Link Between History of Neonatal Jaundice and Early Childhood Allergic Diseases in South Korea

Authors: Seong Wan Kim, Yoong‐A Suh, Seoheui Choi, Moon Sung Park, O Kyu Noh, Jang Hoon Lee

Abstract

Aim

Allergic diseases, such as asthma, allergic rhinitis (AR), and atopic dermatitis (AD) are prevalent chronic childhood conditions, with a rising incidence globally and in South Korea. Neonatal jaundice is a common perinatal condition. However, its association with the subsequent development of allergic diseases remains unclear.

Methods

This retrospective population‐based cohort study used data from the Korean Health Insurance Review and Assessment Service database. Participants were categorised according to the history and severity of neonatal jaundice, defined by treatment status. Logistic regression and Cox proportional hazard models were used to estimate the risk of neonatal jaundice.

Results

Children with neonatal jaundice showed a moderately increased risk of asthma and AR, particularly those who received treatment. Neonatal jaundice requiring treatment showed significantly elevated odds ratios for asthma and AR.

Conclusion

Neonatal jaundice requiring treatment was associated with an increased risk of asthma and AR in early childhood. This suggests that the severity of neonatal jaundice may influence the later development of allergic diseases and highlights the need for closer follow‐up of this population.

Summary In this nationwide cohort study, infants with a higher severity of neonatal jaundice who required treatment were associated with significantly higher risks of asthma and allergic rhinitis by age six.The findings underscore the need for early monitoring and screening for asthma and AR in children with a history of severe neonatal jaundice.Future prospective studies incorporating objective bilirubin levels and phototherapy intensity are warranted.

Introduction

Allergic diseases, including asthma, allergic rhinitis (AR), and atopic dermatitis (AD), are common chronic conditions in early childhood that have shown a rising global prevalence [1]. In South Korea, the prevalence of AR increased from 31% in 1995 to 43.6% in 2015, whereas that of AD increased from 8.8% to 12.7% during the same period [2]. These conditions not only impair quality of life in childhood but can also lead to the progression of other allergic diseases—a phenomenon known as the “allergic march” [3]—and contribute to long‐term health issues in adulthood. Therefore, identifying the underlying pathogenesis and early‐life risk factors is essential for prevention and timely interventions. The perinatal period is a critical window for immune system development [4]. During this time, genetic predisposition and environmental exposure can significantly influence immune regulation and susceptibility to allergic diseases later in life [5]. Previous studies have identified a variety of perinatal risk factors or protective factors for allergic diseases, including breastfeeding [6], family history of allergy [6, 7], administration of systemic antibiotics [8, 9], and history of neonatal respiratory illnesses [10, 11].

Neonatal jaundice is one of the most common clinical conditions in newborns, particularly within the first week of life [12]. Although most cases are mild and resolve without intervention, severe hyperbilirubinemia can result in kernicterus and long‐term neurological complications. To prevent such outcomes, infants with clinically significant neonatal jaundice are treated with phototherapy and, in more extreme cases, exchange transfusion. According to the American Academy of Paediatrics guidelines, the decision to initiate treatment is based on the infant's total serum bilirubin level, age in hours, gestational age, and other risk factors [13]. Thus, the treatment status serves as a reliable surrogate marker of neonatal jaundice severity.

Several recent studies have suggested a potential association between neonatal jaundice and later development of allergic diseases. Aspberg et al. reported that phototherapy‐treated neonates had higher risks of hospitalisation for asthma and asthma diagnoses before the age of 12 years [14, 15]. Other population‐based studies from the United States [16] and Taiwan [17, 18] also explored this relationship using national healthcare databases.

Despite the increasing prevalence of childhood allergic diseases in South Korea and the high incidence of neonatal jaundice, no previous studies have investigated the relationship between these two diseases in the Korean population. Therefore, using comprehensive nationwide health insurance data, this study aimed to evaluate whether a history of neonatal jaundice and its severity, reflected in the treatment status, are associated with an increased risk of developing childhood allergic diseases (asthma, AR, and AD).

Methods

Data Source

Our study utilised healthcare data provided by the Health Insurance Review and Assessment Service (HIRA) of South Korea, which contains comprehensive medical service records for all individuals with a registered national identification number. The HIRA database is based on the National Health Insurance System and includes detailed information such as diagnoses, procedures, prescriptions, and healthcare utilisation at all levels of care. Notably, this dataset covers the entire population of South Korea and allows for the analysis of health service use from birth registration to death. The data were anonymised and customised according to the study objective before being made available for research. We have utilised the HIRA research data (M20240805001).

