Fat malabsorption is well recognized in children and adults with cholestatic liver disease. Decreased bile flow and subsequent decreased intraluminal bile salt concentrations below critical micellar concentration result in malabsorption of fat-soluble molecules, with the degree of malabsorption generally correlating with severity of cholestasis (1). Deficiencies of fat-soluble vitamins A/E/D/K may be present in 20% to 35% of patients with cholestatic liver disease resulting in skin and visual changes, spinocerebellar syndrome, rickets and osteoporosis, and coagulopathy. Many patients are asymptomatic at presentation, stressing the importance of early and adequate supplementation to help prevent long-term consequences (1–4).
Vitamin K is present naturally as phylloquinones synthesized by green plants and menaquinones produced by intestinal bacteria. Dietary phylloquinones provide the main source of vitamin K in humans and, like other fat-soluble vitamins, require incorporation into micelles within the intestinal lumen for efficient absorption. Vitamin K is rapidly metabolized so malabsorption may lead to rapid depletion of limited stores, leading to clinically significant deficiency. Vitamin K serves as a cofactor for the enzyme γ-glutamylcarboxylase, which supports posttranslational carboxylation of glutamic acid residues in many proteins, including coagulation factors such as factors II (prothrombin), VII, IX, X, protein C, protein S, and proteins involved in bone homeostasis, specifically osteocalcin and matrix gla protein. Most of these proteins have multiple carboxylations, and these carboxylated glutamic acid residues have the ability to bind calcium, making the proteins functionally active (1,5). In vitamin K deficiency, there is an increase in abnormal undercarboxylated or partially carboxylated forms of these proteins, which are functionally defective. After these undercarboxylated proteins, particularly undercarboxylated prothrombin, reach a poorly defined level, vitamin K deficiency results in coagulopathy, as measured by prolongation of prothrombin time (PT). Vitamin K deficiency may also contribute to bone disease through inadequate carboxylation of bone matrix proteins, specifically osteocalcin, which represents 15% to 20% of noncollagenous bone. In adults, associations have been demonstrated between low vitamin K intake and decreased bone mineral density (6) and increased risk of hip fractures (7,8).
Traditionally, vitamin K status has been assessed by surrogate markers such as the PT, which measure activity, rather than vitamin K presence. This may greatly underestimate the actual incidence of vitamin K deficiency, as the PT is a late indicator of vitamin K deficiency and can be normal even when prothrombin concentrations fall to half normal values (9). Thus, many patients with cholestasis may have subclinical vitamin K deficiency, placing them at potential risk for bleeding complications and altered skeletal health. New methods of measuring undercarboxylated vitamin K dependent proteins, such as protein induced in vitamin K absence-II (PIVKA-II), are a more sensitive measure of vitamin K status, enabling the clinician to identify early deficiency and subsequently correct vitamin K nutritional status (10,11). The primary aim of this work was to determine the frequency of vitamin K deficiency in cholestatic children and adults as measured by PIVKA-II and assess the relation between PIVKA-II and serum conjugated bilirubin, various serum markers of liver injury (γ-glutamyl transpeptidase [GGT], aspartate aminotransferase [AST], alanine aminotransferase [ALT], alkaline phosphatase), serum bile acids, PT, international normalized ratio (INR), and serum undercarboxylated osteocalcin (ucOC).
PATIENTS AND METHODS
This was a cross-sectional study of patients with cholestatic liver disease recruited from the Gastroenterology/Hepatology clinics at Cincinnati Children's Hospital Medical Center (CCHMC) and the University of Cincinnati. Patients were identified from records of outpatient and inpatient encounters with members of the Division of Pediatric Gastroenterology, Hepatology, and Nutrition at CCHMC and the Division of Digestive Diseases of the University of Cincinnati College of Medicine. Informed consent was obtained from all participants. For patients younger than 18 years, parents provided informed consent and patient's assent was obtained from minors older than 11 years. The study was approved by the institutional review boards of CCHMC and the University of Cincinnati, and the scientific advisory committee of the General Clinical Research Center of CCHMC.
