Efficacy of Starting Very-Low-Dose Cow’s Milk Intake in Children Aged 1–12 Years With Severe Cow’s Milk Allergy: A Retrospective Study

Abstract

Purpose

To avoid the complete elimination of cow’s milk (CM) from the diet, we introduced a very-low-dose (VLD) oral food challenge (OFC) as part of an OFC-based stepwise ingestion protocol in patients with severe CM allergy in 2019, when the OFC protocol at our center was revised. This study investigated the efficacy of initiating CM intake through an OFC-based stepwise ingestion protocol starting from a VLD in achieving a negative full-dose OFC.

Methods

Patients with a history of overt allergic reactions to low-dose (LD; 3 mL CM: 100 mg CM protein) CM or lesser within 6 months and those with complete CM elimination were included. We retrospectively compared the rate of passing a full-dose OFC (200 mL CM: 6,600 mg CM protein) between patients who underwent an LD OFC before 2019 (LD group) and those who underwent a VLD OFC (0.3 mL CM: 10 mg CM protein) after 2019 (VLD group). The period for passing the full-dose OFC was evaluated using Kaplan–Meier survival analyses.

Results

We allocated 270 and 267 patients into the LD and VLD groups, respectively. The median age at initiation of the OFC-based stepwise ingestion protocol was 3.6 (2.3–5.4) and 3.4 (2.0–5.1) years in the LD and VLD groups, respectively. The LD and VLD groups had CM-specific immunoglobulin E levels of 21.9 (10.9–44.0) kUA/L and 20.5 (9.0–47.2) kUA/L, respectively. The baseline clinical characteristics of the two groups showed no significant differences in background. Over 4 years, the LD and VLD groups passed the full-dose OFC at rates of 45% and 60%, respectively, with a significant difference (log-rank test, P = 0.023).

Conclusions

Initiating CM intake through an OFC-based stepwise ingestion protocol starting from a VLD may contribute to achieving a negative full-dose OFC in patients with severe CM allergies.

INTRODUCTION

Cow’s milk (CM) is one of the most prevalent and causative foods for childhood food allergies (FAs) globally.¹,²,³,⁴ In Japan, CM accounts for 21.8% of immediate-type FAs with frequent symptom induction due to accidental ingestion.¹ Therefore, managing FAs, including those caused by CM, should involve minimally necessary causative food elimination based on an accurate diagnosis.¹ Thus, to achieve effective management, the appropriate threshold dose for each patient should be determined, considering their symptom induction thresholds. Additionally, the intake should be gradually increased based on the results of the oral food challenge (OFC).⁵,⁶,⁷,⁸ Japan adopts a universal health insurance system that allows OFC to be conducted as a part of insured medical services. Consequently, performing the OFC in a hospital setting is recommended.¹ The Japanese guidelines currently recommend a three-step approach to performing an OFC: low-dose (LD), medium-dose, and full-dose OFC.¹ Previous reports have indicated that a stepwise OFC approach can be used to avoid completely eliminating causative food, thereby improving FA management and safety.⁹,¹⁰ Additionally, in the United Kingdom and Canada, the “milk ladder” approach has been proposed, with reports on its efficacy and safety¹¹,¹²,¹³,¹⁴ supporting the effectiveness of necessary minimal elimination of causative food.¹⁵

However, in the cases of severe CM FAs, approximately 5% of patients may exhibit immediate-type allergic symptoms even with the intake of CM protein equivalent to 2.9 mg.¹⁶ Under the current management method using minimal necessary elimination, which involves evaluation of symptom-induced threshold level using the three-step OFC, many patients with severe CM FA show positive LD OFC results. These patients should continue a complete causative food elimination. Studies have shown that 42% of patients who have not developed tolerance to CM by the age of 6 years still have a CM allergy at 12 years; those forced to completely eliminate CM at the age of 6 years are at the highest risk of a persistent CM allergy.¹⁷

