ORIGINAL ARTICLE: HEPATOLOGY AND NUTRITION
Risk Factors for Developing Transient
Neonatal Cholestasis


Valerie Champion, yRicardo Carbajal, zJana Lozar,
Isabelle Girard, and
Delphine Mitanchez
ABSTRACT
Objectives: To describe the incidence and the characteristics of neonatal
cholestasis in a cohort of patients with known risk factors and to investigate
additional risk factors.
Methods: A prospective observational study conducted between April 2008
and 2009 involved all neonates admitted in the neonatal ward. They were
divided into high- and low-risk groups for cholestasis. The high-risk group
included preterm birth <34 weeks of gestation, small for gestational age
(SGA), parenteral nutrition (PN) >7 days, abdomino-pelvic or thoracic
surgery. Bilirubinemia was weekly measured in the high-risk group.
Results: Of the 460 newborns studied, 234 were included in the high-risk
group and 226 in the low-risk group. Cholestasis developed in 32 patients
(13.7%) in the high-risk group at mean (SD) age of 14.7 (12.9) days; all were
receiving PN. None of the patients in the low-risk group developed
cholestasis. An analysis was carried out in the 207 patients in the high-
risk group who received PN. The odds ratio (OR) for developing cholestasis
was 2.3 [1.1–5.0] and 5.6 [2.5–12.5] for SGA or surgical patients,
respectively. Cholestasis was associated with neonatal severe conditions,
longer PN duration, and more intravenous macronutrients’ intakes. In
multivariate analysis, SGA and neonatal surgery were strong independent
risk factors for cholestasis, with OR (95% confidence interval [95% CI]) of
4.4 [1.6–12.5] and 4.6 [1.7–12.3], respectively.
Conclusions: Transient neonatal cholestasis is a complication of PN. SGA
and neonatal surgery are additional risk factors. There is no evidence to limit
intravenous protein intakes in preterm.
Key Words: fetal growth retardation, parenteral nutrition, postnatal
nutrition, postoperative complications, premature infant
(JPGN 2012;55: 592–598)
N eonatal cholestasis is a rare condition that affects approxi-
mately 1 of 2500 infants (1). During management of these
patients, the first priority is to recognize conditions requiring
immediate treatment because in those cases, the timing of diagnosis
is directly related to the outcome. Cholestasis is more commonly
reported in premature babies, small-for-gestational-age (SGA)
infants, and surgical patients, especially when prolonged parenteral
Received January 12, 2012; accepted May 6, 2012.
From the
Service de Néonatologie, Hôpital Armand Trousseau, APHP,
Université Pierre et Marie Curie, the yService de Urgences Pédiatriques,
Hôpital Armand Trousseau, APHP, Université Pierre et Marie Curie,
Paris, and the zUniversity Children’s Hospital, Ljubljana, Slovenia.
Address correspondence and reprint requests to Professor Delphine
Mitanchez, Service de Néonatologie, Hôpital Armand Trousseau, 26
Avenue du Docteur Arnold Netter, 75571 Paris Cedex 12, France
(e-mail: delphine.mitanchez@trs.aphp.fr).
The authors report no conflicts of interest.
Copyright # 2012 by European Society for Pediatric Gastroenterology,
Hepatology, and Nutrition and North American Society for Pediatric
Gastroenterology, Hepatology, and Nutrition
DOI: 10.1097/MPG.0b013e3182616916
592 JPGN  Volume 55, Number 5, November 2012
Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
nutrition (PN) of >2 weeks is given. Its incidence varies between
10% and 20% in preterm and SGA depending on the criteria and the
duration of PN (2–4). An incidence as high as 60% has been
reported in surgical neonates (5).
Several studies searching for factors associated with the
development of cholestasis in neonates have been published
(2–4,6–8). Sepsis, delayed enteral intakes, PN duration, amino
acid intakes, and energy overload have been found to be associated
with cholestasis. Male sex, perinatal depression, and shock have
also been implicated as potential risk factors (9,10). Most of these
studies were retrospective chart reviews and some have yielded
conflicting results.
We conducted a prospective study in a cohort of patients with
known risk factors, using a standardized biological testing, to
describe the incidence and the characteristics of neonatal choles-
tasis and to investigate additional risk factors.
METHODS
From April 2008 to May 2009, we prospectively included all
of the neonates admitted during the first month of life to the
neonatal ward of the Armand Trousseau Hospital. Patients were
divided into 2 groups: high and low risk for cholestasis. Patients
with the following criteria were included in the high-risk group:
preterm birth <34 weeks of gestation (WG), small for gestational
age (SGA), PN >7 days, and abdomino-pelvic or thoracic surgery.