Patients represented a convenience sample of available children and adults with cholestasis defined as direct bilirubin >1 mg/dL and/or serum bile acids >30 μmol/L. Patients of either sex were eligible. Inclusion diagnoses were biliary atresia, primary sclerosing cholangitis, α-1 antitrypsin deficiency, neonatal hepatitis, progressive familial intrahepatic cholestasis (PFIC I, II, or III), Alagille syndrome, primary biliary cirrhosis, chronic hepatitis, and idiopathic cirrhosis. Exclusion criteria were concurrent disease known to result in fat malabsorption, such as pancreatic insufficiency, short gut syndrome, and intestinal lymphangiectasia; malignancy; age younger than 6 months or older than 65 years; AIDS; decompensated cirrhosis characterized by intractable ascites and severe coagulopathy (defined as INR >2 not responsive to parental vitamin K therapy); and weight <6 kg. Patients taking antibiotics were not excluded. A brief questionnaire, including such items as age of diagnosis, specific diagnosis, medications, including vitamin supplementations, and hospitalizations and surgeries, was administered by the primary investigator (J.S.). Supplemental information was obtained by reviewing records at the CCHMC.
The patient's blood was drawn for liver biochemistry tests (AST, ALT, alkaline phosphatase, GGT, albumin, total protein, conjugated and unconjugated bilirubin), PT, INR, serum bile acids (normal range 7–9 months: 1–28 μmol/L; 10–12 months: 1–37 μmol/L; >1 year: 1–7 μmol/L), 25-hydroxyvitamin D (normal >32 ng/mL), vitamin A (normal range <1 year: 200–490 μg/L; >1 year: 200–520 μg/L), vitamin E (normal range 3–15 μg/mL), PIVKA II, and ucOC. Liver biochemistry tests, PT, INR, serum bile acids, serum 25-hydroxyvitamin D, serum vitamin A, and serum vitamin E levels were measured using standard laboratory techniques in the laboratories of CCHMC.
Plasma PIVKA-II concentrations were measured using a commercially available enzyme-linked immunosorbent assay kit (Diagnostica Stago; Asnier, France), which uses a mouse monoclonal antibody specific for undercarboxylated prothrombin, but without cross-reactivity to fully carboxylated prothrombin. The lowest limit of detection for this assay is 2 ng/mL. Values <2 ng/mL are considered normal; however, for the purposes of this research protocol, values <2 times the lower limit of detection (4 ng/mL) were considered normal. All control samples were up to 2 standard deviations below this value. All samples were run in duplicate on 2 separate days. Intraassay variability was 6.2% and interassay variability was 10%. Serum ucOC was measured by radioimmunoassay performed by Caren Gundberg (12) at Yale University using previously published techniques. The average value for ucOC is 35% (±15%) in healthy adults and 20% (±18%) in individuals supplemented with vitamin K.
A sample size of 30 was calculated to detect a correlation of at least 0.5 between plasma PIVKA-II and markers of cholestasis. This sample size calculation was based on a 2-sided test with an alpha error of 0.05 and power of 0.80. All statistical analyses were performed using SAS software (version 8.3, SAS Institute, Cary, NC). Because the data were not normally distributed, Spearman correlations were determined among plasma PIVKA-II, serum ucOC, and other variables. Partial correlations were used to assess the relation between plasma PIVKA-II, serum ucOC, and other measured parameters while controlling for vitamin K intake. For the purpose of analysis, all laboratory values below the limit of detection for a particular assay were recorded as the lowest value for that assay. All P values were 2-sided.