In such severe cases, conventional oral immunotherapy (OIT) is a suitable approach.¹⁸,¹⁹,²⁰,²¹ However, conventional CM OIT is initiated from age 4 years and older based on its effectiveness.²² Furthermore, conventional OIT can cause adverse reactions, including anaphylaxis; thus, it cannot be casually started in many cases.²³

Therefore, in 2019, our institution introduced a very-low-dose (VLD) OFC as part of an OFC-based stepwise ingestion protocol to avoid complete CM elimination in patients with severe CM-induced FAs. The rationale was to begin CM intake at a minimal dose in order to shorten the duration of complete elimination, with the ultimate goal of increasing the subsequent negative rate of full-dose OFC. Currently, the efficacy of initiating CM intake through a VLD OFC-based stepwise ingestion protocol on the full-dose OFC negativity rate in CM allergy management remains unclear.

Therefore, in this study, we aimed to investigate the effectiveness of introducing CM intake starting from a VLD within an OFC-based stepwise ingestion protocol in increasing the subsequent negative rate in full-dose OFC.

MATERIALS AND METHODS

Study design and population

This retrospective cohort study in a single center included patients diagnosed with immunoglobulin E (IgE)-related CM allergy who initiated the OFC-based stepwise ingestion protocol at Showa Medical University Hospital between January 1, 2012, and August 31, 2024. In this study, “OFC” refers to a hospital-based diagnostic procedure performed to evaluate clinical reactivity at a specific dose. The “OFC-based stepwise ingestion protocol” refers to a management strategy that includes continued ingestion at home following a negative OFC result. The inclusion criteria were as follows: a clear history of immediate allergic reactions to < 3 mL of CM (100 mg of CM protein) or a positive 3 mL CM OFC result within 6 months, as determined through patient interviews, and complete CM elimination at the time of visit to our institution. The patients were allocated into two groups: those who underwent LD OFC before 2019 (LD group) and those who underwent VLD OFC from 2019 onwards (VLD group). Blood tests were performed within 6 months of the initial OFC to confirm sensitization to CM-specific immunoglobulin E (sIgE) levels ≥ 0.7 kU_A/L. Patients who underwent OFC to confirm the threshold for initiating conventional OIT were excluded. Patients who discontinued follow-up at our hospital during the study or transitioned to conventional OIT were considered dropouts. This study was conducted in accordance with the principles of the Declaration of Helsinki (as revised in 2013) and was approved by the Institutional Review Board of Showa Medical University School of Medicine (approval No. 2023-282-A).

OFC-based stepwise ingestion protocol for CM

Supplementary Fig. S1 illustrates the details of the OFC-based stepwise ingestion protocol. Fresh milk was used in all procedures. The LD group followed a three-step ingestion protocol consisting of LD (LD; 3 mL of CM containing 100 mg of CM protein), medium-dose (24 of mL CM containing 800 mg CM protein), and full-dose (200 mL of CM containing 6,600 mg of CM protein) administration. In contrast, the VLD group followed a four-step ingestion protocol that included an additional VLD (VLD; 0.3 mL of CM containing 10 mg of CM protein) step preceding the LD stage.

After obtaining a negative result at each step, participants were instructed to continue regular ingestion of CM at home, up to the negative dose level. The patients were instructed to consume milk at home, using a syringe or measuring instrument to ensure precise and consistent intake. Only the consumption of fresh milk was permitted in the protocol, and other dairy products such as yogurt and cheese were not included. Although the frequency of home ingestion was left to parental discretion, at least twice-weekly ingestion was recommended. The next step was typically conducted 3–6 months later, as determined by an allergy specialist, based on each patient’s medical history and laboratory findings. If allergic symptoms developed at home, the intake was reduced by one step (for example, from medium- to LD). If symptoms appeared at the LD step in the LD group or the VLD step in the VLD group, complete elimination was advised, and the patient was instructed to seek immediate outpatient evaluation.