All of the other newborns were considered as the low-risk infants.
The study was approved by the ‘‘Comité d’Ethique Appliqué a la
Recherche’’ (Société Française de Néonatologie).
In the high-risk group, bilirubinemia was checked at least
once in the first week and then weekly as long as the infant received
PN. In the low-risk group, jaundice was clinically checked every
day and biological screening was indicated when the babies were
found to be icteric (11). Cholestasis was defined as conjugated
bilirubin >1 mg/dL (17 mmol/L) if the total bilirubin was under
5 mg/dL (85 mmol/L) or as >20% of total bilirubin in the other cases
(1). When cholestasis was detected, diagnostic procedures to
exclude biliary atresia or other hepatic malformations (abdominal
ultrasound), hypothyroidism (TSH, free T4), and cytomegalovirus
infection (detection of cytomegalovirus in urine samples) were
performed. Maternal serologic status for hepatitis B, rubella, tox-
oplasmosis, syphilis, and HIV was systematically checked during
pregnancy. Infants with hepatic malformations or positive viral tests
were excluded from the analysis. The assessment of basic liver
function included serum alanine aminotransferase (ALT), serum
aspartate aminotransferase (AST), and prothrombin time, which
were performed every 2 weeks. Cystic fibrosis was excluded for
every patient by systematic neonatal screening. Infants with central
venous line and PN were checked for secondary sepsis every week,
allowing diagnosis of, among other infections, urinary tract infec-
tion.
The primary outcome was the incidence of cholestasis in
the high-risk group. SGA was defined as a birth weight <10th
 JPGN  Volume 55, Number 5, November 2012
percentile according to neonatal standard curves (12). Surgical
neonates were those requiring surgery within the first week of life
for thoracic or abdomino-pelvic disorder. Maternal and neonatal
data were recorded. Clinical chorioamnionitis was defined by the
presence of temperature >37.88C in a gravid patient without
another localized infection, and any two of these other criteria:
uterine tenderness or foul-smelling amniotic fluid, maternal tachy-
cardia >120 beats/min, and fetal tachycardia >160 beats/min (13).
Respiratory distress was defined by a respiratory support of
>2 hours after birth (nasal oxygen, noninvasive or mechanical
ventilation). Bronchopulmonary dysplasia (BPD) was defined by
oxygen requirement at 36 weeks postmenstrual age (PMA). Mater-
nofetal infection was defined by C-reactive protein >20 mg/L and
either positive superficial bacterial samples or positive blood
or cerebral spinal fluid culture. Late-onset sepsis was defined by
C-reactive protein >10 mg/L and either culture proven or culture
negative after 72 hours of life but treated with antibiotics for at least
7 days. Hemodynamic failure was considered when blood volume
expansions, vasopressor, or inotropic drugs were required. Hypo-
glycemia was defined by 2 capillary glycemias <2.4 mmol/L.
Necrotizing enterocolitis (NEC) was diagnosed according to Bell’s
staging criteria (14). Infants were considered affected from grade
IIa. Perinatal asphyxia was defined by Apgar score at the first
minute <3 and/or Apgar score at the fifth minute <7. Anemia and
thrombopenia were defined by hemoglobin value <12 g/dL and a
platelets count <100,000 cells/mm3.
The amino acid solutions given during the study period were
Pediaven1 and Pediaven2 (Fresenius Kabi France SAS, Sèvres
Cédex, France) in the first 24–48 hours of life and then Vaminolact
(Fresenius Kabi France, Sèvres Cédex, France). Lipid solution was
Medialipide 20% (B. Braun Medical, Billancourt Cedex, France).
Protein infusion was started within the first 24 hours of life to
achieve 1 g  kg1  day1 and advanced to a maximum of 3 to
4.5 g  kg1  day1, according to birth weight and gestational age.
Dextrose was initiated on the first day of life to achieve a glucose
infusion rate between 6 and 8 g  kg1  day1 and advanced to a
maximum of 20 g  kg1  day1 for premature and SGA babies and
22 g  kg1  day1 for surgical babies. All of the infants with PN
were checked at least once per day for capillary glycemia. Lipid
infusion was introduced between 24 and 48 hours of life to achieve
0.5 g  kg1  day1. Daily lipid increment was 0.5 g  kg1 day1 to a
maximum dose of 3 to 3.5 g  kg1  day1. When cholestasis was
diagnosed, parenteral lipid intakes were limited to 2 g  kg1  day1
3 days per week. Average intakes were calculated as the sum of
daily intakes of proteins, lipids, or glucose per kilogram divided by
the number of days of PN.
Infants who were <32 weeks PMAweighed <1500 g, or had a
postsurgical ileus received human milk as starting feeding. Enteral
nutrition was started within the first 24 to 48 hours or when daily
