This study is a cohort studay conducted at Jilin Women And Children Health Hospital in China. The study population consists of newborns born at the hospital between September 2021 and June 2023. The inclusion criteria for study participants comprised neonates who had undergone genotyping for both the G211A polymorphism and the TATA box polymorphism of the UGT1A1 gene. The exclusion criteria encompassed neonates with substantial missing data, as well as those with congenital malformations or significant genetic diseases that could potentially confound the diagnosis of NHB. Clinical information of newborns and their mothers, as well as the genetic polymorphism results for the G211A and the TATA box, were collected from the hospital's medical records system. The study included a total of 3,258 newborns, out of which 372 were diagnosed with NHB, resulting in an incidence rate of 11.42%. The diagnosis of NHB in the study followed the criteria outlined in the Chinese consensus on the diagnosis and treatment of Neonatal hyperbilirubinemia (2014) (10), the specific operation was to measure the TcB value using a Japanese Minolta JM-103 transcutaneous bilirubin meter on a daily basis. Reference was made to the neonatal hour-specific bilirubin nomogram developed by Bhutani VK et al. in the United States. When the TcB value exceeded the 75th percentile of the nomogram, TSB was measured. When the TSB exceeded the 95th percentile of the nomogram, it was defined as hyperbilirubinemia. The study aims to analyze the relationship between clinical factors and UGT1A1 gene polymorphisms and the incidence of NHB in China. A binary logistic regression model is utilized to investigate the impact of clinical factors and UGT1A1 gene polymorphisms on the occurrence of NHB. This research has received approval from the hospital's ethics committee. Most of the newborns in the study were the Han nationality.

All neonates born at the Jilin Women And Children Health Hospital are advised by their doctors to undergo testing for the G211A locus and the TATA box polymorphism of the UGT1A1 gene. The neonates' guardians are free to choose whether or not to proceed with the testing. If guardians choose to proceed, they are required to sign a consent form after being informed of the purpose and procedures of the testing. After 72 h of birth, the newborn will undergo blood collection. Following DNA extraction, PCR amplification and sequencing are performed to determine the polymorphisms of the UGT1A1 gene at the G211A locus and the TATA box. This study solely focuses on the results of these polymorphisms in neonates.

Hemolytic disease of newborn, refers to neonates with a positive direct Coomb's test were diagnosed with hemolytic disease. In this study, the ABO blood type and Rh blood type of the neonate and their mother were considered. Neonatal weight loss, refers to a decrease in neonatal weight of more than 10% of the birth weight after birth. Neonatal intrauterine distress history is defined as the occurrence of hypoxia and acidosis in the fetus during intrauterine development. Neonatal intrauterine distress is primarily caused by various high-risk factors in the mother, fetus, or placenta, and is characterized by abnormal fetal heart rate, a series of metabolic and reactive changes, and a comprehensive presentation that endangers the life and health of the neonate.Neonatal premature rupture of membranes (PROM) history is defined as the rupture of the fetal membranes prior to delivery. In this study, rupture of the fetal membranes more than 18 h before delivery is considered premature rupture of membranes. Maternal gestational hypertension history is defined as the occurrence of hypertension in the mother during pregnancy. Gestational hypertension usually appears after 20 weeks of gestation and is characterized by symptoms such as high blood pressure and proteinuria. Maternal gestational diabetes history is defined as the diagnosis of diabetes mellitus in the mother during pregnancy. Maternal prenatal infection history is defined as the occurrence of infectious diseases in the mother prior to delivery. Common prenatal infections include urinary tract infections, reproductive tract infections, and respiratory tract infections. The diagnosis of these diseases or disease histories was made by doctors in the Neonatology Department of Jilin Women And Children Health Hospital.

We conducted the analysis using BONC DSS Statistics 25.0 data analysis software. For quantitative variables that followed a normal distribution, we presented statistics using mean ± standard deviation. For non-normally distributed quantitative variables, we used the median and interquartile range. Qualitative variables were described using frequency and composition percentages. To compare two groups of quantitative variables, we utilized independent samples t-test for normally distributed data and the Mann-Whitney U test for non-normally distributed data. For comparisons involving three or more groups of quantitative variables, we applied analysis of variance (ANOVA). Two groups of qualitative variables were compared using either the chi-squared test or Fisher's exact probability test. Furthermore, we performed multifactor analysis of NHB using a binary logistic regression model to explore influencing factors. A nomogram was constructed based on the binary logistic regression model results in the R programming language. Model performance was evaluated using ROC curves and the area under the curve (AUC). We considered a P-value less than 0.05 as statistically significant.