When anaphylaxis or shock occurred, CM ingestion was discontinued immediately, and patients were advised to seek immediate outpatient care. Regarding positive results at each step, participants were advised to continue ingestion at a smaller level at home, with the dose set one step below the level at which a positive reaction occurred, following the same protocol. However, if a positive result occurred at the LD step in the LD group or at the VLD step in the VLD group, complete elimination was recommended. For participants showing a positive outcome at a given step, a re-challenge within this stepwise ingestion protocol was considered based on the severity of the induced symptoms. The decision for re-challenge, typically performed 3–12 months after the initial positive test, was made by an allergy specialist.

All ingestion steps were conducted in accordance with the Japanese Guidelines for Food Allergy.¹ The hospital-based procedure was performed under the supervision of an allergist. Antihistamines and leukotriene receptor antagonists were discontinued 3 and 1 days before ingestion, respectively, to avoid confounding effects.²⁴ Each ingestion session was conducted openly, and patients were observed for 3 hours after milk administration. If the responsible allergist confirmed a positive reaction defined as Sampson grade ≥ 2,²⁵ appropriate medications such as antihistamines or inhaled beta-agonists were administered for cutaneous or respiratory symptoms, respectively, and intramuscular adrenaline was administered for anaphylaxis. Grade 1 symptoms were regarded as indeterminate and reassessed through home ingestion to confirm reproducibility. A reaction was defined as positive if reproducible symptoms occurred upon repeated home ingestion. Written informed consent was obtained from all families or guardians.

Outcome definition

The minimum dose for performing the CM OFC before 2019 was 3 mL of CM (LD). However, 0.3 mL of CM (VLD) was subsequently introduced from 2019 onwards, when the OFC protocol at our center was revised. Since then, VLD OFC has been conducted based on the attending physician’s assessment in patients with a history of allergic reaction induction after ingesting < 3 mL of CM¹⁸ or those with high CM-sIgE levels.¹⁹,²⁶

The primary outcome in the LD and VLD groups was the duration from initiation of the OFC-based stepwise ingestion protocol (i.e., the first LD OFC before 2019 or the first VLD OFC after 2019) to achieve a negative full-dose OFC. The secondary outcome measures included an assessment of the safety of the initial LD and VLD OFC. Subgroup analysis within the VLD group was performed to assess the duration from the first VLD OFC to achieving negative results in the LD OFC (3 mL of CM) and medium-dose OFC (24 mL of CM). The analysis also included an evaluation of age at the time of showing VLD OFC negativity and duration from the first VLD OFC to the negative full-dose OFC. Participants in the VLD group were divided into two age categories (< 4 years and ≥ 4 years), based on previous reports suggesting that OIT is generally initiated from 4 years of age or older.²² In addition, patients were stratified by the median CM-sIgE level (< 20.3 kUA/L and ≥ 20.3 kUA/L) at baseline to assess the association between sensitization level and the time required to achieve a negative full-dose OFC result.

Laboratory tests

Serum levels of total IgE and sIgE to CM were measured using Immuno CAP^® (Phadia, Thermo Fisher Scientific, Waltham, MA, USA) 6 months before the first OFC.

Statistical analysis

Kaplan–Meier curves were used for descriptive analyses of the duration until LD, median-dose, and full-dose OFC negative results were obtained. Data were collected until August 31, 2024. The patients were considered lost to follow-up if they dropped out of the study. Notably, all patients were followed up from baseline until a negative full-dose OFC result or censoring. The baseline was defined as the date of initiation of the OFC-based stepwise ingestion protocol, which corresponded to the first LD OFC before 2019 or the first VLD OFC after 2019. Kaplan–Meier curves were analyzed from the date of baseline to the date of full-dose OFC negativity. The χ² test or Mann–Whitney U test was used to perform between-group comparisons of the characteristics; differences between categorical and continuous variables; and safety of the first OFC. The relationship between the factors and failure to achieve a full dose of OFC negative was assessed using the log-rank test and Cox proportional hazards regression for univariate and multivariate analyses, respectively. A logistic regression analysis was conducted, with the dependent variable set as failure to achieve a full-dose OFC negative. Significant factors from the univariate model were used as explanatory variables. The independent variables included: age at initiation of the OFC-based stepwise ingestion protocol, sex (male), past history of anaphylaxis to CM, atopic dermatitis, asthma, other FAs, total IgE levels, sIgE levels for CM, and membership in the LD group, which was adjusted accordingly. Statistical significance was set at P < 0.05. Numerical values are presented as numbers (%) and medians (interquartile range [IQR]). All analyses were performed using JMP Pro 17 (SAS Institute, Cary, NC, USA).