amounts of gastric residual were <30 mL/kg in surgical patients. It
was increased by 12–24 mL  kg1  day1 until reaching full feeds
(150–180 mL  kg1  day1). Time of first enteral feeding was
defined as the moment of first milk administration since birth and
a tolerance of continuous or discontinuous enteral infusion of milk for
>48 hours. Full enteral nutrition was considered as attained when
percutaneous catheter was withdrawn. We recorded days of feeding
interruptions (>12 hours) and day when enteral feedings provided at
least 100 calories  kg1  day1. The absence of enteral feed (no
enteral feed) was defined as an interruption of at least 1 day.
Statistics
All of the potential risk factors for the development of
cholestasis were assessed individually using x2 or Fisher exact
www.jpgn.org
Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Risk Factors for Developing Transient Neonatal Cholestasis
tests for qualitative data and Student t test for continuous variables.
Results are expressed as numbers (%) or mean  SD. P values
<0.05 were considered statistically significant. We used multi-
variate logistic regression analysis to calculate the adjusted odds
ratio (OR) for the development of cholestasis. In the logistic
regression model, we incorporated 4 of the variables found by
the univariate analysis to be P < 0.1. Statistical analysis was
performed using SPSS version 14 (SPSS Inc, Chicago, IL).
RESULTS
During the study period, 468 infants were admitted to
the neonatal unit within their first 6 weeks of life. Among the
468 patients, 242 had at least one of the risk factors (Fig. 1). Eight
of these infants were excluded from the analysis because of the
absence of bilirubin monitoring during >3 weeks. Of these
infants, 4 were extremely preterm between 25 and 26 WG with
appropriate weight for gestational age, 1 was a 31 WG preterm
with small bowel atresia, 1 had a Cantrell pentalogy, 1 had a
polymalformation syndrome, and the last 1 had infantile nephro-
nophtisis. Cholestasis was diagnosed only in the latter infant.
Thus, the high-risk group was constituted by 234 infants. The
main characteristics of the present group are presented in Table 1.
The diagnosis of the 56 surgical disorders were 9 (16%) small bowel
atresia, 8 (14.3%) esophagus atresia, 6 (10.7%) omphalocele,
4 (7.1%) imperforate anus, 4 (7.1%) gastroschisis, 4 (7.1%) meconial
ileus, 2 (3.6%) congenital diaphragmatic hernia, and 19 (33.9 %)
other disorders.
Two hundred twenty-six infants were in the low-risk group
(Table 1). GA was between 34 and 36 weeks for 39.4% of the
babies. Main morbidities were respiratory distress in 91 infants
(40.2%), jaundice in 100 infants (44.2%), maternofetal infection
in 20 infants (8.8%), and perinatal asphyxia in 31 infants
(13.7%).
Cholestasis was diagnosed only in the high-risk group and
developed in 32 out of 234 (13.7%) patients of the group at a mean
(SD; range) age of 14.7 days (12.9; 3–56 days). One SGA baby in
the high-risk group with cholestasis was excluded from analysis
because hepatic arteriovenous anomalies were diagnosed by
abdominal ultrasound. In 53.1% of patients, cholestasis persisted
>2 weeks and remained at discharge in 37.5% of the patients. At
4 weeks of age, cholestasis was still present in 13 of 31 (41.9%) of
patients. Among them, 8 of 13 patients still required PN. At 8 weeks
of age, it only remained in 4 patients (12.9%) and was followed by
full recovery within the first year. There was no case of liver failure.
Aminotransferases blood levels were within normal ranges at
4 weeks, respectively, ALT 18.7 IU/L (range 6–40) and AST
33.2 IU/L (range 13–61). No patient had a low prothrombin time
within the first 2 months.
Among the 459 infants analyzed, we confirmed that SGA,
gestational age under 34 WG, and surgery were significant risk
factors for cholestasis (Table 2). None of the 252 infants who did
not receive PN or who received PN <7 days developed cholestasis.
Of the 234 neonates in the high-risk group, 207 (88.5%)
required PN. Because none of the 27 infants in the group who did
not require PN developed cholestasis, we only considered the 207
babies who received PN for the analysis of additional risk factors.
The main characteristics of the subgroup are presented in Table 3.
Among the 207 neonates with PN, mean gestational age and weight
were similar in infants with and without cholestasis (34.0 vs
33.5 weeks, P ¼ 0.45 and 1763.3 vs 1901.8 g, P ¼ 0.34, respect-
ively).
Univariate analyses for the association between cholestasis
and various neonatal conditions in the 207 high-risk infants who
received PN are reported in Table 4. Preterm birth <32 WG was not
593
 Champion et al JPGN  Volume 55, Number 5, November 2012

