Parenteral nutrition (PN) therapy is lifesaving in premature infants, critically ill patients, and patients with intestinal failure. However, prolonged use is associated with several metabolic complications including hepatic dysfunction. Prematurity and intestinal failure (1) are the most common indications for prolonged PN therapy in infants and children (2). The clinical spectrum of PN-associated hepatic liver disease (PNALD) in children ranges from cholestasis to hepatosplenomegaly, portal hypertension, and failure of intestinal adaptation (3). A peak conjugated bilirubin >2.5 mg/dL has been shown to predict poor outcome (4), and there is increased risk for developing irreversible liver injury if conjugated bilirubin remains persistently elevated >6 mg/dL (5). Approximately 7% to 8% of all pediatric recipients of PN develop biochemical signs of hepatic dysfunction (6–8). By 3 months of PN therapy, more than 40% to 60% of infants develop biochemical evidence of PNALD (9–11). Irreversible PNALD is a leading indication for isolated or combined liver intestinal transplantation (12). PNALD has a multifactor etiology; however, the increasing prevalence with duration of PN therapy suggests the mechanisms include chronic exposure to a toxin or a worsening conditionally essential factor or nutrient deficiency.
Choline is a quaternary amine and has several metabolic roles including source of labile methyl groups, precursor for the neurotransmitter acetylcholine, and structural component of cell membranes, bile, and lipoproteins (13). Adequate amounts of choline are present in breast milk (14), infant formulas (15), and regular diets (16). Choline is also endogenously synthesized from metabolites involved in the transsulfuration pathway (13,17). Choline becomes an essential nutrient when methionine and folate are not available in the diet (18) and when de novo synthesis is impaired (17). Choline deficiency results in profound liver damage characterized by steatosis and oxidative injury in laboratory animals (19,20). Healthy adults fed a choline-deficient diet developed decreased plasma concentrations of free choline and increased serum alanine aminotransferase activity consistent with hepatic dysfunction (21). Standard PN with intravenous lipid emulsions contains “phosphatidylcholine” but insignificant amounts of free choline (22,23). Children and adult patients on long-term PN with intravenous lipids and without significant enteral absorption had decreased plasma concentrations of free choline, and the degree of deficiency was correlated with severity of abnormalities in hepatic transaminases (22,24,25). Furthermore, adults on chronic PN who were supplemented with choline had resolution of PN-associated steatosis (24,26). There are limited data about choline status in infants on chronic PN therapy. The current evidence in support of an association between choline deficiency and hepatic injury, lack of significant amounts of free choline in PN, increased requirements, and higher likelihood of immature endogenous biosynthetic pathways suggests that infants on prolonged PN therapy may be at greater risk for choline deficiency, and that this may play a role in the onset and progression of PNALD. Therefore, the aim of our study was to examine the status of choline and related metabolites in infants on chronic PN therapy.
The subjects were infants on PN therapy and a comparison group consisting of breast-fed or formula-fed infants in the neonatal intensive care units (NICUs) at Children's Memorial Hospital and Prentice Women's Hospital, both located in Chicago, IL. Subjects were continuously enrolled. The inclusion criterion for infants on PN was duration of PN of at least 4 weeks providing approximately 50% or more of daily energy intake. Infants with known primary liver disease (eg, bilary atresia, metabolic liver disease, neonatal hepatitis), therapy with extracorporeal membrane oxygenation, renal failure, and history of congenital toxoplasmosis, cytomegalovirus, and herpesvirus or HIV infection were excluded from the study. The controls were infants ages 4 weeks or more who were exclusively enterally fed with either breast or infant formula with no history of liver disease or PN therapy in the preceding 2 weeks. The following information was obtained at enrollment: gestation age (weeks), birth weight (g), weight at enrollment, diagnoses, duration of PN, percentage of total energy intake from PN and lipid emulsions, and enteral feedings. Informed consent was obtained before the study from the parent(s) and/or guardian(s). Institutional review boards at Children's Memorial Hospital and Northwestern University (Chicago, IL) approved the study.
