In order to develop an approach to the diagnosis and management of the jaundiced newborn, it is necessary to understand the nonpathologic factors that can affect bilirubin levels in the normal newborn infant as well as the natural history of neonatal bilirubinemia. Many factors have been identified in large epidemiologic studies as having some effect on neonatal bilirubin levels,1 but their clinical relevance is often questionable. Those that have been shown in recent studies to have an important influence on total serum bilirubin (TSB) levels are listed in Table 6-1.

Mean maximum TSB concentrations in East Asian, Native American, and some Hispanic infants (primarily those of Mexican descent) are significantly higher than those in white infants.2–6 In a study of Hispanic infants, 31% had peak TSB levels >15 mg/dL6 compared with 3–10% of infants in other US populations.7,8 The mechanisms responsible for these differences are unknown, although there is some evidence that in the Native American population, increased bilirubin production plays a role.5 Black infants in the United States and Great Britain have lower TSB levels than white infants.3,9–11

Neonatal jaundice runs in families. Khoury et al.12 studied a population of 3301 newborns born to male US army veterans between 1966 and 1986. If one or more previous siblings had a TSB >12 mg/dL, the subsequent sibling was three times more likely than controls (10.3% vs. 3.6%) to develop a TSB >12 mg/dL, and if a prior sibling had a TSB level >15 mg/dL, the risk in the subsequent sibling was increased 12.5-fold (10.5% vs. 0.9%). These relationships applied whether or not the siblings were breastfed or formula fed. The familial nature of hyperbilirubinemia has also been documented in Chinese and Danish infants.13,14

The genetics of neonatal jaundice and the inborn errors of hepatic bilirubin uridine diphosphate glucuronosyltransferase (UGT) expression are discussed in detail in Chapter 1. Gilbert’s syndrome is not generally classified as a pathologic entity in a newborn, but newborn infants who are heterozygous or homozygous for Gilbert’s syndrome (i.e., they have the variant UGT genotype) have significantly elevated bilirubin levels in the first 2–4 days, compared with homozygous normal infants.15,16 In addition to a decreased ability to conjugate bilirubin, infants with Gilbert’s syndrome also appear to have an increase in red cell turnover and therefore bilirubin production.16 In a population of Scottish, primarily breastfed newborns, with TSB levels of >5.8 mg/dL after 14 days of life, 31% were homozygous for the 7/7 Gilbert’s syndrome promoter genotype compared with only 6% of a control group that did not have prolonged hyperbilirubinemia.17 By itself, the presence of Gilbert’s syndrome has a relatively modest effect on TSB levels in the neonate15 but, when combined with other icterogenic factors, Gilbert’s syndrome plays a ubiquitous role in the pathogenesis of neonatal hyperbilirubinemia. When the Gilbert’s genotype is combined with other icterogenic factors such as breastfeeding,17,18 G6PD deficiency,19 ABO incompatibility,20 and pyloric stenosis,21 there is a dramatic increase in the newborn’s risk for hyperbilirubinemia. In a striking example of the contribution of genetic mutations to hyperbilirubinemia in otherwise healthy infants, Huang et al.22 studied 72 Taiwanese infants with TSB levels ≥20 mg/dL. The factors identified as contributing to hyperbilirubinemia in this study included a genetic polymorphism of the organic anion transporter protein OATP-2, a coding sequence gene polymorphism for the hepatic bilirubin conjugating enzyme uridine diphosphate glucuronosyltransferase 1A1 (UGT1A1), and breastfeeding. By itself, breastfeeding was associated with an odds ratio (OR) of 4.6 for the risk of developing a TSB ≥20 mg/dL. The combination of the OATP-2 gene polymorphism with a variant UGT1A1 gene at nucleotide 211 increased the OR to 22, and when these two genetic variants were combined with breastfeeding, the OR was 88.22 As genetic testing is not routinely performed in jaundiced newborns, it is difficult to know what, if any, contribution these polymorphisms make to the individual jaundiced infant. Nevertheless, it is clear that in the presence of a certain genotype, breastfeeding will contribute to the development of marked hyperbilirubinemia.23

Smoking

Some studies suggest that infants of mothers who smoke during pregnancy have lower TSB levels than infants of nonsmokers,9,24 but others have not found this.25,26 These data are confounded by the fact that women who smoke during pregnancy are less likely to breastfeed their infants (OR 0.45, 95% CI 0.31–0.64).27