Neonates hospitalized in Trousseau during the 1st month n = 468

Infants with one of these risk factors n = 242: preterm birth

< 34 WG, SGA, parenteral nutrition > 7 days, abdomino-
pelvic or thoracic surgery

Excluded infants n = 8 (absence of bilirubin

monotoring): extreme prematurity (4),
polymalformative syndrom (2), nephronophtisis
(1), prematurity with small bowell atresia (1)

High risk group n = 234 Low risk group n = 226

Parenteral nutrition n = 207 No parenteral nutrition n = 27

Cholestasis n = 32 No cholestasis n = 175 Cholestasis with hepatic arterio No cholestasis n = 26 No cholestasis n = 226
venous anomalies n = 1

Cholestasis < 14 days n = 15 Cholestasis ≥ 14 days n = 17

Cholestasis remaining after hospitalization n = 12 Cholestasis disappearing after hospitalization n = 5

Cholestasis remaining after the first year n = 0

FIGURE 1. Incidence and evolution of cholestasis in studied population.

an additional factor associated with cholestasis in the subgroup. No
associated risk factors for cholestasis were identified among
maternal morbidities (diabetes, preeclampsia, chorioamnionitis,
and prolonged rupture of membranes).
TABLE 1. Main characteristics of the high- and low-risk groups
included in the study
The association between cholestasis and potential nutritional
risk factors in the high-risk group of infants who received parental
nutrition is presented in Table 5.
In multivariate analysis, after adjustment on PN duration,
SGA and neonatal surgery were strong independent risk factors for
cholestasis (Table 6).

High-risk group,z Low-risk group,§ DISCUSSION

n ¼ 234 n ¼ 226 To the best of our knowledge, the present study is the first
prospective study analyzing risk factors for developing neonatal


Gestational age, wk 33.9 (3.7) 37.8 (2.4) cholestasis in a large group of neonates. We prospectively checked


Birth weight, g ,y 1906.5 (718.4) 2991.1 (614.6) for cholestasis in all of the neonates admitted in the unit. We
Male sex 121 (51.7) 115 (50.9 distinguished a low-risk group that was clinically checked daily and
SGA 81 (34.6) 0 (0) biologically if necessary, and a high-risk group with standardized
Prematurity <32 wk 70 (29.9) 0 (0) follow-up. The latter group included neonates who were preterm
Prematurity <34 wk 130 (55.6) 0 (0) <34 WG, SGA, received PN >7 days, or required surgery. None of
Prematurity 34–36 wk 47 (20.0) 89 (39.4) the patients in the low-risk group developed cholestasis. Only
Surgery 56 (23.9) 0 (0) patients who received PN developed cholestasis and SGA and
Parenteral nutrition 207 (88.5) 92 (40.7) neonatal surgery were found to be the main associated risk factors


Hospital stay duration, d 36.4 (24.5) 9.0 (7.3) for developing cholestasis. Similarly to other investigators, we
found that infants who developed cholestasis had more severe
Values are n (%) unless indicated otherwise. PN ¼ parenteral nutrition; neonatal conditions and more feeding interruptions. They were
SGA ¼ small for gestational age. exposed to PN for a longer duration and received more intravenous


Values are mean (SD).
y
Birth weight for one infant missed. macronutrients.
z
Infants presenting at least one of the following: preterm birth <34 WG, In case of cholestatic jaundice, some clinical conditions are
SGA (birth weight <10th percentile), PN > 7 days, abdomino-pelvic or related to serious etiologies and need prompt diagnosis and therapy.
thoracic surgery. Early diagnosis in these cases is potentially life-saving and may lead
§
Infants who did not present any of the high-risk criteria. to better outcomes because of better support that may help avoid

594 www.jpgn.org

Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
 JPGN  Volume 55, Number 5, November 2012 Risk Factors for Developing Transient Neonatal Cholestasis

TABLE 2. Demographic and comorbid factors associated with cholestasis in the whole studied population (n U 459), including high- and low-risk


infants

Cholestasis

N n % Crude odds ratio (95% CI) P

SGA
No 379 18 4.8 1 (reference) <0.001
Yes 80 14 17.5 4.2 (2.0–9.0)
Prematurity <34 wk
No 329 17 5.2 1 (reference) 0.02
Yes 130 15 11.5 2.4 (1.2–5.0)
Surgery
No 403 13 3.2 1 (reference) <0.001
Yes 56 19 33.9 15.4 (7.0–33.7)
Parenteral nutrition >7 d
No 317 0 0 NA NA
Yes 142 32 22.5

CI ¼ confidence interval; NA ¼ not applicable; SGA ¼ small for gestational age.