Analysis of Choline and Metabolites
Our initial intent was to measure plasma-free choline, phosphatidylcholine, glycerolphosphocholine, and phosphocholine concentrations; however, because of an omission in the protocol, the specimens were processed and stored as whole blood. The whole-blood samples were stored at −70°C before analysis. All of the specimens were analyzed simultaneously to avoid any potential variation. Thus, the analyses were performed on whole blood and not plasma specimens. Choline, phosphocholine, glycerolphosphocholine, and phosphatidylcholine were extracted using the method of Bligh and Dyer (27). Aqueous and organic compounds were separated, analyzed, and quantified directly using liquid chromatography/electrospray ionization-isotope dilution mass spectrometry (LC/ESI-IDMS) after the addition of internal standards labeled with stable isotopes to correct for recovery (28). The concentrations of choline and related metabolites in whole blood were reported in nanomoles per milliliter.
The data were expressed as mean ± standard deviation. Comparisons between groups were made using 2-sample t tests or analysis of variance (ANOVA) with subsequent pairwise comparisons using a Bonferroni adjustment, as appropriate. Nonparametric tests were also performed and conclusions were the same. Correlation analysis was performed to examine the relation between PCho and GPCho. A trend test (29) was used to further explore the association between PN and PCho concentrations. Statistical significance was set at P < 0.05. Analyses were performed using Stata, version 10 (StataCorp, College Station, TX).
Twenty-eight infants were enrolled in the study. The controls were healthy low-birth-weight infants retained in the nursery primarily for clinical stability and weight gain before discharge home. The infants on PN and controls were similar in postnatal age at the time of enrollment in the study. The rest of the subject characteristics are displayed in Tables 1 and 2. The infants on PN (n = 14) were slightly older in gestation age and had significantly higher birth weight. The mean percentage of daily energy intake from PN was 78% and ranged from 49% to 100% (Table 1). The data about intravenous lipid intake (g kg−1 day−1) represented the average intake during the 7 days before enrollment. There was a trend for higher total energy intake among infants receiving <90% compared with those receiving >90% of daily energy intake from PN (116 ± 23 vs 100 ± 9 kcal/day, P = 0.057) that was attributed to enteral feeding. However, direct comparison of total energy intake necessitated an estimation of energy expended in thermic effect of food (30,31) and losses from malabsorption. Eleven of the control infants were fed fortified breast milk and 3 exclusively fed infant formula (Enfamil EnfaCare Lipil [22 cal/fl oz enterocollitis], Mead Johnson Nutrition, Glenview, IL).
The whole-blood concentrations of choline, glycerolphosphocholine, and phosphocholine in the exclusively enterally fed and parenterally fed infants did not differ significantly (Table 3). However, the whole blood concentrations of phosphatidylcholine were significantly greater in the parenterally fed infants (P = 0.04). The mean amount of intravenous lipids administered was 2.8 ± 0.64 g kg−1 day−1, and only 1 infant received <2.5 g kg−1 day−1 (Table 1). There was no consistent correlation between intravenous lipid intake and phosphatidylcholine concentrations among the infants receiving >2.5 g kg−1 day−1. However, the infant who received the least amount of intravenous lipids (0.9 g kg−1 day−1) also had the lowest whole-blood concentrations of phosphatidylcholine (1137 mmol/mL), thus supporting an association. When the subjects were separated on the basis of amount of PN received, 50% of the infants were receiving 90% or more of their daily energy intake from PN. This group on the most PN (≥90% of daily energy intake) had significantly decreased whole-blood concentrations of phosphocholine compared with infants who received <90% of daily energy intake from PN (Fig. 1; P value for trend <0.05) and controls. The whole-blood concentrations of phosphocholine were significantly correlated with glycerolphosphocholine in the exclusively enterally fed infants (controls). However, there was no strong relation found between whole-blood concentrations of phosphocholine and glycerolphosphocholine in the infants on PN. Of note, the same patterns were seen when 1 of the controls with PCho (>60 nmol/mL) and GPCho (>100 nm/mL) was excluded from the analysis as a potential outlier (data not shown).