Study Population

To utilise contemporary data and ensure the availability of healthcare records up to the age of 6 years, the authors selected a cohort of children born in 2015, thereby minimising potential confounding by temporal trends. Additionally, to capture all relevant medical services used during the perinatal period, we ensured that the dataset obtained covered the full spectrum of healthcare utilisation from birth. Consequently, the HIRA provided data on 35 727 individuals who met the inclusion criteria and matched the study objectives.

We excluded children with risk factors for allergies in the postnatal period, as identified in the HIRA database. For instance, children with respiratory illnesses and those with infectious diseases who received systemic antibiotics were excluded from the study. Furthermore, children with a documented history of hospitalisation in the neonatal intensive care unit for other systemic diseases, with the exception of neonatal jaundice, were excluded from the study in order to include only healthy children without the influence of other diseases. After their exclusion, 29 690 participants were finally included in the study. Among them, 10 102 were identified as having a history of neonatal jaundice. Among them, 7418 improved without medical intervention, whereas the remaining 2684 received treatment with either phototherapy or exchange transfusion (Figure 1).

Outcomes

The primary outcome was the diagnosis of each allergic disease by the age of 6 years. Based on previous studies using national databases similar to HIRA [14, 15, 17, 18, 19], we minimised overdiagnosis by including only patients whose allergic disease was recorded as the primary diagnosis in at least three outpatient visits or at least one inpatient admission. Diagnoses were made by paediatricians and coded according to the Korean Standard Classification of Diseases (KCD) in the HIRA database. As a secondary outcome, the cumulative incidence of each allergic disease by the age of 6 years was compared.

Data Collection

To identify other risk factors for allergic diseases in the perinatal period, we categorised the study participants according to sex, preterm birth, small for gestational age, large for gestational age (LGA), the history of affected by maternal disease such as preeclampsia, gestational diabetes mellitus, polyhydramnios and oligohydramnios, and multiple births using the diagnosis code in the Korean Standard Classification of Diseases (KCD). However, other important risk factors, such as a family history of allergic diseases and mode of delivery (e.g., caesarean section)—could not be included in our analysis because the HIRA database does not provide information on participants' family history. Procedure codes for phototherapy and exchange transfusion were used to categorise the participants according to neonatal jaundice severity. The diagnostic and procedural codes used in this study are presented in Table 1. Data on hospital visits were used to obtain additional information regarding allergic diseases.

Statistical Analysis

All variables in this study were categorical and presented as frequencies and percentages. Group comparisons were performed using Pearson's chi‐squared test. Statistical significance was defined as a p‐value < 0.05. Prior to the main analysis, multicollinearity among the confounding risk factors in the perinatal period was assessed. Pairwise correlation coefficients (R‐value) were calculated, and any variable pairs with R‐value ≥ 0.7 were identified. Variance inflation factors (VIFs) were also examined, and all confounding factors had VIFs < 5, confirming the absence of multicollinearity. The risk of allergic disease diagnosis was analysed using binomial logistic regression with adjusting for the confounding risk factors. Cumulative prevalence was assessed using survival analysis with Kaplan–Meier curves and log‐rank test, with the censoring set at the age of 6 years. Final estimates were presented as hazard ratios (HRs) from the Cox proportional hazards models, and the proportional hazards assumption was tested for all significant covariates. All statistical analyses were performed using the R software.

Ethical Considerations

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of Ajou University Hospital (Protocol AJOUIRB‐EX‐2024‐373 and 2024‐06‐18). The study was conducted without obtaining individual informed consent. Under Article 16(3) of the Korean Bioethics and Safety Act, the requirement for consent was waived because obtaining consent from such a large cohort was impracticable; the analyses used fully anonymized data, and the research posed minimal risk to participants.

Results

When comparing the perinatal variables between children with and without a history of neonatal jaundice, we found that those with jaundice were predominantly male and had lower incidences of preterm births, LGA, and maternal diseases. A similar trend was observed when groups categorised according to neonatal jaundice severity and treatment status were compared (those without a history of neonatal jaundice, those with a history of neonatal jaundice and kept under observation, and those with a history of neonatal jaundice that required treatment). Significant differences were found among the three groups in terms of sex and the incidence of LGA and maternal diseases. In addition, the incidence of preterm births was relatively higher among those with a history of neonatal jaundice that required treatment (Table 2).