RESULTS

Baseline Characteristics

The LD and VLD groups included 359 and 318 patients, respectively. Excluding those who met the exclusion criteria, the analysis included 270 and 267 patients in the LD and VLD groups, respectively (Fig. 1). Table 1 presents the patient characteristics. The two groups showed no significant differences in background. The median OFC interval was 4 months (IQR, 3–6 months).

Fig. 1. Study flow chart.

LD, low-dose; CM, cow’s milk; OFC, oral food challenge; VLD, very-low-dose; OIT, oral immunotherapy.

Table 1. Characteristics at OFC-based stepwise ingestion protocol started.

Characteristics		LD group (n = 270)	VLD group (n = 267)	P value
Age (year)		3.6 (2.3–5.4)	3.4 (2.0–5.1)	0.107†
Male, Sex		179 (66)	158 (59)	0.093*
Past history of anaphylaxis to CM		111 (41)	121 (45)	0.326*
Allergic complications
	Atopic dermatitis	88 (33)	74 (28)	0.224*
	Asthma	52 (19)	48 (18)	0.701*
Total IgE level (IU/mL)		496 (246–1,057)	458 (195–1,062)	0.177†
CM-specific IgE (kUA/L)		21.9 (10.9–44.0)	20.5 (9.0–47.2)	0.791†

Variables are expressed number (%) or median (interquartile range).

OFC, oral food challenge; LD, low dose; VLD, very-low-dose; CM, cow’s milk; IgE, immunoglobulin E.

^*Statistics were performed using the χ² test.

^†Statistics were performed using the Mann Whitney U test.

Time to negative full-dose OFC

Fig. 2 shows the Kaplan–Meier curve for the negative rate of full-dose OFC. A significant difference was observed between the VLD and LD groups, with the VLD group having a higher negative rate for full-dose OFC (P = 0.023). Four years after initiation of the OFC-based stepwise ingestion protocol, the full-dose OFC negative rates were 45% and 60%, respectively. Additionally, 26 patients (10%) and 11 (4%) in the LD and VLD groups did not achieve an increased threshold, respectively, and were subsequently transitioned to conventional OIT, indicating a significantly higher transition rate to conventional OIT in the LD group (P = 0.014).

Fig. 2. Full-dose OFC negative rate.

OFC, oral food challenge; VLD, very-low-dose; LD, low-dose.

Safety

Table 2 presents the safety of the initial LD and VLD OFC in the LD and VLD groups, respectively. The positive rates in each group were 141 (53%) and 112 (42%) in the LD and VLD groups, respectively (P = 0.021). The rates of anaphylaxis induction and medication administration, including intramuscular adrenaline, were significantly lower in the VLD group.

Table 2. Safety of first OFC.

Characteristics		LD group (n = 270)	VLD group (n = 267)	P value
OFC positive rate		141 (53)	112 (42)	0.021
Anaphylaxis		25 (9)	13 (5)	0.041
Organ system affected
	Skin	79 (29)	69 (26)	0.378
	Respiratory	90 (33)	51 (19)	< 0.001
	Gastrointestinal	61 (23)	21 (8)	< 0.001
	Neurological	35 (13)	13 (5)	< 0.001
	Cardiovascular	11 (4)	2 (0.7)	0.009
Treatment
	Antihistamines	70 (26)	49 (18)	0.044
	Inhaled beta adrenergic agonists	66 (24)	23 (9)	< 0.001
	Intramuscular adrenaline	12 (4)	1 (0.3)	< 0.001

Variables are expressed number (%).