Thirty-three patients were enrolled in this study. Two patients (1 with biliary atresia and 1 with idiopathic cholestasis) withdrew due to difficulty obtaining a sufficient quantity of blood for study laboratories. Of the 31 patients who fulfilled eligibility criteria, 24 (77%) were children (<18 years old) and 17 (55%) were female. The mean age of the study patients was 12.8 years (median 5.7, range 0.7–54.1 years) (Table 1). The most common diagnoses were Alagille syndrome (n = 10) and biliary atresia (n = 9), accounting for 60% of the study population. Other diagnoses included primary biliary cirrhosis (4), progressive familial intrahepatic cholestasis (2), congenital hepatic fibrosis (1), primary sclerosing cholangitis (1), α-1 antitrypsin deficiency (1), hepatitis C (1), and idiopathic cholestasis (2). Mean serum conjugated bilirubin and bile acids were 1.8 mg/dL and 276 μmol/L, respectively. Other laboratory values for study patients are listed in Table 1.
Of the 31 patients, 28 had all laboratories drawn at the time of study enrollment. Three patients did not have PT or INR results available on the day of study. One patient had a PT and INR drawn within 2 days of initial laboratory assessment and this value was used in the analysis. Laboratory records were reviewed for the other 2 patients and both had normal PT/INR before and after study laboratories and a stable medical history without change in medications or clinical status. For these patients, the PT and INR values obtained closest to the time of study enrollment were used for analysis. Of the 31 patients, 9 (29%) had increased INRs, defined as >1.3 (Tables 2 and 3). Twenty-one patients (67%) had an elevated (abnormal) plasma PIVKA-II (Tables 2 and 3). All patients with increased INRs had increased plasma PIVKA-II. The highest plasma PIVKA-II (5831.4 ng/mL) and most elevated INR (3.45) were observed in a pediatric patient with Alagille syndrome. This occurred despite receiving supplemental oral vitamin K at a dosage of 162.3 μg/kg/day, and corrected after receiving parenteral vitamin K. Of the 21 patients with plasma PIVKA-II elevation, 15 (71%) were receiving supplemental vitamin K therapy, with dosages ranging from 7.8 to 700 μg/kg/day. In patients with normal plasma PIVKA-II, 9 (90%) were receiving supplemental vitamin K (dosage range 10–289 μg/kg/day). Seven patients (6 adults, 1 child) were receiving no supplemental vitamin K and 6 (86%) of these patients had elevated PIVKA-II levels.
Other fat-soluble vitamin deficiencies were observed in 77% of the study participants. The most common was vitamin D deficiency as assessed by serum 25-hydroxyvitamin D, observed in 22 (71%) of patients (Table 2). Nine (29%) patients had decreased serum vitamin A levels, whereas 2 patients (6%) had decreased serum vitamin E levels (Table 2). Two patients had both vitamin D and vitamin E deficiency, whereas 7 patients had vitamin D and vitamin A deficiency. In total, 56% of patients with vitamin A deficiency, 77% of patients with 25-hydroxyvitamin D deficiency, and 100% of patients with vitamin E deficiency had elevated PIVKA-II.
When we evaluated the relation between diagnosis and increased INR and plasma PIVKA-II levels (Table 3), we found a consistently higher frequency of increased PIVKA-II level compared with INR increases across all diagnoses. This may be a direct reflection of severity of cholestasis with a strong correlation among serum bile acids, serum conjugated bilirubin, and PIVKA-II (P = 0.002 and P < 0.001, respectively) (Table 4).
To assess whether markers of the severity of an individual's cholestatic liver disease such as serum conjugated bilirubin and elevated INR were associated with elevated plasma PIVKA-II, Spearman correlations were performed between plasma PIVKA-II, serum ucOC, and other measured parameters. Plasma PIVKA-II directly correlated with PT (P = 0.02), INR (P = 0.02), and the severity of cholestasis as measured by serum conjugated bilirubin (P < 0.001) and serum bile acids (P = 0.002) (Table 4, Fig. 1). Additional positive correlations were observed between plasma PIVKA-II and serum ucOC, serum AST, ALT, and alkaline phosphatase. There was an inverse correlation between plasma PIVKA-II and serum 25-hydroxyvitamin D levels (P = 0.01) (Fig. 1). No significant correlations were found between plasma PIVKA-II and other measured laboratories (Table 4).