We found that whole-blood–free choline concentrations did not differ significantly in exclusively enterally fed versus infants on PN therapy. Whole-blood phosphatidylcholine concentrations were significantly increased in infants on PN therapy, possibly related to infusion of intravenous lipids. PN therapy providing >90% of the daily energy intake was associated with decreased phosphocholine concentrations, which suggested a choline-deficient state (32). These metabolic changes are evidence of decreased choline status in infants on chronic PN providing a significant percentage of the daily energy intake.
This report is unique in that whole blood and not plasma concentrations of free choline and related metabolites were measured in infants on PN and exclusively enterally fed infants. Our study measured free choline in whole blood and found higher mean concentrations (Table 3) than 21.8 ± 2.2 nmol/mL based on plasma measurements reported by Ilcol et al (14). Whole-blood measurements reflected the total concentration of free choline in cellular (erythrocytes) and noncellular (plasma) blood compartments. Plasma and whole-blood–free choline concentrations are weakly, but significantly positively, correlated (r = 0.322, P < 0.05) (33). Erythrocyte choline accounts for the major difference between whole-blood– and plasma-free choline concentrations. In healthy adults, the erythrocyte and whole-blood concentrations of free choline are strongly correlated (r = 0.939, P < 0.0001), whereas plasma and erythrocyte-free choline concentrations are less significantly correlated (r = 0.297, P < 0.05) (33). Choline is actively accumulated into erythrocytes independent of plasma concentrations (34); however, choline does not appear to be transported out of the red blood cells (RBCs) when plasma concentration is low, as in a deficient state. Whether this represents a disruption in RBC choline transport or whether it is simply an actively mediated process that does not necessarily respond to a need to maintain homeostasis between intra- and extracellular compartments is unknown.
The biosynthesis of choline starts with s-adenosyl methionine (SAM) produced in the transsulfuration pathway. SAM is involved in the sequential methylation of phospatidylethanolamine to form phosphatidylcholine (13). Phosphotidylcholine is converted to free choline through intermediates lysophosphocholine, glycerolphosphocholine, and phosphocholine (Fig. 2). We found that glycerolphosphocholine concentrations were similar in infants on PN and controls; however, phosphocholine was significantly reduced in infants who received the most PN (Fig. 1). Therefore, the more likely reason for decreased phosphocholine was not impaired degradation of glycerolphosphocholine to phosphocholine but increased use of phosphocholine to generate free choline (Fig. 2). Laboratory animals fed a choline-deficient diet experienced a >80% decline in hepatic concentrations of phosphocholine compared with an approximately 60% decline in free choline (32,35). Thus, phosphocholine is a labile metabolite mainly located in tissue and more sensitive to choline status than concentrations of free choline. The group on prolonged PN was of higher birth weight than controls but had decreased phosphocholine consistent with metabolic evidence of choline deficiency. Five of the 6 infants receiving >90% versus 3 of the 8 receiving <90% of daily energy intake from PN had a history of intestinal surgery. However, infants receiving <90% versus >90% of daily energy intake from PN did not differ in birth weight (1786 ± 1266 vs 2171 ± 982 g, P = NS), weight at enrollment (2660 ± 1332 vs 3528 ± 1193 g, P = NS), or duration of PN therapy (6.24 ± 1.22 vs 6.76 ± 2.42 weeks, P = NS). Therefore, history of intestinal surgery appeared to be associated with greater dependence on PN, and enteral feeding quantities more than trophic volumes may have exerted a beneficial effect on choline status. The controls were a convenience sample of healthy exclusively enterally fed infants of ages at least 4 weeks. Thus, they were mostly low-birth-weight (premature) infants retained in the nursery primarily for clinical stability and weight gain before discharge. Contrary to expectation, they did not demonstrate any metabolic signs of choline deficiency, thus suggesting increased dependence on PN and not birth weight was the main risk factor. Sondheimer et al (10) also found no association between development of PNALD and gestational age or birth weight. Therefore, infants with a prolonged increased requirement for PN may be at greater risk for developing choline deficiency.
