Food allergies manifest in approximately 4%–6% of the pediatric population and frequently present symptoms within the first 2 years of life [1,2]. The most common food antigens are cow’s milk, hen’s egg, wheat, peanuts, tree nuts, and seafood [3,4]. The diagnosis of immunoglobulin E (IgE)-mediated food allergy includes detailed clinical history, skin prick test (SPT), and serum-specific IgE (sIgE) measurements; however, in suspicious cases, the diagnosis should be confirmed with oral food challenge (OFC). Nonetheless, OFC is a time- and staff-intensive procedure and carries a risk of severe reactions, which may lead to anaphylactic shock [5].

The basophil activation test (BAT) is a flow cytometry test that measures the expression of activation markers on the surface of blood basophils. The BAT has been shown to reflect the allergic status of patients sensitized to food, inhalant, and insect venom antigens and has gained importance in eligibility for a specific therapy and monitoring of the response to therapy for an allergy [6]. The blood sample taken is compared with the suspected antigen in a tube, which prevents patients from being exposed to the antigen under investigation. For patients and their families, it is a potentially safer and preferable test because there is no chance of an allergic reaction that could result in anaphylaxis.

Studies have shown that cord blood cells or cytokines in high-risk infants predict allergic diseases such as recurrent wheezing attacks later in life [7-9]. One of the discussed cell populations in predicting allergic outcomes is basophil progenitor cells in cord blood; however, it is still unclear whether these cells have predictive value [8,10]. It could be crucial in newborns to anticipate the food allergies that will manifest later to avoid severe reactions like anaphylaxis. The aim of this study is to determine whether cow’s milk allergy (CMA) in infants can be predicted using BAT with cord blood samples by stimulating basophils with milk protein antigens (cow’s milk and casein). CMA is most frequently caused by caseins, but they can also be promoted by whey proteins [11]. CMA is defined as a reproducible adverse reaction to one or more cow’s milk component proteins [12].

Basophils are expected to be sensitized earlier for a positive result in a BAT test and its reactivity to an unexposed food is not an expected reaction in a cord blood sample, however several studies have shown that, newborns may develop sensitivity because of environmental factors and epigenetic changes during the mother's pregnancy [7]. As not all children born to parents with allergies will develop allergic disorders, and not all children with allergies will be born to parents with atopy, this study may allow for individualized monitoring of mother and baby carrying a high risk for food allergies.

A total of 30 mother-child pairs were included in the study. Demographic characteristics of the study participants are shown in Tables 1 and 2. Postnatal age at the time the survey was carried out had a mean of 13.6 ±1.5 months. All of the cases were born with normal birth weights, and most underwent cesarean section delivery (Table 1). The majority of them were breastfed but not exclusively, mostly for at least 6 months. Formula was also used in about one-third of the cases, primarily during the first 3 months (Table 1). Allergic symptoms mostly emerged after the 6th month of birth (Table 1).

Mother, father, and sibling atopy rates were 26.7%, 13.3%, and 13.3%, respectively. Of the mothers, just 17% were taking regular medication, and 60% had been exposed to smoke during pregnancy. Only a small group of cases (17%) had pets at home (Tables 1 and 2).

It was found that 9 of the cases (30.0%) developed symptoms of food allergy, and 8 of them had CMA. Two of the cases with CMA were diagnosed with a history of CMA and SPT (Table 3). Six cases had a convincing history of allergic reactions following the ingestion of cow’s milk protein; however, only 3 of them gave consent for SPT and no evidence of sensitization observed (casein was not included in any of the SPTs performed in our city in that period) (Table 3).

We compared the 2 groups, cases developing allergies or not, with univariate analysis (Table 4). Variables with P values of <0.200 were cases fed with formula, atopy in the family, nutrition type during pregnancy, smoking during pregnancy, neutrophil-to-lymphocyte ratio (NLR), and casein-BAT score. As the median values of cow's milk BAT and casein-BAT were 4.1 (0.1–24.6) and 2.6 (0.1–15.0), we compared the 2 groups according to these median levels (Table 4). No difference was found for cow's milk BAT (P=0.69); however, there was a significant statistical difference in the comparison of casein-BAT, with casein BAT ≥2.6 present in 8 of 9 infants (88.9%) who developed food allergy symptoms and in 8 of 21 (38.1%) of those without symptoms (P=0.017). With one exception, all cases that developed an allergy exhibited a casein-BAT score of 2.6 or higher. The case with a low casein-BAT score was followed for recurrent wheezing and flare-up of atopic dermatitis, and the rest of the allergic cases were followed up for allergic skin lesions (Table 3). Since the receiver operating characteristic (ROC) analysis did not identify a statistically significant cutoff value for casein-BAT (area under the curve=0.651, P=0.197), it was not possible to determine an optimal threshold for classification. Therefore, casein-BAT values were dichotomized based on the median and included as a categorical variable in the logistic regression analysis.

Risk factors were determined by creating a regression model using formula use, nutrition during pregnancy, family atopy, smoking during pregnancy, NLR, and casein-BAT. It has been found that cord blood casein-BAT score of 2.6 or higher increases the risk of developing food allergy symptoms in the first year of life by 33.2 times (P=0.03) (Table 5).

We computed a post hoc power for the observed association between casein BAT ≥2.6 (vs. <2.6) and 12-month food allergy symptoms using GPower v3.1.9.7 (exact unconditional test for 2 independent proportions, 2-sided; α=0.05; n=9 vs 21; control proportion=0.38); the achieved power (1–β) was 0.93.