In order to explore the relationship between clinical factors and UGT1A1 gene polymorphism and the incidence of NHB in Chinese newborns, the collected factors were categorized into quantitative and qualitative variables for analysis. In the quantitative variables, the gestation age of the neonates ranged from 28 to 42 weeks, with an average of 39.12 ± 1.11 weeks. The birth weight of the neonates ranged from 1,790 to 5,470 g, with an average of 3,399.43 ± 396.40 g. The discharge age of the neonates ranged from 1 to 15 days, with a median of 5 days and a interquartile range of 1 day, The numbers of gestational ranged from 1 to 7 times, with a median of 1 time and a interquartile range of 1 time. The parity ranged from 1 to 3 times, with a median of 1 time and a interquartile range of 0 times. The numbers of abortions ranged from 0 to 6 times, with a median of 0 times and a interquartile range of 1 time. The time of breastfeeding initiation ranged from 0 to 3 h, with a median of 0.5 h and a interquartile range of 0.7 h. The time to first meconium ranged from 0.5 to 28 h, with a median of 8 h and a interquartile range of 10 h.The time when stool turns yellow ranged from 1 to 9 h, with a median of 3 h and a interquartile range of 1 h. The average daily frequency of stools ranged from 0 to 10 times, with a median of 3 times and a interquartile range of 1 time. Table 1 presents the differences in the quantitative variables between the neonatal hyperbilirubinemia group and the non-hyperbilirubinemia group. There were significant differences observed between the hyperbilirubinemia group and the non-hyperbilirubinemia group in terms of newborn gestational age, newborn discharge age, and the time when stool turns yellow (P ≤ 0.05). The results show that newborns with hyperbilirubinemia had a significantly lower mean gestational age (38.9 ± 1.1) compared to newborns without hyperbilirubinemia (39.2 ± 1.1, P ≤ 0.001). Furthermore, the hyperbilirubinemia group had a longer duration of hospitalization (P = 0.007) and a longer time for the stool to turn yellow (P ≤ 0.001). Table 2 illustrates the relationship between qualitative variables in clinical factors of newborns and their mothers and the occurrence of hyperbilirubinemia. Significant differences were observed between the hyperbilirubinemia group and the non-hyperbilirubinemia group in terms of maternal blood type, mode of delivery, neonatal cephalohematoma, hemolytic disease of newborn, neonatal weight loss, neonatal premature rupture of membranes (PROM) history, neonatal amniotic fluid status, and maternal gestational diabetes history. The incidence of hyperbilirubinemia in newborns with maternal blood types A (14.1%), B (10.8%), and O (17.4%) was significantly higher than in newborns with maternal blood type AB (8.7%) (P ≤ 0.001). The rate of hyperbilirubinemia in newborns delivered through natural labor (14.7%) and assisted delivery (30.2%) was significantly higher than in newborns delivered by cesarean section (10.9%) (P ≤ 0.001). Newborns with the following conditions, namely, neonatal cephalohematoma (68.7%), hemolytic disease of newborn (67.3%), neonatal weight loss (44.4%), and neonatal premature rupture of membranes (PROM) history (16.1%), exhibited significantly higher incidence rates of hyperbilirubinemia compared to newborns without these conditions (P ≤ 0.001). For maternal gestational diabetes history, we found that a total of 799 study subjects had a diabetes history.The incidence of hyperbilirubinemia in newborns whose mothers had a history of diabetes during pregnancy (13.9%) was higher than in newborns whose mothers had no history of gestational diabetes (P = 0.043). Additionally, newborns with clear amniotic fluid (13.6%) had a significantly higher incidence of hyperbilirubinemia than newborns with turbid amniotic fluid (9.5%) (P = 0.028).

To explore the relationship between UGT1A1 gene polymorphism and NHB. We primarily observed the genetic polymorphism results at two loci of the UGT1A1 gene, namely, the G211A polymorphism and the promoter TATA box (Table 3). The results demonstrate that genetic polymorphism at the G211A locus significantly increases the incidence of hyperbilirubinemia in newborns (P = 0.011). The incidence of hyperbilirubinemia is significantly higher in newborns with heterozygous and homozygous mutations at the UGT1A1 gene G211A locus (13.7%) compared to those with the wild type (10.2%) (P = 0.003). The variation in the TATA box does not significantly impact the incidence of NHB (P = 0.136). Joint analysis of both loci reveals that the incidence of hyperbilirubinemia in newborns with genetic polymorphism is higher than those wild type.