Twenty-nine patients had serum analyzed for ucOC, with a mean ucOC of 25% (median 21.3%, range 0%–57.5%). In total, 14 (48%) patients had abnormal values. One (14%) of the unsupplemented individuals had abnormal ucOC values (defined as >35%, value 40.8%), whereas 13 (59%) patients taking supplemental vitamin K had abnormal ucOC values (defined as >20%, range 21%–57.5%). Of those patients with abnormal ucOC values, 100% of unsupplemented patients and 77% of supplemented patients had elevated PIVKA-II. Spearman correlations were performed between serum ucOC and other measured parameters. Serum ucOC directly correlated with plasma PIVKA-II (P = 0.003) and serum albumin (P = 0.004), and inversely correlated with serum vitamin A (P < 0.01). No significant correlation was observed between PT, INR, or markers of cholestasis such as conjugated bilirubin or bile acids.
Because most patients (77%) were taking oral vitamin K supplements, Spearman correlations were also determined between supplemental vitamin K intake and other measured parameters. Vitamin K intake was positively correlated with serum bile acids and alkaline phosphatase, and the correlation between vitamin K intake and serum conjugated bilirubin approach significance (Table 5), indicating patients with more significant cholestatic liver disease were more likely to be taking higher dosages of supplemental vitamin K. However, no correlation was observed between vitamin K intake and PT, INR, plasma PIVKA-II, or serum ucOC (Table 5). Subsequently, Spearman correlations between plasma PIVKA-II, serum ucOC and other laboratories were reanalyzed, controlling for vitamin K intake. Apart from the correlation between PIVKA-II and alkaline phosphatase, all previously correlated parameters persisted, with the correlations between plasma PIVKA-II, serum ucOC, conjugated bilirubin, and bile acids remaining most significant (data not shown). Similarly, significant correlations between serum ucOC with previously noted parameters remained when controlling for vitamin K intake (data not shown).
Fat and fat-soluble vitamin malabsorption in cholestatic liver disease results from intraluminal bile salt concentrations falling below the critical micellar level. Deficiencies of fat-soluble vitamin deficiencies are a well-known complication of cholestasis. The results of the current study indicate that vitamin K deficiency, as assessed by plasma PIVKA-II, affects two thirds of cholestatic children and adults, with 77% having deficiencies of other fat-soluble vitamins. Of note, since vitamin E/lipid ratios were not measured, we may have underestimated vitamin E deficiency in our population.
Increased undercarboxylated prothrombin or decreased vitamin K levels have previously been identified in adults with cholestatic liver disease (13–15). Although there were discrepancies between measured vitamin K levels and PT, a limitation of these studies is that plasma vitamin K levels may not be representative of total body vitamin K stores (5). Most studies on adults have included only primary biliary cirrhosis, and therefore, the findings may not be generalizable to other cholestatic liver diseases, particularly those causing childhood cholestasis. There is a single study evaluating vitamin K deficiency in childhood cholestasis in which elevated plasma PIVKA-II was found in 48% of children of mild to moderate cholestatic liver disease, with only 4.7% having prolonged PT. Plasma PIVKA-II was positively correlated with markers of cholestasis and severity of liver disease as measured by the Child-Turcotte-Pugh classification (16). The present study similarly indicates that vitamin K deficiency, as determined by elevated PIVKA-II levels, is common in cholestatic liver disease despite oral vitamin K supplementation. Prothrombin time, a surrogate marker of vitamin K status, identified only 43% of those with increased plasma PIVKA-II, suggesting that PT underestimates the prevalence of vitamin K deficiency.
