associated with egg white sensitization. We found that increased birth weight and length were associated with food sensitization, particularly to egg whites and wheat.

Informed Consent Statement:
The Ethics Committee waived the requirement for informed consent for this study, and the data were analyzed anonymously.

Data Availability Statement:
Data is unavailable because the owner of this database is Chang Gung Memorial Hospital.