Alkaline phosphatase is a membrane-bound metallo-enzyme with several isoforms including hepatic, bone, intestinal, and lipoprotein bound (36). It is the most frequently elevated enzyme during prolonged PN therapy (37–39). The elevation is commonly attributed to hepatobiliary injury and/or metabolic bone disease (39–41). In addition, alkaline phosphatase is essential for the dephosphorylation of phosphocholine to free choline (42) (Fig. 2). Buchman et al (24) examined the response of serum hepatic transaminases and alkaline phosphtase in adult patients on chronic PN providing >80% of nutritional needs that were randomized to continual therapy with routine PN versus PN supplemented with choline. Within 2 weeks, there was a significant decline in serum concentrations of alkaline phosphatase in the group receiving choline supplements. We speculate that this response may have resulted from replenished choline status resulting in decreased need for endogenous biosynthesis of free choline. We did not measure serum or whole-blood alkaline phosphatase concentration, therefore, our current data are unable to confirm the observation of Buchman et al. (24). Further study is warranted to determine whether increased alkaline phosphatase activity in patients on chronic PN also represents a metabolic response to choline deficiency.
PN therapy was associated with increased whole-blood concentrations of phosphatidylcholine, which could be related to therapy with intravenous lipids. The dose of intravenous lipids ranged from 0.9 to 3.5 g kg−1 day−1, and only 1 infant received less than 2.5 g kg−1 day−1. There was no consistent correlation between dose of intravenous lipids and whole-blood concentrations of phosphatidylcholine. However, the infant who received the least dose of lipids (0.9 g kg−1 day−1) also had the lowest whole-blood concentration of phosphatidylcholine. It is assumed that phosphatidylcholine in intravenous lipid emulsions is readily available for biosynthesis of free choline. Sheard et al (23) examined a group of malnourished adult patients on lipid-supplemented and lipid-restricted PN therapy and reported improved choline status in the group that received intravenous lipids. However, the patients were also being fed enterally, which may have contributed to plasma choline concentrations. Misra et al (25) and Buchman et al (22) found evidence of choline deficiency in children on chronic PN, despite therapy with intravenous lipid emulsions. Thus, the phosphatidylcholine in intravenous lipid emulsions may not be readily available for biosynthesis of free choline. Phosphatidylcholine in lipid emulsions is a constituent of the hydrophilic phospholipid membrane enveloping a hydrophobic triglyceride core (liposome) (43). Following hydrolysis and uptake of triglycerides from liposomes, the remaining phospholipids may get taken up by cell membranes, transferred to other plasma lipoproteins, or re-enriched with a cholesterol from other circulating lipoproteins to form lipoprotein X (43). Thus, only a proportion of the phosphatidylcholine present in intravenous lipid emulsions becomes readily available for biosynthesis of choline. Rapid infusion or use of 10% lipid emulsion solutions is associated with increased formation of lipoprotein X (43) and thus potentially even less available phospholipids.
In summary, this study of whole-blood concentrations of free choline and related metabolite in infants on chronic PN and enteral feeds is unique. Our initial intent was to measure plasma-free choline and related metabolites in plasma. However, we found that whole-blood measurements could also be used to define choline deficiency. PN therapy was associated with increased whole-blood phosphatidylcholine concentrations; however, free choline, glycerolphosphocholine, and phosphocholine concentrations did not differ to a significant degree in PN and exclusively enterally fed infants. Phosphocholine, the labile precholine metabolite, was significantly decreased in those receiving a substantial amount (>90%) of their energy intake from PN. Plasma-free choline concentration is a valid indicator of choline status, but when whole-blood measurements are used, phosphocholine concentration may be a more sensitive indicator for choline deficiency. Choline supplementation should be investigated in infants who require prolonged PN and the role of whole-blood PCho concentrations as a marker of choline status further evaluated. The lack of a homeostatic mechanism between RBC and plasma choline also requires further investigation.