Deficiencies of other fat-soluble vitamins were present in 77% of the study patients, with vitamin D deficiency in 71% and vitamin A deficiency in 29%. Vitamin E deficiency occurred less frequently. Overall, the prevalence of vitamin D deficiency in the present study is higher than in previous reports (2,4,13,15,17,18). This discrepancy is likely due to our use of a more stringent definition of vitamin D deficiency defined as 25-hydroxyvitamin D levels <32 ng/mL. Values of 25-hydroxyvitamin D >32 ng/mL are now recognized by authorities in the field as the level required for optimal calcium absorption and bone health, and thus this value has become the standard for vitamin D sufficiency (19). Studies in children, adolescents, and adults have shown an inverse relation between serum PTH and 25-hydroxyvitamin D at levels below 32 ng/mL (20–23). The prevalence of vitamin A and E deficiency in our study is consistent with some previously reported frequencies in patients with cholestatic liver disease (2,4,13,15,18), whereas others have reported a higher prevalence (3,17,24–27). These differences may relate to the methods used to assess deficiency, particularly for vitamin E. Cholestatic liver disease can lead to hyperlipidemia, which may artificially raise vitamin E levels to normal values because vitamin E is carried in the high-density lipoprotein and low-density lipoprotein fraction of lipoproteins, thus masking true vitamin E deficiency (28,29). The measurement of the ratio of serum vitamin E to total lipids is a more accurate indicator of vitamin E status (29), and because serum total lipids, cholesterol, and triglycerides were not measured in our study, vitamin E deficiency may be underestimated in our study population. In the present study, vitamin A deficiency may be overestimated because vitamin A levels were not normalized for serum retinol binding protein, which may be reduced in decompensated liver disease (30). Because few patients had evidence of advanced hepatic decompensation as assessed by reduced serum albumin concentrations, this may not represent a significant problem in our study population.
Previous studies have found associations between vitamin K deficiency, either measured by low plasma vitamin K levels (15) or elevated undercarboxylated prothrombin (13,16), and the severity of cholestasis in adults as determined by elevated conjugated bilirubin, serum bile acids, and alkaline phosphatase. This study confirmed these findings, and we identified correlations between plasma PIVKA-II and serum aminotransferases. In the present study, the 100% concordance between prolonged PTs and elevated plasma PIVKA-II levels, significant correlation between plasma PIVKA-II and serum ucOC, another vitamin K–dependent protein, and the presence of other fat-soluble vitamin deficiencies in 81% of the patients with increased plasma PIVKA-II strongly suggests that abnormal PIVKA-II was the result of vitamin K deficiency. Although we cannot categorically conclude that malabsorption was the cause of the elevated PIVKA-II, given the above associations, this is the most plausible explanation.
Not surprisingly, vitamin K intake was positively correlated with the severity of cholestasis, suggesting that clinicians recommend supplementation in the presence of increasing cholestasis; however, plasma PIVKA-II elevation was common in these patients despite vitamin K therapy. It would be expected that a negative correlation between vitamin K intake and PT, INR, and PIVKA-II levels would be identified, but in fact there were no correlations found. This suggests that even with supplementation, intake is not sufficient to overcome the absorptive defect in these patients and current practices of oral vitamin K supplementation may be inadequate to maintain vitamin K nutriture in cholestatic liver disease. A new mixed micellar preparation of vitamin K1 (Kanakion MM) is available in Europe, but unfortunately, absorption of this formulation in cholestatic infants remains poor and unpredictable (31). In addition, few clinicians recommend parenteral vitamin K unless extremely prolonged PTs develop. Future studies examining the dosage and route of vitamin K administration are, therefore, indicated.
Clinical characteristics and medical management may differ between adult and pediatric patients. In the present study, there was a trend for the pediatric patients to be more cholestatic, with higher conjugated bilirubin and serum bile acids, factors that are highly correlated with PIVKA-II elevation. Additionally, pediatric patients were more likely to be receiving supplemental vitamin K replacement. Only 1 adult patient with Alagille syndrome studied at the pediatric center was receiving vitamin K therapy.