One infant with hepatic arteriovenous anomalies excluded from analysis.

complications of liver disease. The most serious one is biliary findings. The largest study was reported by Robinson and Ehrenk-
atresia for which the outcome is related to the timing of surgical ranz (8), who conducted a retrospective case–control trial including
intervention, with better outcomes found before 45–60 days. We >200 patients, <34 WG, and receiving PN 7 days or more. They
checked all of the patients for differential diagnosis, according to found that 58% of SGA infants developed cholestasis, and the OR
the current guidelines (1). Abdominal ultrasound was normal in all for SGA was close to our finding (3.3 [1.6–6.6]). In the cholestatic
of the patients, although the present finding cannot be solely relied group, infants with SGA had lower birth weight, but they were more
on to exclude biliary atresia. We did not indicate more invasive mature than the AGA infants (8). In a large retrospective study
procedure because pale stools were observed only transitorily in conducted over a 2-year period, Wright et al (15) found that infants
two infants. Alpha-1-antitrypsin deficiency was eliminated in case with the highest risk for cholestasis were preterms with lower birth
of persistent cholestasis without improvement for >3 weeks. We weight, NEC, sepsis, or who required longer ventilation, more
did not check the patients for galactosemia and tyrosinemia because hemodynamic support, and longer PN. After logistic regression
none of them was acutely ill. Furthermore, urinary tract infection was analysis, only birth weight, duration of PN, and the type of amino
easily excluded because all of the infants with central venous line and acids remained as significant risk factors. Baserga and Sola (6)
PN were checked weekly for secondary sepsis. Finally, cholestasis suggested that intrauterine growth restriction (IUGR) is the main
either improved or stabilized in all of the cases when intravenous condition that predisposes preterm to development of cholestasis. In
lipids were decreased and enteral nutrition was started. Thus, patients a retrospective review of extremely low birth weight infants
should be investigated when neonatal cholestasis appears during <1000 g receiving PN >7 days, they found that the incidence of
prolonged PN, but the investigations can be targeted according to cholestasis in SGA infants of this group was 56 vs 27% in AGA (OR
the clinical status and the evolution. If pale stools persist, invasive 3.3 [1.3–8.5]). Some authors have shown that IUGR causes hepa-
methods should be used to eliminate biliary atresia (11). tocellular dysfunction and leads to a decreased capacity to metab-
Previous studies investigated the relation between cholesta- olize proteins during the first postnatal weeks of life and to an
sis, prematurity, and SGA and their results are consistent with our altered hepatic fatty acid metabolism (16,17). All of these elements,
consistent with our results, support that birth weight has a higher
effect on the risk of cholestasis than prematurity.
TABLE 3. Main characteristics of the 207 infants who received parental Surgery was previously identified as a risk factor for cho-
nutrition in the high-risk group lestasis in small retrospective studies (18,19). In a large retro-
spective study reviewing 1366 patients who received PN at least
n % 14 days, the highest likelihood of developing cholestasis was found

in patients who had jejunal atresia (64%) and gastroschisis (43%)
Gestational age, wk 33.5 (3.7)

(5).
Birth weight, g ,y 1880.3 (745.2)
In many of the studies, the most consistently reported risk
Male sex 108 52.2
factor for neonatal cholestasis is the duration of PN. We also found
SGA 58 28.0
that PN was related to cholestasis that led us to adjust multivariate
Prematurity <32 wk 70 33.8
analysis on PN duration. PN duration is linked to bowel disuse that
Prematurity <34 wk 128 61.8
may lead to a lower stimulation of the enterohepatic function, a
Prematurity 34–36 wk 36 17.4
decreased gastrointestinal hormone release such as gastrin and
Surgery 56 27.0
cholecystokinin, and a decrease in bile transport (20). We found
Hospital stay duration, d 39.6 24.2
that the total amount of macronutrients intakes was significantly
Values are n and % unless indicated otherwise. SGA ¼ small for gesta- higher in the cholestatic group; however, in multivariate analysis,
tional age. daily protein intakes were not associated with cholestasis. We can

hypothesize that the association of daily amino acid intakes with
Values are mean (SD).
y
Birth weight for 1 infant was missing. cholestasis found in univariate analysis was related to PN duration.

www.jpgn.org 595

Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
 Champion et al JPGN  Volume 55, Number 5, November 2012

TABLE 4. Demographic and comorbid factors associated cholestasis in the high-risk group with parenteral nutrition (207)

