Effects of Protein/Nonprotein Caloric Intake on Parenteral Nutrition–Associated Cholestasis in Premature Infants Weighing 600–1000 Grams  
From the * Department of Pediatrics, NYU Medical
Center, New York, New York; and the Divisions of Correspondence: Anupama Chawla, MD, Department of Pediatrics, HSC T-11, Rm 080, SUNY at Stony Brook, Stony Brook, NY 11794-8111. Electronic mail may be sent to anchawla{at}notes.cc.sunysb.edu.
Background: Parenteral nutrition-associated cholestasis (PNAC) has
historically been a significant cause of morbidity and mortality in neonates
undergoing parenteral feeding. Studies examining the causes of cholestasis in
the PN-dependent neonate have produced a wide range of data, with some
conflicting results. Increased protein/nonprotein calorie ratios, increased
glucose concentrations, and increased lipid concentrations have all been
implicated as possible causes of PNAC. However, these studies were done in the
pre-TrophAmine (neonatal-specific amino acid parenteral nutrition [PN]
formulation) era. With the introduction of TrophAmine, infants are now
receiving higher concentrations of protein, often being advanced rapidly even
when nonprotein calories may not be sufficiently advanced to meet the infants'
caloric needs. To the best of our knowledge, no studies have been conducted to
evaluate the protein/nonprotein calorie ratio as a cause of PNAC in the
TrophAmine era. Methods: A retrospective chart review of 25
cholestatic and 25 noncholestatic PN-dependent premature neonates was
conducted. All neonates weighed between 600 and 1000 g. Cholestasis was
defined as a serum total bilirubin (TB) Parenteral nutrition-associated cholestasis (PNAC) has historically been a significant cause of morbidity and mortality in neonates undergoing parenteral feeding. Studies examining the causes of cholestasis in the parenteral nutrition (PN)-dependent neonate have produced a wide range of data, with some conflicting results. The literature suggests that overfeeding is associated with the development of PNAC.1,2 The relative percentages of macronutrients in various PN infusates have also been studied to determine factors increasing the risk of PNAC. For example, increased percentage of lipids in parenteral feeds has been associated with the development of PNAC.3 Decreasing the lipid supply in the PN solution has been shown to normalize bilirubin in 1 study of PNAC in the newborn.4 A 1980 study compared the effects of 2 PN solutions, differing only in protein content, on the development of cholestatic jaundice in infants. Neonates receiving amino acids at a high rate of protein intake (3.6 g/kg/d) developed PNAC earlier and had a higher peak bilirubin level than a comparable group of infants receiving a lower protein intake (2.3 g/kg/d). This study also examined carbohydrate content in PN solutions and found that infants receiving greater amounts of dextrose developed PNAC at an earlier age than infants receiving less carbohydrate.5 Other studies have shown no correlation between the incidence of jaundice and the amount of amino acids administered.6 Yet another study found no significant differences between cholestasic and noncholestatic infants with respect to the relative percentages of protein, glucose, and lipid calories in the PN infusates.6 Despite these differing findings, many have considered that the use of adult PN formulations with inappropriate macronutrient composition for children was a possible cause of PNAC. This has led to the development of pediatric PN formulations, with possibly more appropriate amino acid compositions. Two popular infant-specific PN formulations created to mimic the amino acid profile of premature infants are Aminosyn PF (APF; Abbott, North Chicago, IL) and TrophAmine (TA; McGraw, Irvine, CA). The first clinical trial using TA reported a significant decrease in the incidence of PNAC compared with historical controls.7 Subsequent studies have, however, disagreed about the benefits of using one infant-specific formulation over another.8 A recent, much larger study found a >2-fold increase in the incidence of PNAC when the use of APF was compared with exclusive TA use.9 TA has thus become the PN formulation of choice to prevent PNAC in the neonate.10 Despite the advances in producing more infant-specific formulations and a resultant decline in the incidence of PNAC, this continues to remain a significant clinical problem, affecting as many as 12.8%–33.3% of PN-dependent premature infants. With the introduction of TA,7 PN formulations with relatively higher amino acid concentrations are currently being used. Infants are now receiving higher concentrations of protein, often being advanced rapidly even when nonprotein calories are not sufficiently advanced to meet the infants' caloric needs. The appropriate amount of nonprotein calories required to spare the use of protein to meet caloric needs may play a role in the pathogenesis of PNAC. Studies of infants receiving TA exclusively as the protein source are needed to examine whether the different amounts of protein concentrations and protein/nonprotein calorie ratios play a role in the development of PNAC.