Statistics were performed using the χ² test.

OFC, oral food challenge; LD, low dose; VLD, very-low-dose.

Time to negative individual-dose OFC

Regarding the subgroup analysis, the graph depicting the time to achieve a negative result for VLD, LD, medium-dose, and full-dose OFC in the VLD group is shown in Fig. 3. The time for 50% of patients to achieve a negative result was 1.1, 2.6, and 3.1 years for low-, medium-, and full-dose OFC, respectively. The subsequent status of full-dose OFC negativity was evaluated longitudinally based on the age at which a negative result was confirmed for VLD OFC (Supplementary Fig. S2). The older the age at protocol initiation where a negative result was confirmed for VLD OFC, the less likely it was to achieve a negative result for full-dose OFC.

Fig. 3. Individual-dose OFC negative rate in the VLD group (n = 267).

OFC, oral food challenge; VLD, very-low-dose.

Time to negative full-dose OFC between < 4 and ≥ 4-year-olds

Furthermore, since conventional OIT is recommended to be initiated after 4 years of age,²² we divided the VLD group into two subgroups: those who underwent VLD OFC at < 4 years and those at ≥ 4 years. We analyzed the negative rate for full-dose OFC (Fig. 4). The results showed that the group that underwent VLD OFC at < 4 years had a significantly higher full-dose OFC negativity rate than did the group with undergone VLD OFC at ≥ 4 years (log-rank test, P < 0.001). Moreover, we examined the time to achieving a negative full-dose OFC result after dividing participants into two groups based on whether their CM-sIgE level was above or below the median. The analysis showed that participants with lower CM-sIgE levels had a significantly higher rate of achieving a negative full-dose OFC result compared with those with higher CM-sIgE levels (log-rank test, P < 0.001). These findings are presented in Supplementary Fig. S3. A multivariate analysis was conducted to identify risk factors that could inhibit a negative result in full-dose OFC. The analysis revealed that being in the LD group was the only significant factor (adjusted hazards ratio, 3.54 [2.82–5.87]; P < 0.001) (Supplementary Table S1). Atopic dermatitis was non-significant in the univariable analysis but showed a borderline association in the multivariable analysis (P = 0.053).

Fig. 4. Full-dose OFC negative rate by < 4- and ≥ 4-year-olds in the VLD group (n = 267).

OFC, oral food challenge; VLD, very-low-dose.

DISCUSSION

In this study, we investigated whether starting to intake VLD influenced the negativity rate in tolerance for CM allergy management. Our results indicated that introducing VLD OFC to patients with severe CM allergies, thereby starting CM intake early and avoiding complete elimination (early small ingestion [ESI]), effectively increased the threshold for CM and ultimately may contribute to tolerance acquisition. This finding may have important implications in the fundamental strategies for managing CM allergies in the future. In this study, the term “OFC” refers to a hospital-based diagnostic food challenge, whereas the “OFC-based stepwise ingestion protocol” represents a management strategy involving continued ingestion at home following a negative OFC result. The observed effects should therefore be interpreted as the impact of initiating CM intake through this stepwise protocol, rather than the effect of OFC itself.

The natural tolerance acquisition rate for CM allergy in Japan is reportedly 64% by age 5 years.²⁷ Other reports on the natural tolerance acquisition rate for CM allergy include 57% by age 3 years in Europe²⁵ and 50% by age 8.7 years in South Korea.²⁸ These reports do not account for severity; therefore, we infer that the natural tolerance acquisition rate for severe cases is likely to be lower. However, the full-dose OFC negative rate in the VLD group, which included more severe cases, was comparable to the natural tolerance acquisition rates reported in Japan and other countries. These results suggest that the ESI strategy may be applicable to a wide range of children, including those with severe CM allergy. Furthermore, it is particularly significant in patients with more severe FAs.