Cholestasis

N n % Crude odds ratio (95% CI) P

Sex
Female 99 12 12.1 1 (reference) 0.20
Male 108 20 18.5 1.7 (0.8–3.6)
SGA
No 149 18 12.1 1 (reference) 0.03
Yes 58 14 24.1 2.3 (1.1–5.0)
Prematurity <37 wk
No 43 9 20.9 1 (reference) 0.27
Yes 164 23 14.0 0.6 (0.3–1.5)
Prematurity <32 wk
No 70 9 12.9 1 (reference) 0.46
Yes 137 23 16.8 0.7 (0.3–1.7)
Surgery
No 152 13 8.6 1 (reference) <0.001
Yes 55 19 34.6 5.6 (2.5–12.5)
Asphyxia
No 181 22 12.2 1 (reference) <0.001
Yes 24 10 41.7 5.2 (2.0–13.0)
Respiratory distress syndrome
No 77 8 10.4 1 (reference) 0.12
Yes 130 24 18.5 1.9 (0.8–4.6)
Intubation
No 106 9 8.5 1 (reference) 0.006
Yes 101 23 22.8 3.2 (1.4–7.3)
Maternofetal infection
No 187 30 16.0 1 (reference) 0.48
Yes 20 2 10.0 0.6 (0.1–2.6)
Hemodynamics disturbance
No 177 20 11.3 1 (reference) <0.001
Yes 30 12 40.0 5.2 (2.2–12.4)
Blood transfusion
No 152 11 7.2 1 (reference) <0.001
Yes 55 21 38.2 7.9 (3.5–18.0)
Thrombopenia
No 163 18 11.0 1 (reference) <0.001
Yes 44 14 31.8 3.8 (1.7–8.4)
Bronchopulmonary dysplasia
No 188 23 12.2 1 (reference) <0.001
Yes 19 9 47.4 6.5 (2.4–17.6)
Necrotizing enterocolitis
No 202 29 14.4 1 (reference) 0.01
Yes 5 3 60.0 8.9 (1.4–55.9)


Nosocomial sepsis
No 174 18 10.3 1 (reference) <0.001
Yes 33 14 42.4 6.4 (2.7–14.9)

CI ¼ confidence interval ; SGA ¼ small for gestational age.

All of the cases of sepsis were related to catheter coagulase-negative staphylococci infections.

The Pediatrix Amino Acid Study Group found that cholestasis in
preterm infants was related to higher cumulative doses of amino
acids (2); however, in a randomized controlled trial, they showed
that early higher doses of amino acids were not associated with an
increase in the risk of cholestasis (21). Therefore, we think that
protein intakes should not be limited in the first days of life for the
smallest babies.
The deleterious effects of lipids on the hepatobiliary system
have been suggested in children by Colomb et al (22). Colomb et al
suggested that the clearance of endotoxins may be reduced by
accumulation of lipids in Kuppfer cells and that peroxidation of
lipids produces toxic metabolites. Similar to others, we found an
association between cholestasis and daily lipid intakes (23). These
intakes were decreased to 2 g/kg 3 times per week when cholestasis
was diagnosed, according to the recent European guidelines (24).
On the basis of the factors associated with cholestasis,
Wagner et al (25) hypothesize about the potential role of oxidative
stress as a cause of cholestasis. Molecular mechanisms of biliary
secretion involve different transporters, which can be inhibited by
mediators of the oxidative stress (26). Newborns are especially