Subjects Infants whose results are reported here were all patients of the NICU at the Children's Medical Center of Stony Brook University Hospital. All premature infants weighing 600–1000 g, with a gestational age  32
weeks, were included in the study. Infants with any metabolic liver disease,
major congenital anomalies, or conditions requiring major surgery were
excluded. A retrospective chart review of these patients was conducted after
approval by the institutional review board of Stony Brook University.
Study Design
Statistical Methods
Table I shows the subject demographics of neonates in both groups. There was no significant difference in mean bodyweight, mean gestational age, race, or the male:female ratio between both study groups. Table II contains comorbid factors, namely, bronchopulmonary dysplasia, intraventricular hemorrhage, retinopathy of prematurity, and number of septic episodes. These factors were compared between the 2 groups and were also without statistical differences.
Analysis of parenteral macronutrient intake in both groups is displayed in Table III. There were no significant differences between total caloric intake, total protein intake, percentage protein calories, percentage nonprotein calories, protein (g/kg):nonprotein (kcal) ratio, total lipid intake, percentage lipid calories, glucose infusion rate, percentage carbohydrate calories, and total nonprotein caloric intake. Factors independent of macronutrient intake are displayed in Figure 1. Our analysis showed that increased duration of PN therapy and length of hospital stay is significantly associated with PNAC. However, the number of days before the start of enteral feedings did not differ between cholestatic and noncholestatic infants. Table IV shows there was no difference in renal function as evaluated by blood urea nitrogen and creatinine.
With the persistence of PNAC in the neonatal intensive care unit (NICU), understanding of any causative factors is imperative. Early, aggressive PN is now recommended in the NICU to prevent many of the complications associated with prematurity. Despite these new recommendations, protein administration is not standardized either for starting dose or rate of increase. However, the trend is to start at higher concentrations and to rapidly advance protein intake to a goal of 3.5–4.0 g/kg/d.11–13 However, as mentioned earlier, studies in the pre-TA era showed that increased protein concentrations were associated with the development of PNAC. Studies of infants receiving TA exclusively as the protein source are needed to examine if the different amounts of protein concentrations and protein/nonprotein calorie ratios play a role in the development of PNAC. Our study of 50 premature neonates showed no differences in total protein intake or the protein/nonprotein calorie ratio between the cholestatic and noncholestatic groups. Therefore, we conclude that providing higher protein concentrations and rapidly advancing protein intake while not advancing higher nonprotein calories is not associated with the development of PNAC. Excess lipid and carbohydrate intake have also been associated with PNAC in the pre-TA era. Infants in our study who received TA as their sole protein source showed no differences in total lipid and carbohydrate intake or the percentages of calories from lipids and carbohydrates between the 2 study groups. We conclude that previous studies implicating excess lipid and carbohydrate intake are not a cause for concern in the TA era, and the recommendation for aggressive PN is not associated with PNAC. Like previous studies,14 we found that the cholestatic neonates had a significantly increased duration of PN therapy. Our cholestatic group also displayed a significantly increased length of hospital stay. Contrary to some previous studies,2 our results showed no significant differences in macronutrient factors, including the protein/nonprotein calorie ratio, between cholestatic and noncholestatic premature PN-dependent infants. There was a trend noted to the length of time these patients were not fed orally, although this did not reach statistical significance (Figure 1).
Protein to nonprotein calorie ratio was not an etiology in the development of cholestasis in PN neonates weighing 600–1000 g. Renal function as elicited does not seem to contribute to the cholestatic status of these infants. Therefore, providing adequate protein calories should not be limited in this patient population, as suggested by previous studies in the pre-TA era. However, PNAC is significantly associated with prolonged duration of PN therapy, during which other factors like gastrointestinal dysfunction in absence of enteral nutrition, various hepatotoxins in the PN infusate, and small intestinal flora might also play a role in the development of PNAC.8,14 Further studies involving larger numbers of infants receiving TA are necessary to determine whether nutrition differences have any effect on hepatocyte secretion and whether bile flow plays a role in the development of PNAC. Received for publication December 22, 2006. Accepted for publication March 26, 2007.
Journal of Parenteral and Enteral Nutrition, Vol. 31, No. 6,
487-490 (2007)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
2.0 mg/dL, with a serum direct
bilirubin (DB) 
Top
MATERIALS AND METHODS
RESULTS