Diabetes

Infants of insulin-dependent diabetic mothers (IDM) are more likely to become jaundiced than are control infants28–32 and hyperbilirubinemia appears to be most prominent in macrosomic newborns29 and those with an increased birth weight/length ratio.31 With better diabetic control, as is currently the norm in developed countries, the risk of hyperbilirubinemia in this population should decrease.29 The hyperbilirubinemia in these infants is most likely the result of an increase in bilirubin production, which is directly related to the degree of macrosomia.33 Mean carboxyhemoglobin levels (an index of heme catabolism and bilirubin production) were 1.5 ± 0.19% (SD) in IDMs who were large for gestation (LGA) compared with 1.10 ± 0.27% in those who were appropriate for gestation (AGA) and 1.19 ± 0.33% (P < .05) in control infants.33 IDMs also tend to have higher hematocrits30,31 and increased erythropoietin levels in cord blood.34 Diabetic mothers have three times more β-glucuronidase in their breast milk than nondiabetic mothers.32 This enzyme enhances the enterohepatic reabsorption of bilirubin35 (see section “Breastfeeding and Jaundice”).

Drugs

Induction and Augmentation of Labor by Oxytocin

Multiple studies and several controlled trials have shown an association between the use of oxytocin to induce or augment labor and an increased incidence of neonatal hyperbilirubinemia,36 although the mechanism for this is unclear. An association has also been found between the total dose of oxytocin used for induction and the incidence of neonatal hyperbilirubinemia.37

Anesthesia and Analgesia

Anesthetic and analgesic agents readily cross the placenta and produce measurable levels in the newborn38 and the use of bupivacaine in epidural anesthesia has been associated with neonatal jaundice.26 Bupivacaine reduces red blood cell filterability in vitro and red cell survival in the rat so that increased bilirubin production could be the explanation for the described association between neonatal jaundice and maternal bupivacaine administration.39

Phenobarbital

Phenobarbital is a potent inducer of microsomal enzymes and increases bilirubin conjugation and excretion as well as bile flow.40 When given in sufficient doses to the mother, the infant, or both, phenobarbital is effective in lowering serum bilirubin levels in term and preterm infants in the first week of life,40–42 but concerns about long-term toxicity in the infant and even adults militate against the use of phenobarbital in pregnant women for this purpose.43,44

Other Drugs

The TSB concentrations at 48 and 72 hours are lower in infants of mothers who have received narcotic agents, barbiturates, aspirin, chloral hydrate, reserpine, and phenytoin, while diazepam and oxytocin lead to higher levels.45 Infants of heroin-addicted mothers46 have lower bilirubin levels as do the infants of mothers who received the analgesic and antipyretic, antipyrine, before delivery.47 Antenatal betamethasone, when used to accelerate lung maturation, did not increase neonatal TSB levels48 but antenatal dexamethasone was associated with an increased incidence of TSB levels >15 mg/dL in a group of preterm infants.49

Beta-agonists used for tocolysis do not affect carbon monoxide (CO) production in preterm neonates of mothers who receive these drugs50,51 and were not associated with higher bilirubin levels.51–53

Delivery Mode

In one study, breastfed, vaginally delivered, term newborns had higher TSB and transcutaneous bilirubin (TcB) levels than those delivered by cesarean section (c-section),54 although this difference was not found in a controlled trial of delivery route for very low birth weight (VLBW) infants.55 Breastfed infants delivered by c-section take in significantly fewer calories in the first 6 days than those delivered vaginally (Table 6-2).56 Thus, one would except TSB levels to be higher in infants delivered by c-section (see section “Breastfeeding and Jaundice”). The average cumulative intake in a vaginally delivered infant in the first 6 days of life is 450 ± 285 mL (SD) versus 358 ± 218 mL (SD) in c-section infants (P = .001).56 Infants delivered by c-section have lower hematocrits on day 6 than those delivered vaginally.54

Vacuum Extraction

Infants delivered by vacuum extraction26,57,58 are more likely to have a cephalhematoma and/or bruising of the scalp and, therefore, more likely to develop hyperbilirubinemia. Because the catabolism of 1 g of hemoglobin yields 35 mg of bilirubin, bruising and cephalhematomas can contribute significantly to the infant’s bilirubin load.11,59,60

Placental Transfusion and the Timing of Umbilical Cord Clamping

A Cochrane database review of 11 trials of 2989 mothers61 and their infants found that in those whose cords were clamped early, significantly fewer required phototherapy for jaundice (RR 0.59, 95% CI 0.38–0.92). The late clamped group also had a significant increase in hemoglobin levels. The blood volume of term and preterm newborns is increased when there is a delay in cord clamping.62 In a study of infants 28–36 weeks of gestation, Saigal et al. found that if cord clamping was delayed, the mean TSB level at 72 hours was 7.7 mg/dL compared with 3.2 mg/dL in the early clamped group.63 On the other hand, in a recent study of late preterm infants, although the late cord clamped group showed higher hemoglobin levels, there was no relationship between delayed clamping and pathologic jaundice or polycythemia and the need for phototherapy.64

Umbilical Cord Bilirubin Levels

The association between TSB levels in the cord blood and subsequent peak TSB concentrations was described more than 50 years ago65 and has since been confirmed in infants with and without hemolytic disease.66–68

Gestation

In term and late preterm newborns, by far the most important single clinical factor associated with the subsequent risk of hyperbilirubinemia is the infant’s gestational age.11,69–71 This association has been confirmed consistently in multiple studies, although the magnitude of this risk has only recently been quantified (Table 6-3).11,69–71 How decreasing gestation and other factors contribute to the risk of subsequent hyperbilirubinemia is discussed in detail in Chapter 9.