In the logistic regression analysis of the risk of early childhood allergic diseases, children with a history of neonatal jaundice demonstrated relatively higher risks of asthma and AR than did those without jaundice; however, these associations were not significant. However, when the children were stratified by neonatal jaundice severity and treatment status, a significantly higher risk of asthma was observed in those who did not receive treatment (OR 1.12, 95% CI: 1.05, 1.23, p < 0.001) and an even higher risk was noted in those who received treatment (OR 1.18, 95% CI: 1.10, 1.26, p < 0.001) Similarly, a significant increase in the risk of AR was observed in children who were treated for neonatal jaundice (OR 1.08, 95% CI: 1.01, 1.15, p = 0.023). No significant association was observed between the history of neonatal jaundice and risk of AD (Table 3).

In the survival analysis of the cumulative incidence of asthma, although the Cox proportional hazards model showed a trend toward an increased risk in the group with a history of neonatal jaundice, the result was not significant. However, when stratified by treatment status and neonatal jaundice severity, both untreated (OR 1.10, 95% CI: 1.04, 1.16, p < 0.001) and treated children (OR 1.16, 95% CI: 1.09, 1.23, p < 0.001) with a history of neonatal jaundice showed significantly higher HRs compared with those without such a history. The Cox model confirmed an increased risk of AR among the children with a history of neonatal jaundice (OR 1.04, 95% CI: 1.01, 1.07, p = 0.018). Stratified analysis according to treatment status further demonstrated a significantly elevated HR in the treated group. Consistent with the logistic regression results, no significant association was observed between the history of neonatal jaundice and risk of AD in the survival analysis (Table 4).

Discussion

This study investigated the association between the history of neonatal jaundice and subsequent development of childhood allergic diseases, including asthma, AR, and AD. Overall, the history of neonatal jaundice was not significantly associated with the development of allergic diseases in childhood. However, stratified analyses revealed a significantly increased risk of asthma and AR in both the untreated and treated children with a history of neonatal jaundice. Although the odds ratio indicates only a modest increase in risk, the low p value confirms this association is unlikely due to chance, consistent with a mild–to–moderate effect size.

These findings are consistent with those of several previous studies, suggesting a link between neonatal hyperbilirubinemia and risk of allergic diseases. As mentioned previously, Aspberg et al. found that receipt of phototherapy was associated with higher rates of asthma diagnosis [15] and hospitalisation [14]. Wei et al. also reported an increased risk of allergic diseases in patients with a history of neonatal jaundice in a survival analysis [19]. However, Kuzinewicz et al. found that phototherapy itself did not increase the risk of asthma. When comparing the total serum bilirubin levels, the risk increased when the level was 9–18 mg/dL and was the highest when it was in the 15–18 mg/dL range [16]. Huang et al. also analysed data from 1959 to 1965, when phototherapy was unavailable, and found that the risk of asthma increased with a maximum total serum bilirubin level ≥ 6 mg/dL, with the greatest increase at ≥ 15 mg/dL [20]. Considering these findings, elevated serum bilirubin levels, rather than phototherapy itself, may contribute to increased susceptibility to allergies, and the results of the present study can be interpreted in the same way.

Compared with previous studies, the risk estimates for allergic diseases observed in the present study are relatively moderate. However, in a meta‐analysis by Kuniyoshi et al. that included the abovementioned studies, the pooled risk was not significantly increased with a history of neonatal jaundice or receipt of phototherapy [21]. Notably, in contrast to many previous studies, our analysis was restricted to otherwise healthy children after excluding those with other systemic diseases that could affect allergic outcomes during the neonatal period. This methodological difference may partly explain why previous studies have reported comparatively higher risk estimates associated with a history of neonatal jaundice and receipt of phototherapy.

However, the relationship between the history of neonatal jaundice and risk of AD was not confirmed in the present study. Although some studies have reported an increased risk of AD associated with a history of neonatal jaundice or receipt of phototherapy [19, 22], Tham et al. found no significant association [23], and Ku et al. reported a decreased risk of AD following phototherapy [24]. Phototherapy is occasionally used to treat AD, although the wavelengths and therapeutic protocols used differ substantially from those used for treating neonatal jaundice [25]. These findings suggest that not only the severity of neonatal jaundice but also the receipt of phototherapy itself may influence the development of AD. Given this conflicting evidence, further investigations are warranted to clarify this relationship.