To further investigate the impact of genetic polymorphism at the G211A and TATA box loci on NHB in China, we collected transcutaneous bilirubin measurement values from the second to fifth day after birth. All newborns had their transcutaneous bilirubin values collected on the second day after birth. Due to factors like discharges and loss to follow-up, we were unable to collect transcutaneous bilirubin values for all newborns on the third to fifth day. Nonetheless, we analyzed the differences in transcutaneous bilirubin values from second to fifth day between different genotypes at the two loci and when both loci were jointly analyzed for the entire population, the non-hyperbilirubinemia group, and the hyperbilirubinemia group. We found that in the entire population and the non-hyperbilirubinemia group, significant differences in transcutaneous bilirubin values on the fourth day were observed between different genotypes at the two loci (Table 4). However, no differences were observed in the hyperbilirubinemia group. In the total population and non-hyperbilirubinemia group, the transcutaneous bilirubin levels of newborns with genotype AG and AA were significantly higher than those with genotype GG (P ≤ 0.001), the transcutaneous bilirubin measurement values of newborns with genotypes TA₆/TA₇ and TA₇/TA₇ were significantly lower than those of A₆/TA₆ (P ≤ 0.03). When both loci were jointly analyzed, newborns with the presence of any genetic polymorphism or both genetic polymorphisms had higher transcutaneous bilirubin measurement values compared to newborns without genetic polymorphism (P ≤ 0.02). The above differences were not found in newborns in the hyperbilirubinemia group.

We utilized binary logistic regression models to analyze the risk factors for NHB (Table 5). The model indicates that the following factors are associated with an increased risk of NHB: the time when stool turns yellow [P ≤ 0.001, OR 1.266 (95% CI: 1.125–1.425)]; neonatal cephalohematoma [P ≤ 0.001, OR 33.642 (95% CI: 21.823–51.861)]; hemolytic disease of newborn [P ≤ 0.001, OR 33.849 (95% CI: 18.589–61.636)]; neonatal weight loss [P ≤ 0.001, OR 11.275 (95% CI: 7.842–16.209)]; neonatal premature rupture of membranes (PROM) history [P = 0.021, OR 1.422 (95% CI: 1.056–1.917)]; genetic polymorphism at the UGT1A1 gene G211A locus. Gestational age is a protective factor [P ≤ 0.001, OR 0.766 (95% CI: 0.686–0.855)]. Compared to natural labor, cesarean section is a protective factor [P = 0.011, OR 0.711 (95% CI: 0.546–0.926)], while assisted delivery is a risk factor [P = 0.022, OR 2.207 (95% CI: 1.121–4.346)]. Figure 1 presents a nomogram and the ROC curve of the NHB risk prediction model based on the analysis of risk factors. The nomogram (Figure 1A) includes factors such as neonatal gestational age (Weeks), the time when stool turns yellow (Days), mode of delivery (Delivery, CS, Caesarean Section; NL, Natural labor; AD, Assisted Delivery), neonatal cephalohematoma (Cephalohematoma), hemolytic disease of the newborn (HDN), neonatal weight loss (Weightloss), neonatal premature rupture of membranes history (PROM), and genetic polymorphism at the G211A locus (Genetype). Each line on the nomogram is scaled according to the range of values for each factor. The length of each line reflects the contribution of that factor to the risk of NHB. By adding up the scores for each factor, a total score is obtained, and using this total score, a vertical line is drawn to estimate the corresponding probability of developing hyperbilirubinemia. The ROC curve is used to evaluate the binary logistic regression model (Figure 1B). The points on the ROC curve represent different cut-off points. The area under the ROC curve (AUC) indicates the model's ability to distinguish between patients and non-patients. A larger AUC suggests a stronger ability to discriminate and more accurate predictions. In this model, the AUC is 0.804(95% CI: 0.777–0.831), indicating that the model has good discrimination ability and is valuable for predicting the risk of NHB after birth.

There are certain limitations to this study. Firstly, the sample size is limited, and the development of the prediction model was conducted in a single cohort. Therefore, the generalizability of the study results has certain limitations. Secondly, unfortunately, there were cases of neonates being discharged or lost to follow-up, which led to incomplete collection of transcutaneous bilirubin values on the third to the fifth days. Thirdly, in our study, the diagnosis of NHB was based on the hour-specific bilirubin nomogram developed by Bhutani VK et al. However, it is important to note that the accuracy of this nomogram may be limited as it did not consider the impact of ethnicity and other high-risk factors. This limitation underscores the necessity of exploring risk factors for NHB within different ethnic groups. Finally, this study used transcutaneous bilirubin measurements rather than conjugated bilirubin measurements, let alone unconjugated bilirubin measurements that can penetrate the neonatal blood-brain barrier and have direct neurotoxicity. These shortcomings have some limitations in explaining the impact of UGT1A1 gene polymorphism on the incidence rate of NHB in China, especially the impact of TATA box polymorphism. We believe that with the deepening of research and technological progress, the impact of more clinical risk factors and gene polymorphisms on NHB will become increasingly clear.