This study has some limitations. The sample size is small, particularly in regards to the adult population. Although the study had sufficient power to evaluate correlations between plasma PIVKA-II and markers of cholestasis within the study population, differences between various cholestatic conditions and adult and pediatric patients and predictors of plasma PIVKA-II elevation within these subgroups could not be assessed. The study patients are a convenience sample and may not be completely representative of children and adults with cholestasis; however, there is no reason to believe that our patients would not be representative of all pediatric and adult cholestatic patients. Measurement of serum lipids and retinal binding protein were not performed, which may have influenced the identified prevalence of deficiencies of vitamin E and A in this population. Finally, many patients (58%) were taking antibiotics at the time of study enrollment, which may alter intestinal bacteria and thus influence menoquinone production. However, most of these patients (89%) were receiving oral vitamin K supplementation and it is doubtful that the antibiotics had any measurable effect on their vitamin K metabolism.
In summary, despite vitamin K supplementation elevated plasma PIVKA-II is common in cholestatic liver disease and frequently accompanies other fat-soluble vitamin deficiencies, suggesting continued vitamin K deficiency in these patients. Patients with advanced cholestasis, with increased serum conjugated bilirubin and serum bile acids, are at highest risk for PIVKA-II elevation. Prothrombin time may underestimate the prevalence of vitamin K deficiency as determined by elevated plasma PIVKA-II. Better strategies for vitamin K assessment and guidelines for specific dosing in cholestatic liver disease should be developed.
The authors would like to acknowledge the following individuals for their thoughtful comments, advice, and their assistance in this project: the physicians and nurses of the Division of Gastroenterology, Hepatology, and Nutrition at CCHMC, Kenneth Sherman, MD, Stephen Zucker, MD, Caren Gundberg, PhD, the General Clinical Research Center nurses and staff, and all of the parents and patients.
1. Sokol RJ. Fat-soluble vitamins and their importance in patients with cholestatic liver diseases. Gastroenterol Clin North Am 1994; 23:673–705.
2. Phillips JR, Angulo P, Petterson T, et al. Fat-soluble vitamin levels in patients with primary biliary cirrhosis. Am J Gastroenterol 2001; 96:2745–2750.
3. Sokol RJ, Kim YS, Hoofnagle JH, et al. Intestinal malabsorption of vitamin E in primary biliary cirrhosis. Gastroenterology 1989; 96:479–486.
4. Munoz SJ, Heubi JE, Balistreri WF, et al. Vitamin E deficiency in primary biliary cirrhosis: gastrointestinal malabsorption, frequency and relationship to other lipid-soluble vitamins. Hepatology 1989; 9:525–531.
5. Shearer MJ. Vitamin K. Lancet 1995; 345:229–234.
6. Booth SL, Broe KE, Gagnon DR, et al. Vitamin K intake and bone mineral density in women and men. Am J Clin Nutr 2003; 77:512–516.
7. Feskanich D, Weber P, Willett WC, et al. Vitamin K intake and hip fractures in women: a prospective study. Am J Clin Nutr 1999; 69:74–79.
8. Booth SL, Tucker KL, Chen H, et al. Dietary vitamin K intakes are associated with hip fracture but not with bone mineral density in elderly men and women. Am J Clin Nutr 2000; 71:1201–1208.
9. Sokoll LJ, Sadowski JA. Comparison of biochemical indexes for assessing vitamin K nutritional status in a healthy adult population. Am J Clin Nutr 1996; 63:566–573.
10. Widdershoven J, van Munster P, De Abreu R, et al. Four methods compared for measuring des-carboxy-prothrombin (PIVKA-II). Clin Chem 1987; 33:2074–2078.
11. Motohara K, Kuroki Y, Kan H, et al. Detection of vitamin K deficiency by use of an enzyme-linked immunosorbent assay for circulating abnormal prothrombin. Pediatr Res 1985; 19:354–357.
12. Gundberg CM, Nieman SD, Abrams S, et al. Vitamin K status and bone health: an analysis of methods for determination of undercarboxylated osteocalcin. J Clin Endocrinol Metab 1998; 83:3258–3266.