Although the underlying mechanisms linking the occurrence of neonatal jaundice to the later development of allergic diseases remain unclear, several hypotheses have been proposed. Unconjugated bilirubin may suppress interleukin‐2, a cytokine that plays a critical role in regulating the balance between T helper cell (Th) subsets, thereby shifting the immune response toward a Th2‐dominant profile rather than a Th1 response [26, 27]. In neonates, Th1‐mediated immunity is limited because of developmental immaturity, and Th2 polarisation induced by bilirubin may enhance immunoglobulin E‐mediated allergic sensitisation [28]. On the other hand, bilirubin also possesses antioxidant properties [29], which can reduce oxidative stress and suppress inflammatory pathways, potentially exerting protective effects against allergic diseases [30]. Considering these dual roles of bilirubin, further research is required to elucidate its role in the immunopathogenesis of allergic conditions.

This study has some limitations. First, as this was an observational study, causal inferences could not be made. Second, the severity of neonatal jaundice was approximated based on the data on treatment status, which may not fully reflect the clinical spectrum of hyperbilirubinemia. The HIRA dataset used in this study did not contain data on bilirubin levels, and the information on the initiation time, duration, or intensity of phototherapy was also unavailable, preventing their inclusion in our analyses. Although the treatment status serves as a proxy for disease severity, it does not fully capture it. Third, data on important potential confounding factors, such as family history of allergy, environmental exposure, emergency or elective delivery, and breastfeeding, were not available in the dataset. Fourth, as this study aimed to utilise medical service records starting from the first day of life, data were obtained only for children assigned a national identification number immediately after birth by the HIRA. Therefore, children with delayed registration, possibly owing to early postnatal medical complications, were excluded from the analysis. Although this may have introduced selection bias, we specifically aimed to investigate the association between the history of neonatal jaundice and the development of allergic diseases in otherwise healthy children, and this selection criterion was consistent with the study's objective. Methods such as inverse probability weighting (IPW) are commonly used to adjust for selection bias [31]; they require data on both included and excluded individuals to estimate the probability of inclusion [32]. In this study, IPW could not be applied because the dataset was preprocessed and provided by the HIRA, and no information was available for the children who were excluded owing to delayed registration. Therefore, appropriate modelling for IPW adjustment was not feasible.

Despite these limitations, the strengths and implications of the present study are noteworthy. First, even after applying the aforementioned exclusion criteria, a large number of participants were included in the analysis. Second, the HIRA database contains complete healthcare utilisation data from birth onwards, ensuring that there was no loss to follow‐up until the age of 6 years. Third, despite the absence of certain clinical variables, the study accounted for a wide range of perinatal confounding factors and excluded children with a history of infection or respiratory diseases to avoid independent effects on the development of allergic diseases. To the best of our knowledge, this is the first population‐based study to evaluate the association between the history of neonatal jaundice and the risk of allergic diseases in Korea. These findings may provide useful insights for clinicians monitoring children with a history of neonatal jaundice, especially those who require treatment, by informing them of follow‐up care and early screening for allergic diseases, such as asthma and AR.

In conclusion, our findings suggest that a history of neonatal jaundice requiring treatment is associated with moderately increased risks of asthma and AR in early childhood. Based on these results, clinicians should consider monitoring children with a history of neonatal jaundice, especially those who require treatment for the potential development of allergic diseases during follow‐up, which may help guide anticipatory care and early interventions. Further prospective cohort studies with detailed clinical information such as family history of allergic diseases and immunological data such as serum IgE or Interleukin‐2 are warranted. In particular, future research should incorporate objective measurements, such as bilirubin levels, duration and intensity of phototherapy, and additional biomarkers of immune function, to better elucidate the causal relationship between the history of neonatal jaundice and allergic disease development.

Author Contributions

Seong Wan Kim: conceptualisation, data curation, formal analysis, investigation, methodology, writing – original draft preparation. Yoong‐A Suh: writing – review and editing. Seoheui Choi: writing – review and editing. Moon Sung Park: writing – review and editing. O Kyu Noh: supervision, data curation, formal analysis, investigation, methodology, writing – review and editing. Jang Hoon Lee: supervision, conceptualisation, methodology, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.