Although baseline characteristics appeared similar between the VLD and LD groups, the higher negative full-dose OFC rate observed in the VLD group may be partly explained by the fact that patients who underwent the VLD challenge avoided complete elimination and started intake at an earlier stage. Previous studies from Japan have suggested that the duration of complete elimination may influence tolerance acquisition.¹⁷,²⁹ Multivariate analysis results similarly indicated that being in the LD group was a significant factor. Globally, methods such as the milk ladder, which avoids complete elimination and progressively increases intake from small amounts, have been proposed.³⁰ These approaches have been shown to enhance threshold levels and tolerance acquisition rates.¹³,³¹ Symptom-free ingestion initiation of the causative food also increases this threshold, supported by conventional OIT studies.²² Furthermore, we have conducted similar analyses in patients with hen’s egg allergies and demonstrated that avoiding complete elimination and starting with a small intake at an early stage can induce tolerance.¹⁰ The guidelines in Japan also recommend that assessing the reintroduction of CM in all children exhibiting CM allergy is advised as prolonged total exclusion, which may lead to more persistent allergy and increased dietary and social exclusion.¹ Therefore, this study supports the importance for minimally necessary elimination. Nevertheless, this finding should be interpreted with caution due to the retrospective nature of the study, as natural tolerance acquisition cannot be fully excluded.

In the VLD group, a significant difference was observed in the induction of full-dose OFC negativity based on whether CM intake started before or after age 4. Reports have indicated that initiating small amounts of CM intake early in patients with a history of immediate reactions to it and in those requiring complete elimination can lead to early tolerance acquisition.³² Similar findings have been reported for peanut FAs, where early introduction has been effective.³³,³⁴ The results of this study suggest that starting CM intake at a VLD at a younger age may be associated with a shorter period of complete elimination and potentially be associated with earlier tolerance acquisition.

However, in patients whose VLD OFC results are positive and who must adhere to complete elimination, future research should likely focus on implementing OFC with even smaller amounts than VLD to further evaluate the efficacy of the ESI strategy. Additionally, the results of this study do not necessarily indicate that the patients have acquired complete tolerance to CM. This is due to several factors, including the possibility of allergic reactions when consuming amounts greater than those set in the full-dose OFC. A risk of allergic reactions may occur if CM is consumed after a specific period of avoidance or under conditions of stress, such as exercise or illness. These risks have been well documented in studies on conventional OIT.³⁵,³⁶ The patients in this study were originally those with severe CM allergies with a history of allergic reactions to small amounts of CM intake. The ESI strategy involves continuous intake repeatedly from small amounts and gradual increases while confirming the rise in threshold levels, which is similar to conventional OIT. Therefore, caution is required when confirming tolerance acquisition in these patients.

In contrast to previously published LD conventional OIT protocols for CM allergy,³⁷ our VLD OFC-based stepwise ingestion protocol is not intended as an immunotherapy procedure for inducing desensitization. Instead, it functions as a stepwise diagnostic and management method to minimize the period of complete elimination and to introduce very small, physiologically tolerable amounts of CM under controlled hospital settings. While both methods share the concept of gradual allergen exposure, conventional OIT typically involves daily ingestion of defined allergen doses aimed at immune modification, whereas our VLD OFC-based stepwise ingestion protocol serves primarily to identify a safe threshold and to maintain minimal continuous exposure within each individual’s tolerance level. Therefore, our intervention should be regarded as a distinct clinical strategy rather than a variant of LD conventional OIT.

We previously conducted a similar analysis in children with severe hen’s egg allergy and reported the findings.¹⁰ Comparing the data on severe hen’s egg allergy to that on severe CM allergy, the full-dose OFC negative rate for severe CM allergy was lower than the 95% observed for egg allergy. Notably, several factors might account for this difference.

