This trial will be carried out in two stages in the sick baby room of the Children's Hospital of China Medical University (CMUH). The first stage is: the enzyme active reaction kit (usually called API ZYM kit) was used to determine the β-glucuronidase activity of 9 strains of Lactobacillus and 4 strains of Bifidobacterium, screening out the probiotics that have the best inhibitory efficacy of intestinal β-glucuronidase activity; The second stage is: using the results of the first stage analysis, treating the newborns of jaundice by oral probiotics.

Hyperbilirubinemia is one of the common diseases in neonates. According to the literature, the incidence of neonatal jaundice is around 60~70% in Western countries, and even higher among newborns of Asian ethnicity. There are many causes of neonatal hyperbilirubinemia, one of the important mechanisms is the lack of intestinal microbial flora, which causes high level of conjugated bilirubin to be converted to unconjugated bilirubin via β-glucuronidase in the intestine. The high level of unconjugated bilirubin will be absorbed or returned to the blood via the enterohepatic circulation, resulting in jaundice. The main treatment of neonatal jaundice is phototherapy: unconjugated bilirubin is converted to a non-toxic isomer by blue light (wavelength 425-457 nm).

At present, it has been confirmed that Escherichia coli (E. coli) in the intestine of animals produces two harmful bacterial enzymes, β-glucosidase and β-glucuronidase. Therefore, in recent years, scientists have used probiotics that inhibit the growth characteristics of pathogenic bacteria to study the treatment of neonatal jaundice. Chen Yi-Ji et al. have shown that probiotics inhibit the activity of intestinal β-glucuronidase, preventing the conversion of conjugated bilirubin to unconjugated bilirubin, thereby reducing the level of unconjugated bilirubin in the blood. Funda Tuzun et al.'s research indicates that bifidobacteria probiotics prevent breast milk jaundice, and probiotics prevent jaundice by increasing intestinal peristalsis and microbial flora. In addition, many studies have shown that Lactobacillus and Bifidobacterium species have inhibitory efficacy on intestinal β-glucuronidase activity, but there is no specific indication of which strain has the best inhibitory efficacy. Based on the above research, the question investigators need to clarify is whether oral probiotics can correct the intestinal microbial flora of newborns of jaundice? Is there any difference in the improvement of neonatal jaundice by the addition of probiotics in breast milk, formula or mixed milk? Therefore, this study will first screen out the probiotics that have the best inhibitory efficacy of intestinal β-glucuronidase activity. In the second stage, the results of the first stage will be used to perform a neonatal jaundice oral probiotic double-blind randomized trial and the therapeutic efficacy will be observed.

This trial will be carried out in two stages in the sick baby room of the Children's Hospital of China Medical University (CMUH). The first stage is: the API ZYM kit was used to determine the β-glucuronidase activity of 9 strains of Lactobacillus and 4 strains of Bifidobacterium, screening out the probiotics that have the best inhibitory efficacy of intestinal β-glucuronidase activity; The second stage is: using the results of the first stage analysis, treating the newborns of jaundice by oral probiotics. The inclusion criteria of jaundice infants were as follows: the full-term infant (≧37 weeks) had a jaundice index greater than 15 mg/dl on the fourth day after birth, while neonates with hypothyroidism, trisomy 21, maternal blood type A, B and O incompatibility, gastrointestinal disease, Glucose-6-Phosphate Dehydrogenase deficiency (G6PD deficiency), hemangioma, cephalhaematoma or hemorrhages, severe asphyxia (stage III), fetal chromosomal anomalies, cyanotic congenital heart disease, omphalocele, early onset sepsis or liver failure were excluded. The second stage of jaundice newborns will be divided into three groups: (I. experimental group) light therapy + Lactobacillus salivarius AP-32 (L. salivarius AP-32), (II. experimental group) light therapy + Bifidobacterium animalis subspecies lactis CP-9 (B. animalis subsp. lactis CP-9), (III. control group) light therapy + placebo. The study will be conducted for 7 days, the serum bilirubin level will be measured by the same group of experienced nurses every day, and all babies are monitored for adverse conditions such as vomiting, diarrhea or bloating.

Calculation of sample size in the second stage: According to Chou Hung-Chieh et al.'s research, the incidence of neonatal hyperbilirubinemia was 30.5% for a concentration ≧15 mg/dl. investigators assume that the experimental group has a 50% improvement compared with the control group. If investigators allow 5% chance of type 1 error (α=0.05) and 10% chance of type 2 error(β=0.1), then the required sample size in each group will be 60 people (total 180 babies).

The research data will be based on statistical software (called IBM SPSS) for analytical statistics. Normally distributed data were analyzed using Student's t-test two-tailed assay, and non-normally distributed data were analyzed using the Wilcoxon signed-rank test. Data were presented as mean ± standard deviation (SEM) or median, with P values < 0.05 being considered statistically significant.

Inclusion Criteria:

1. Full-term infants (≧37 weeks).
2. Jaundice index greater than 15 mg/dl on the fourth day after birth.

Exclusion Criteria:

1. Hypothyroidism
2. Trisomy 21
3. Maternal blood type A, B and O incompatibility
4. Gastrointestinal disease
5. Glucose-6-Phosphate Dehydrogenase deficiency (G6PD deficiency)
6. Hemangioma
7. Cephalhaematoma or hemorrhages
8. Severe asphyxia (stage III)
9. Fetal chromosomal anomalies
10. Cyanotic congenital heart disease
11. Omphalocele
12. Early onset sepsis
13. Liver failure