596 www.jpgn.org

Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
 JPGN  Volume 55, Number 5, November 2012
TABLE 5. Characteristics and potential nutritional risk factors for the development of cholestasis in the 207 infants of the high-risk group who
received PN (mean W SD)
Cholestasis, n ¼ 32
Gestational age, wk
Birth weight, g
Intravenous protein, g  kg1  d1
Intravenous glucose, g  kg1  d1
Intravenous lipids, g  kg1  d1
PN duration, d
First enteral feeds, d
Feeding interruption duration, d
Delay before enteral caloric intakes, 100 calories  kg1  d1
PN ¼ parenteral nutrition.
prone to oxidative stress because they have reduced antioxidant
defenses and are often exposed to situations such as high oxygen
concentration administration, infection, or inflammation that
enhance oxidative stress (27). Similar to most of previous studies
dealing with the present subject did (2–4,8,15), ours also found that
severe neonatal conditions such as asphyxia, NEC, secondary
sepsis, hemodynamic failure, and BPD were associated with cho-
lestasis. These comorbid factors are covariate markers for the
degree of severity of illness, but they are also related to oxidative
stress, as a cause or a consequence (28). Furthermore, there is much
evidence that fetal growth restriction is a condition that exposes to
oxidative damage whether it is related to preeclampsia (28) or to an
unexplained cause (29). Oxidative stress is also believed to be an
integrated part of the surgical stress response (30).
PN adds to these situations that enhance oxidative stress. The
solutions are contaminated with various sources of peroxides
originating from the lipid emulsions, vitamins, and interactions
between nutrients and ambient light (31). Protecting parenteral bags
from light limits formation of peroxides and could have potential
positive clinical effects (32). On the contrary, the use of certain
ingredients in PN formulas may enhance antioxidant capacity of the
neonates. Lipid solutions containing fish oil have anti-inflammatory
effects due to their abundance of omega-3 fatty acids and the high
amount of antioxidant a-tocopherol. Data have emerged to show
that fish oil–based lipid emulsion may be both safe and efficacious
to reverse PN-induced liver disease in neonates and infants (33).
Preliminary data are encouraging, but such a product deserves
further evaluation using a larger number of samples. Although
the above oxidative stress hypothesis may be tempting, it remains so
far a hypothesis. Biochemical indicators of oxidative stress have not
been studied to support these conclusions.
The strength of the present study is its prospective design;
however, despite including >200 patients at risk, the major limita-
tion of the present study was the small number of cases of
cholestasis that precluded detailed multivariate analysis using
multiple parameters. Furthermore, the weekly bilirubin checking
TABLE 6. Multivariate analysis of the risk factors for cholestasis in the
207 infants of the high-risk group who received parenteral nutrition
Adjusted OR P caloric intake on parenteral nutrition associated cholestasis in premature
Small for gestational age, % 4.4 (1.6–12.5) 0.005
PN duration, d 1.1 (1.0–1.1) <0.001
Intravenous protein intake, g  kg1  d1 2.0 (0.8–4.5) 0.12
Surgery, % 4.6 (1.7–12.3) 0.003
OR ¼ odds ratio; PN ¼ parenteral nutrition.
www.jpgn.org
Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
Risk Factors for Developing Transient Neonatal Cholestasis
No cholestasis, n ¼ 175 P
34.0  4.3 33.5  3.6 0.45
1763.3  830.7 1901.8  729.0 0.34
2.52  0.54 1.77  0.79 <0.001
13.96  3.27 9.96  3.20 <0.001
1.87  0.56 1.24  0.97 <0.001
35.84  18.81 13.31  12.96 <0.001
6.44  10.86 1.91  3.32 0.03
11.94  11.26 2.81  5.70 <0.001
37.56  22.37 13.77  11.14 <0.001
and the subsequent diminution of parenteral lipid intakes precluded
the analysis of risk factors related to lipid intake. Finally, another
limitation of the study is a relative small group of patients to include
variable causes of infantile cholestasis including rare diseases. Our
conclusions are mainly related to transient cholestatic disease and
cannot be generalized.
In conclusion, we confirmed that transient neonatal choles-
tasis is a complication of PN, and that SGA and neonatal surgery are
additional risk factors. Although neonatal cholestasis is otherwise a
rare occurrence, other causes should be systematically excluded by
clinical examination and noninvasive investigations to avoid the
risk of delaying the diagnosis of a serious etiology. PN is a risk
factor that cannot be easily avoided. Our data do not question
aggressive PN with high protein intakes that is recommended in
current guidelines (34). Parenteral fish oils may also offer an
interesting preventive measure in the future. Our data suggest that
cholestasis could be a consequence of oxidative stress. To confirm
the present hypothesis, further studies are needed to investigate
markers of oxidative stress in such conditions.
REFERENCES
1. Moyer V, Freese DK, Whitington PF, et al. Guideline for the evaluation
of cholestatic jaundice in infants: recommendations of the North
American Society for Pediatric Gastroenterology, Hepatology and
Nutrition. J Pediatr Gastroenterol Nutr 2004;39:115–28.
2. Steinbach M, Clark RH, Kelleher AS, et al. Demographic and nutritional
factors associated with prolonged cholestatic jaundice in the premature
infant. J Perinatol 2008;28:129–35.
3. Hsieh MH, Pai W, Tseng HI, et al. Parenteral nutrition-associated
cholestasis in premature babies: risk factors and predictors. Pediatr
Neonatol 2009;50:202–7.
4. Costa S, Maggio L, Sindico P, et al. Preterm small for gestational age