13. Kaplan MM, Elta GH, Furie B, et al. Fat-soluble vitamin nutriture in primary biliary cirrhosis. Gastroenterology 1988; 95:787–792.
14. O'Brien DP, Shearer MJ, Waldron RP, et al. The extent of vitamin K deficiency in patients with cholestatic jaundice: a preliminary communication. J R Soc Med 1994; 87:320–322.
15. Kowdley KV, Emond MJ, Sadowski JA, et al. Plasma vitamin K1 level is decreased in primary biliary cirrhosis. Am J Gastroenterol 1997; 92:2059–2061.
16. Mager DR, McGee PL, Furuya KN, et al. Prevalence of vitamin K deficiency in children with mild to moderate chronic liver disease. J Pediatr Gastroenterol Nutr 2006; 42:71–76.
17. Argao EA, Specker BL, Heubi JE. Bone mineral content in infants and children with chronic cholestatic liver disease. Pediatrics 1993; 91:1151–1154.
18. Jorgensen RA, Lindor KD, Sartin JS, et al. Serum lipid and fat-soluble vitamin levels in primary sclerosing cholangitis. J Clin Gastroenterol 1995; 20:215–219.
19. Hollis BW. Circulating 25-hydroxyvitamin D levels indicative of vitamin D sufficiency: implications for establishing a new effective dietary intake recommendation for vitamin D. J Nutr 2005; 135:317–322.
20. Guillemant J, Cabrol S, Allemandou A, et al. Vitamin D-dependent seasonal variation of PTH in growing male adolescents. Bone 1995; 17:513–516.
21. Bowden SA, Robinson RF, Carr R, et al. Prevalence of vitamin D deficiency and insufficiency in children with osteopenia or osteoporosis referred to a pediatric metabolic bone clinic. Pediatrics 2008; 121:e1585–e1590.
22. El-Hajj Fuleihan G, Nabulsi M, Choucair M, et al. Hypovitaminosis D in healthy schoolchildren. Pediatrics 2001; 107:E53.
23. Rajakumar K, Fernstrom JD, Janosky JE, et al. Vitamin D insufficiency in preadolescent African-American children. Clin Pediatr (Phila) 2005; 44:683–692.
24. Herlong HF, Russell RM, Maddrey WC. Vitamin A and zinc therapy in primary biliary cirrhosis. Hepatology 1981; 1:348–351.
25. Herlong HF, Recker RR, Maddrey WC. Bone disease in primary biliary cirrhosis: histologic features and response to 25-hydroxyvitamin D. Gastroenterology 1982; 83:103–108.
26. Sokol RJ, Balistreri WF, Hoofnagle JH, et al. Vitamin E deficiency in adults with chronic liver disease. Am J Clin Nutr 1985; 41:66–72.
27. Jeffrey GP, Muller DP, Burroughs AK, et al. Vitamin E deficiency and its clinical significance in adults with primary biliary cirrhosis and other forms of chronic liver disease. J Hepatol 1987; 4:307–317.
28. Behrens WA, Thompson JN, Madere R. Distribution of alpha-tocopherol in human plasma lipoproteins. Am J Clin Nutr 1982; 35:691–696.
29. Sokol RJ, Heubi JE, Iannaccone ST, et al. Vitamin E deficiency with normal serum vitamin E concentrations in children with chronic cholestasis. N Engl J Med 1984; 310:1209–1212.
30. Mourey MS, Siegenthaler G, Amedee-Manesme O. Regulation of metabolism of retinol-binding protein by vitamin A status in children with biliary atresia. Am J Clin Nutr 1990; 51:638–643.
31. Pereira SP, Shearer MJ, Williams R, et al. Intestinal absorption of mixed micellar phylloquinone (vitamin K1) is unreliable in infants with conjugated hyperbilirubinaemia: implications for oral prophylaxis of vitamin K deficiency bleeding. Arch Dis Child Fetal Neonatal Ed 2003; 88:F113–F118.
© 2009 Lippincott Williams & Wilkins, Inc.