First, children with CM allergies have lower natural tolerance acquisition rates than those with egg allergies.²⁸,³⁸ Additionally, the age at which natural tolerance is acquired is later for CM allergies.³⁹ These differences likely influence the observed outcomes. Second, the threshold for inducing symptoms is lower in children with CM allergy than in those with hen’s egg allergy. Reports comparing eliciting dose (ED)01 and ED05 for children indicate that ED01 and ED05 are 4.2 mg and 87.7 mg, respectively, for hen’s egg allergy. However, these values are 0.3 mg and 14.4 mg, respectively, for CM allergy.³⁹ Third, CM allergy is more likely to induce severe symptoms such as anaphylaxis than hen’s egg allergy.³⁹,⁴⁰,⁴¹ When small amounts of intake trigger severe symptoms, the fear of reintroducing the allergen can lead to prolonged periods of complete elimination.⁴² Compared with our previous study on hen’s egg allergy, this study on CM allergy involved patients who were, on average, over a year older and included more patients with anaphylaxis history to the causative food.¹⁰ These results support the tendency for longer periods of complete elimination in CM allergy cases. From these perspectives, acquiring tolerance to CM allergy appears to be more challenging. Additionally, previous studies on conventional OIT have shown lower desensitization and tolerance acquisition rates for CM than for hen’s egg,⁴³ suggesting that increasing the threshold is more difficult for CM allergy.

The importance of ensuring the safety of FA treatments, such as through OFC, has recently been emphasized in clinical practice.¹ Our comparison of the safety of initial LD OFC and VLD OFC demonstrated that VLD OFC poses a lower risk. This finding indicates the effectiveness of VLD OFCs and their greater utility compared with that of LD OFC. However, given that anaphylactic symptoms were induced in a certain number of patients, comprehensive risk management remains essential. Therefore, patients and their guardians must be fully informed about the risks associated with OFC.

This study has some limitations. First, the intervals between OFCs were non-uniform, based on individual judgments by the outpatient primary physicians. However, the impact on outcomes was considered minimal because the policy regarding the interval for progressing to the next OFC did not change at the research facility during the periods when the LD and VLD groups were created. Second, this study lacked detailed information on home ingestion frequency in all patients. The frequency of ingestion might have affected the threshold. However, since the guidance for ingestion frequency at home did not change between the two groups, the difference between the groups was considered minimal. Third, safety at home was not evaluated in all patients. Continuous ingestion was performed at home after confirming OFC negativity; however, allergy symptom induction, medication usage, and frequency of hospital visits were not evaluated. Fourth, this study is the lack of detailed information regarding adherence to at-home ingestion protocols, including the frequency of milk intake and the occurrence of adverse reactions. Because these factors may have influenced the long-term outcomes, the interpretation of the long-term effectiveness and safety should be made with caution. Fifth, the sensitization status in all patients in this study was not confirmed using skin prick test. In addition, this study did not measure casein-sIgE or CM- and casein-specific IgG4. Finally, this was a retrospective, single-center study. Therefore, standardizing patient backgrounds, including CM allergy severity and the age at which OFC was initiated, proves difficult. In addition, younger patients who initiated the protocol earlier may have included those who would have naturally acquired tolerance over time, which could not be fully accounted for in this retrospective analysis. Future research should involve prospective, multicenter studies that include various food items to determine whether similar management strategies could lead to an increase in the threshold.

In conclusion, this study revealed that introducing VLD OFC for early intake as part of the ESI strategy effectively reduced the period of complete elimination of CM in patients with severe CM allergies. This approach increases the likelihood of obtaining subsequent negative results for the full-dose OFC. Therefore, as a management strategy for patients with CM allergy, including severe cases, early implementation of the ESI strategy may facilitate tolerance acquisition.

SUPPLEMENTARY MATERIALS

Supplementary Table S1

Risk factor for failure to induce full-dose OFC negative in univariate and multivariate analyses

Supplementary Fig. S1

Cow’s milk OFC-based stepwise ingestion protocol.

Supplementary Fig. S2

Full-dose OFC negative rate by age in very-low-dose group (n = 267).

Supplementary Fig. S3

Full-dose OFC negative rate by < 20.3 and ≥ 20.3 kUA/L CM-sIgE in very-low-dose group (n = 267).