1. O'Keefe SJD, Buchman AL, Fishbein TM, et al
. Short bowel syndrome and intestinal failure: consensus definitions and overview. Clin Gastroenterol Hepatol 2006; 4:6–10.
2. Colomb V. Home parenteral nutrition
: the pediatric point of view. Nutrition 1999; 15:172–173.
3. Kelly DA. Intestinal failure-associated liver disease: what do we know today? [see comment]. Gastroenterology 2006; 130(2 suppl 1):S70–S77.
4. Spencer AU, Neaga A, West B, et al
. Pediatric short bowel syndrome: redefining predictors of success. Ann Surg 2005; 242:403–412.
5. Nasr A, Avitzur Y, Ng VL, et al
. The use of conjugated hyperbilirubinemia greater than 100 micromol/L as an indicator of irreversible liver disease in infants with short bowel syndrome. J Pediatr Surg 2007; 42:359–362.
6. Beale EF, Nelson RM, Bucciarelli RL, et al
. Intrahepatic cholestasis associated with parenteral nutrition
in premature infants. Pediatrics 1979; 64:342–347.
7. Pereira GR, Sherman MS, DiGiacomo J, et al
. Hyperalimentation-induced cholestasis. Increased incidence and severity in premature infants. Am J Dis Child 1981; 135:842–845.
8. Bell RL, Ferry GD, Smith EO, et al
. Total parenteral nutrition
-related cholestasis in infants. JPEN J Parenter Enteral Nutr 1986; 10:356–359.
9. Whitington PF. Cholestasis associated with total parenteral nutrition
in infants. Hepatology 1985; 5:693–696.
10. Sondheimer JM, Asturias E, Cadnapaphornchai M. Infection and cholestasis in neonates with intestinal resection and long-term parenteral nutrition
. J Pediatr Gastroenterol Nutr 1998; 27:131–137.
11. Kelly DA. Liver complications of pediatric parenteral nutrition
: epidemiology. Nutrition 1998; 14:153–157.
12. Kaufman SS, Atkinson JB, Bianchi A, et al
. Indications for pediatric intestinal transplantation: a position paper of the American Society of Transplantation. Pediatr Transplant 2001; 5:80–87.
13. Zeisel SH. Choline
: an essential nutrient for humans. Nutrition 2000; 16:669–671.
14. Ilcol YO, Ozbek R, Hamurtekin E, et al
status in newborns, infants, children, breast-feeding women, breast-fed infants and human breast milk. J Nutr Biochem 2005; 16:489–499.
15. Holmes-McNary MQ, Cheng WL, Mar MH, et al
esters in human and rat milk and in infant formulas. Am J Clin Nutr 1996; 64:572–576.
16. Fischer LM, Scearce JA, Mar M-H, et al
. Ad libitum choline
intake in healthy individuals meets or exceeds the proposed adequate intake level. J Nutr 2005; 135:826–829.
17. Chawla RK, Berry CJ, Kutner MH, et al
. Plasma concentrations of transsulfuration pathway products during nasoenteral and intravenous hyperalimentation of malnourished patients. Am J Clin Nutr 1985; 42:577–584.
18. Zeisel SH, Da Costa KA, Franklin PD, et al
, an essential nutrient for humans. FASEB J 1991; 5:2093–2098.
19. Koppe SWP, Sahai A, Malladi P, et al
. Pentoxifylline attenuates steatohepatitis induced by the methionine choline
deficient diet. J Hepatol 2004; 41:592–598.
20. Sundaram SS, Whitington PF, Green RM. Steatohepatitis develops rapidly in transgenic mice overexpressing Abcb11 and fed a methionine-choline
-deficient diet. Am J Physiol Gastrointest Liver Physiol 2005; 288:G1321–G1327.