Birth Weight

Because birth weight is generally a direct reflection of gestational age, it has long been known that decreasing birth weight is associated with an increased risk of hyperbilirubinemia. Currently, it is impossible to determine the natural history of bilirubinemia in the VLBW (<1500 g) or extremely low birth weight (ELBW <1000 g) infant because a large proportion of these infants receive phototherapy. Small studies in the 1950s72,73 showed a direct correlation between decreasing birth weight and maximum TSB concentrations and also showed that achievement of the peak TSB concentration occurs later in preterm infants than in term infants. In those days, kernicterus in LBW (<2500 g) infants was not uncommon.74

Given the association described above between decreasing birth weight and TSB levels, it is a surprise to find that in term and late preterm infants, after controlling for gestational age, there is a positive association between increasing birth weight and significant hyperbilirubinemia.57,75,76 IDMs who are macrocosmic29 and have an increased birth weight/length ratio31 are more likely to develop hyperbilirubinemia than those who are AGA and some larger infants could be undiagnosed IDMs. Other potential explanations for the increase in TSB associated with larger birth weight are the need for vacuum extraction delivery57 and scalp injury during delivery11,69,76 and the fact that physicians may be less concerned about jaundice in larger babies and use higher TSB levels for initiating phototherapy.76 Flaherman et al.76 evaluated 111,009 infants born between 1995 and 1998. They used logistic regression to control for gestational age, scalp injury diagnosis, maternal diabetes, method of delivery, and other confounders, and found that higher birth weight was associated with TSB levels >20 mg/dL in 36–38 weeks gestation infants, but not in infants ≥39 weeks. Birth weight had a small but significant effect on hyperbilirubinemia only for less mature babies. The relationship between increasing birth weight and hyperbilirubinemia is intriguing and requires further investigation to elucidate its pathophysiology.76

Gender

As a group, male infants have higher bilirubin levels than females,11,26,71,77,78 although occasional studies have not found this relationship.70 In the reported cases of kernicterus, there is a striking preponderance of males.79,80

Caloric Intake and Weight Loss

There is a consistent association between hyperbilirubinemia and weight loss in the first few days after birth.26,70,71,77,78,81 Caloric deprivation is associated with increases in TSB in animals and humans82 and there is a reciprocal relationship between caloric intake and the degree of hyperbilirubinemia in Gilbert’s syndrome.83 In adults, caloric deprivation is also associated with increased hemolysis.84 In a randomized controlled trial, infants with birth weights between 1250 and 2000 g were assigned to be fed within 2 hours of birth or 24–36 hours after birth. The earlier fed infants had significantly lower TSB levels between 96 and 144 hours.85 The primary mechanism responsible for the lower TSB appears to be a decrease in the enterohepatic circulation of bilirubin.82,86

Type of Diet

Infants fed a casein hydrolysate formula had significantly lower TSB levels from days 10 to 18 than those in infants fed standard casein or whey-predominant formulas.87,88 The cumulative stool output of infants fed a casein hydrolysate formula was lower than that in infants fed the other formulas,87 suggesting that factors other than stool output and its effect on the enterohepatic circulation must explain these observations. The casein hydrolysate formula (Nutramigen) contains an inhibitor of β-glucuronidase,89 an enzyme that acts on the hydrolysis of bilirubin glucuronide and therefore facilitates the enterohepatic absorption of unconjugated bilirubin.35 When fed saccharolactone, an inhibitor of β-glucuronidase, rats excrete less bilirubin in their bile suggesting that inhibition of β-glucuronidase decreased intestinal absorption of bilirubin.90

Breastfeeding and Jaundice

The vast majority of studies in the last 30 years have found a strong association between breastfeeding, elevated TSB levels in the first few days, and an increased risk of subsequent significant hyperbilirubinemia (Figure 6-1).11,22,71,75,91–94 A pooled analysis of 12 studies of more than 8000 newborns showed that breastfed infants were about three times more likely to develop TSB levels of ≥12 mg/dL and six times more likely to develop levels of ≥15 mg/dL than formula-fed infants (Figure 6-1).93Table 6-4 lists some recent studies that have quantified the risk of hyperbilirubinemia in exclusively breastfed infants.