Because intrauterine factors are considered to play a role in the developmental origins of many diseases, including food allergies, there has been interest in analyzing cord blood for possible biomarkers of atopic disease [7,8,16]. Finding abnormalities in many of the cellular and genetic components of cord blood in children who are at risk of atopy is the main objective of biomarker research. Several possible cord blood biomarkers of atopic risk have been identified, such as level of total IgE, decrease in the ratio of T-helper cell type 1 (Th1) to type 2 (Th2), low interleukin (IL)-10 (a Th1-related cytokine), increased CCL22 (a Th2-related chemokine), decreased frequency and function of Tregs, and lower levels of IL-12 and IL-15, which have been associated with allergic diseases, such as atopic dermatitis, food allergy, and recurrent wheeze [16-20]. However, the findings have been inconclusive at the moment, and there is particularly a lack of evidence regarding prospective studies.

Allergic reactions and anaphylaxis are the result of mast cell and basophil activation and degranulation. Results from both humans and mice have shown that basophils are present in circulation from birth; however, mast cells emerge only in the skin during embryonic stages, which appear to be immature and unable to respond to IgE until after birth [21,22]. In neonates, basophils are more readily available cells in the peripheral blood than mast cells and are, therefore, more accessible for testing even in the cord blood [23,24]. This advantage of basophils in neonates and the relationship of BAT with various allergic diseases provided the rationale in our research for studying BAT in cord blood [5,8]. While neonatal basophils theoretically need to be sensitized before they can respond to casein, further studies are needed to determine how basophils of some of our cases had a response to casein. Even though it is unknown if basophils are sensitized with sIgE before birth, studies have shown that basophils are highly sensitive to some antigens and have a high level of FcεRI expression in utero [25,26]. Understanding how BAT functions in this context could yield valuable insights into the development of allergic conditions later in life.

BAT evaluates blood basophils following their exposure to specific antigens or controls [6,27]. When activated, basophils increase the expression of various activation markers on their surface, which can be quantified using flow cytometry [6]. The most commonly utilized marker for assessing basophil activation is CD63, a lysosomal-associated membrane protein. CD63 is initially located on the membrane of secretory lysosomes within the basophil but becomes expressed on the cell's plasma membrane during degranulation, which makes it the most frequently used marker to assess basophil activation [6]. Despite its use to diagnose and monitor allergic diseases, BAT has several limitations. As well as the standardization issue, it also requires large volumes of fresh blood and needs to be processed immediately. In our study, we studied 30 mother-child pairs who consecutively applied to our center for birth and gave informed consent for collecting the cord blood samples to study BAT and to conduct a questionnaire at the end of the first year. Initially, our study aimed to evaluate responses to cow's milk and casein using the BAT test and examine the potential development of CMA during follow-up. However, our findings revealed that BAT by cow's milk was not linked to subsequent milk or food allergies. Instead, we discovered that BAT specifically triggered by casein was associated with a higher likelihood of developing food allergies later on. This may be important in cases where close monitoring for food allergies is required, and earlier introduction of complementary foods may be considered in view of tolerance development.

Although cord blood analysis is a popular method of study to measure allergy risk in later stages of life, to our knowledge, the BAT has never been used as a method in cord blood. Although the use of BAT in clinical practice is promising, there is still a need for standardization [5]. Even though some efforts have been made, the main unmet need is to establish a cutoff value for each antigen [28]. Although we were unable to determine a cutoff value for casein-BAT, we showed on the basis of the median value that food allergies occurred more frequently in the first year in people with casein-induced basophil activation values of 2.6 and above. It is noteworthy that one patient who developed an allergy exhibited a score lower than 2.6 and had a history of recurrent wheezing and flare-ups of atopic dermatitis. In contrast, the remaining allergic cases had a score of 2.6 and above, and all were followed up for allergic skin lesions. Basophils play an important role in the pathogenesis of allergic skin lesions and the high BAT score in the cord blood of these patients suggests that basophil sensitivity may be present from birth. Other cases with high casein-BAT scores did not have allergies in the study period; however, these patients will be monitored longer to see if they develop allergies.

Out of the 9 cases where allergies developed, SPT was conducted in 5 instances. The allergens were identified in 3 of these SPTs (hen's egg, strawberry, banana, and cow's milk). We could not show the relationship between allergies and casein due to the absence of casein in SPT tests in 5 cases and no SPT in 4 cases; however, the relationship between casein-BAT and food allergy may be important in showing that these cases are likely to be predisposed to food allergies from birth.

The most important limitation of the study is the low number of cases, which prevented us from making subgroup evaluations such as the type of food allergy or the type of symptoms. Therefore, the results presented should be interpreted with caution and require further validation. The second limitation: we were unable to identify a significant cutoff value in the ROC analysis since we used the median casein-BAT value in the logistic regression analysis as a categorical variable. With a larger number of patients, the casein-BAT test can be standardized, and a cutoff value can be obtained for a valid use. One can speculate as to why the BAT with cow's milk does not yield the same results as casein. BAG-F2 contains multiple milk proteins, and the relative representation of casein epitopes may be lower compared with BAG-F78. Furthermore, in complex extracts, conformational changes or masking effects may reduce the accessibility of casein epitopes, which may explain why casein BAT was positive despite a lack of significant reactivity to cow’s milk extract. Lastly, we conducted a single-center, prospective pilot feasibility study to obtain preliminary association signals. Because post hoc power is driven by the observed effect size and our estimates have wide CIs, these results should be viewed as hypothesis-generating and used to inform a priori calculations for future larger studies. The higher frequency of food allergy detected in our study was probably due to a selection bias consisting of patients admitted to a tertiary center.

In conclusion, a high casein-BAT value in cord blood may be predictive for food allergy that may develop in the first year of life or even later. Antigen-specific BAT evaluation in cord blood may pave the way for changes in the lifestyle, environmental factors, and diet of allergic children if validated in a larger number of patients.