21. Zeisel SH, Blusztajn JK. Choline
and human nutrition. Annu Rev Nutr 1994; 14:269–296.
22. Buchman AL, Moukarzel A, Jenden DJ, et al
. Low plasma free choline
is prevalent in patients receiving long term parenteral nutrition
and is associated with hepatic aminotransferase abnormalities. Clin Nutr 1993; 12:33–37.
23. Sheard NF, Tayek JA, Bistrian BR, et al
. Plasma choline
concentration in humans fed parenterally. Am J Clin Nutr 1986; 43:219–224.
24. Buchman AL, Ament ME, Sohel M, et al
. Choline deficiency
causes reversible hepatic abnormalities in patients receiving parenteral nutrition
: proof of a human choline
requirement—a placebo-controlled trial. JPEN J Parenter Enteral Nutr 2001; 25:260–268.
25. Misra S, Ahn C, Ament ME, et al
. Plasma choline
concentrations in children requiring long-term home parenteral nutrition
: a case control study. JPEN J Parenter Enteral Nutr 1999; 23:305–308.
26. Buchman AL, Dubin MD, Moukarzel AA, et al
. Choline deficiency
: a cause of hepatic steatosis during parenteral nutrition
that can be reversed with intravenous choline
supplementation. Hepatology 1995; 22:1399–1403.
27. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 1959; 37:911–917.
28. Koc H, Mar M-H, Ranasinghe A, et al
. Quantitation of choline
and its metabolites in tissues and foods by liquid chromatography/electrospray ionization-isotope dilution mass spectrometry. Anal Chem 2002; 74:4734–4740.
29. Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985; 4:87–90.
30. FAO/WHO/UNU. Expert Consultation. Energy and protein requirements. WHO Tech Rep Ser
31. National Research Council. Recommended Dietary Allowances
. 10th ed. Washington, DC: National Academy Press; 1989.
32. Pomfret Ed KA, Zeisel SH. Effects of choline deficiency
and methotrexate treatment upon rat liver. J Nutr Biochem 1990; 1:533–541.
33. Ilcol Y, Hizli B, Eroz E, Ulus I. Plasma, whole blood, erythrocyte and urine free choline
concentrations in Turkish healthy subjects. FEBS 2007; 274(suppl 1):286 [abstr].
34. Martin K. Concentrative accumulation of choline
by human erythrocytes. J Gen Physiol 1968; 51:497–516.
35. Zeisel SH, Mar MH, Zhou Z, et al
. Pregnancy and lactation are associated with diminished concentrations of choline
and its metabolites in rat liver. J Nutr 1995; 125:3049–3054.
36. Van Hoof VO, De Broe ME. Interpretation and clinical significance of alkaline phosphatase isoenzyme patterns. Crit Rev Clin Lab Sci 1994; 31:197–293.
37. Luman W, Shaffer JL. Prevalence, outcome and associated factors of deranged liver function tests in patients on home parenteral nutrition
. Clin Nutr 2002; 21:337–343.
38. Salvino R, Ghanta R, Seidner DL, et al
. Liver failure is uncommon in adults receiving long-term parenteral nutrition
. JPEN J Parenter Enteral Nutr 2006; 30:202–208.
39. 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; 27:284–290.
40. Lipkin EW, Ott SM, Klein GL, et al
. Serum markers of bone formation in parenteral nutrition
patients. Calcif Tissue Int 1990; 47:75–81.
41. MacMahon P, Blair ME, Treweeke P, et al
. Association of mineral composition of neonatal intravenous feeding solutions and metabolic bone disease of prematurity. Arch Dis Child 1989; 64:489–493.
42. Murai S, Saito H, Shirato R, et al
. An improved method for assaying phosphocholine
and glycerophosphocholine in mouse tissue. J Pharmacol Toxicol Methods 2001; 46:103–109.
43. Ferezou J, Bach AC. Structure and metabolic fate of triacylglycerol- and phospholipid-rich particles of commercial parenteral fat emulsions. Nutrition 1999; 15:44–50.