infants are not at higher risk for parenteral nutrition-associated choles-
tasis. J Pediatr 2010;156:575–9.
5. Christensen RD, Henry E, Wiedmeier SE, et al. Identifying patients, on
the first day of life, at high-risk of developing parenteral nutrition-
associated liver disease. J Perinatol 2007;27:284–90.
6. Baserga MC, Sola A. Intrauterine growth restriction impacts tolerance
to total parenteral nutrition in extremely low birth weight infants.
J Perinatol 2004;24:476–81.
7. Blau J, Sridhar S, Mathieson S, et al. Effects of protein/nonprotein
infants weighing 600–1000 grams. JPEN J Parenter Enteral Nutr
2007;31:487–90.
8. Robinson DT, Ehrenkranz RA. Parenteral nutrition-associated choles-
tasis in small for gestational age infants. J Pediatr 2008;152:59–62.
9. Albers MJ, de Gast-Bakker DA, van Dam NA, et al. Male sex predis-
poses the newborn surgical patient to parenteral nutrition-associated
cholestasis and to sepsis. Arch Surg 2002;137:789–93.
597
 Champion et al
10. Jacquemin E, Maurage C, Borderon JC, et al. Early cholestasis in
premature infants receiving total parenteral nutrition: a possible con-
sequence of shock and hypoxia. Eur J Pediatr Surg 1995;5:259–61.
11. Benchimol EI, Walsh CM, Ling SC. Early diagnosis of neonatal chole-
static jaundice: test at 2 weeks. Can Fam Physician 2009;55:1184–92.
12. Leroy B, Lefort F. The weight and size of newborn infants at birth. Rev
Fr Gynecol Obstet 1971;66:391–6.
13. Popowski T, Goffinet F, Maillard F, et al. Maternal markers for detecting
early-onset neonatal infection and chorioamnionitis in cases of pre-
mature rupture of membranes at or after 34 weeks of gestation: a two-
center prospective study. BMC Pregnancy Childbirth 2011;11:26.
14. Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enter-
ocolitis. Therapeutic decisions based upon clinical staging. Ann Surg
1978;187:1–7.
15. Wright K, Ernst KD, Gaylord MS, et al. Increased incidence of
parenteral nutrition-associated cholestasis with aminosyn PF compared
to trophamine. J Perinatol 2003;23:444–50.
16. Boehm G, Senger H, Braun W, et al. Metabolic differences between
AGA- and SGA-infants of very low birthweight. I. Relationship to
intrauterine growth retardation. Acta Paediatr Scand 1988;77:19–23.
17. Boehm G, Senger H, Muller D, et al. Metabolic differences between
AGA- and SGA-infants of very low birthweight. II. Relationship to
protein intake. Acta Paediatr Scand 1988;77:642–6.
18. Aspirot A, Su W, Flageole H, et al. Cholestasis associated with small
bowel atresia: do we always need to investigate? J Pediatr Surg 2007;
42:873–7.
19. Beath SV, Davies P, Papadopoulou A, et al. Parenteral nutrition-related
cholestasis in postsurgical neonates: multivariate analysis of risk fac-
tors. J Pediatr Surg 1996;31:604–6.
20. Touloukian RJ, Seashore JH. Hepatic secretory obstruction with total
parenteral nutrition in the infant. J Pediatr Surg 1975;10:353–60.
21. Clark RH, Chace DH, Spitzer AR. Effects of two different doses of
amino acid supplementation on growth and blood amino acid levels in
premature neonates admitted to the neonatal intensive care unit: a
randomized, controlled trial. Pediatrics 2007;120:1286–96.
22. Colomb V, Jobert-Giraud A, Lacaille F, et al. Role of lipid emulsions in
cholestasis associated with long-term parenteral nutrition in children.
JPEN J Parenter Enteral Nutr 2000;24:345–50.
598
Copyright 2012 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.
JPGN  Volume 55, Number 5, November 2012
23. Shin JI, Namgung R, Park MS, et al. Could lipid infusion be a risk for
parenteral nutrition-associated cholestasis in low birth weight neonates?
Eur J Pediatr 2008;167:197–202.
24. Koletzko B, Goulet O, Hunt J, et al. Guidelines on Paediatric Parenteral
Nutrition of the European Society of Paediatric Gastroenterology,
Hepatology and Nutrition (ESPGHAN) and the European Society for
Clinical Nutrition and Metabolism (ESPEN), Supported by the Euro-
pean Society of Paediatric Research (ESPR). J Pediatr Gastroenterol
Nutr 2005;41(Suppl 2):S1–87.
25. Wagner M, Zollner G, Trauner M. New molecular insights into the
mechanisms of cholestasis. J Hepatol 2009;51:565–80.
26. Sekine S, Yano K, Saeki J, et al. Oxidative stress is a triggering factor for
LPS-induced Mrp2 internalization in the cryopreserved rat and human
liver slices. Biochem Biophys Res Commun 2010;399:279–85.
27. Saugstad OD. Oxidative stress in the newborn: a 30-year perspective.
Biol Neonate 2005;88:228–36.
28. Gitto E, Pellegrino S, Gitto P, et al. Oxidative stress of the newborn
in the pre- and postnatal period and the clinical utility of melatonin.
J Pineal Res 2009;46:128–39.
29. Burton GJ, Jauniaux E, Charnock-Jones DS. The influence of the
intrauterine environment on human placental development. Int J Dev
Biol 2010; 54:303–12.
30. Kucukakin B, Gogenur I, Reiter RJ, et al. Oxidative stress in relation to
surgery: is there a role for the antioxidant melatonin? J Surg Res
2009;152:338–47.
31. Chessex P, Watson C, Kaczala GW, et al. Determinants of oxidant stress
in extremely low birth weight premature infants. Free Radic Biol Med
2010;49:1380–6.
32. Khashu M, Harrison A, Lalari V, et al. Impact of shielding parenteral
nutrition from light on routine monitoring of blood glucose and trigly-
ceride levels in preterm neonates. Arch Dis Child Fetal Neonatal Ed
2009;94:F111–5.
33. de Meijer VE, Gura KM, Meisel JA, et al. Parenteral fish oil as
monotherapy for patients with parenteral nutrition-associated liver
disease. Pediatr Surg Int 2009;25:123–4.
34. Simmer K. Aggressive nutrition for preterm infants: benefits and risks.
Early Hum Dev 2007;83:631–4.
www.jpgn.org