Journal of Hepatology 41 (2004) 3843 www.elsevier.

com/locate/jhep

Normal protein diet for episodic hepatic encephalopathy: results of a randomized study
pez-Hell n2, Merce rdoba1,*, Juan Lo Planas3, Pilar Sab n4, Francesc Sanpedro1, Juan Co Francisco Castro1, Rafael Esteban1, Jaume Guardia1
2 1 ` noma de Barcelona, Barcelona, Spain Servei de Medicina Interna-Hepatologia, Hospital Universitari Vall dHebron, Barcelona, Universitat Auto mica i Biologia Molecular, Hospital Universitari Vall dHebron, Barcelona, Universitat Auto ` noma de Barcelona, Centre dInvestigacions en Bioqu Barcelona, Spain 3 ` noma de Barcelona, Barcelona, Spain Unitat de Suport Nutricional, Hospital Universitari Vall dHebron, Barcelona, Universitat Auto 4 ` noma de Barcelona, Barcelona, Spain Servei de Farmacia, Hospital Universitari Vall dHebron, Barcelona, Universitat Auto

See Editorial, pages 147 148

Background/Aims: Protein-restricted diets are usually prescribed for cirrhotic patients with hepatic encephalopathy. However, protein restriction may worsen the nutritional status without resulting in an improvement of hepatic encephalopathy. We designed a study to assess the effects of the amount of protein in the diet on the evolution of episodic hepatic encephalopathy. Methods: Cirrhotics admitted to the hospital because of an episode of encephalopathy (n 5 30) were randomized to receive a low-protein diet with progressive increments or a normal protein diet for 14 days, in addition to standard measures to treat hepatic encephalopathy. Protein synthesis and breakdown were studied at day 2 and day 14 with the glycine-N15 infusion method. Results: The outcome of hepatic encephalopathy was not signicantly different between both groups of treatment. Protein synthesis was similar for low and normal protein diet, but those of the low-protein diet group showed higher protein breakdown. Conclusions: Diets with a normal content of protein, which are metabolically more adequate, can be administered safely to cirrhotic patients with episodic hepatic encephalopathy. Restriction of the content of protein of the diet does not appear to have any benecial effect for cirrhotic patients during an episode of encephalopathy. q 2004 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Encephalopathy; Low-protein diet; Ammonia

1. Introduction Hepatic encephalopathy is a common complication of cirrhosis that accounts for a large number of hospital admissions. Treatment of encephalopathy is based on the correction of precipitating factors and administration of drugs to decrease intestinal generation of ammonia [1]. Protein restriction has classically been considered a mainstay of treatment in hepatic encephalopathy [2].
Received 29 October 2003; received in revised form 17 February 2004; accepted 5 March 2004; available online 20 April 2004 * Corresponding author. Tel.: 34-93-274-6140; fax: 34-93-274-6068. rdoba). E-mail address: jcordoba@vhebron.net (J. Co

However, this recommendation has been contested due to the lack of scientic proof [3]. The inclusion of protein restriction as a component of therapy for hepatic encephalopathy has been based on old uncontrolled observations, most of them anecdotal [4]. There have been no prior studies of cirrhotic patients admitted to the hospital because of an episode of hepatic encephalopathy in which patients were randomized to receive different amounts of protein in the diet. Cirrhotic patients exhibit increased protein requirements to achieve balanced nitrogen metabolism [5]. In these patients, malnutrition has been associated with decreased survival [6]. Thus, limiting the amount of protein may worsen their clinical condition. Indeed, in severe alcoholic

0168-8278/$30.00 q 2004 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2004.03.023

 rdoba et al. / Journal of Hepatology 41 (2004) 3843 J. Co

39

liver disease, high-protein diets may result in a better outcome [7]. These data have led the European Society for Parenteral and Enteral Nutrition (ESPEN) to acknowledge the importance that an adequate amount of protein may have on the outcome of cirrhotic patients [8]. The current recommendation is to limit protein restriction during an episode of hepatic encephalopathy to a moderate intake (0.5 g/kg/day) and to shortly thereafter return to a normalto-high protein intake (1 1.5 g/kg/day). For all these reasons we designed a randomized study to assess the effects of the amount of protein in the diet on the evolution of the mental state during an episode of hepatic encephalopathy.

2. Methods 2.1. Participants

We selected our patients from subjects presenting to the emergency room of Hospital Vall dHebron, Barcelona, Spain. Patients were considered eligible if they were cirrhotic patients showing clear signs of episodic hepatic encephalopathy. The diagnosis of cirrhosis was based on prior liver biopsy or on a combination of clinical and imaging data demonstrating liver failure, portal hypertension and cirrhotic appearance of the liver. Episodic hepatic encephalopathy was diagnosed on the basis of a sudden change in mental state from a previously normal consciousness that could not be attributed to another cause. Exclusion criteria to participate in the study were: impossibility to provide nasogastric feeding (e.g. gastrointestinal bleeding, intestinal ileum), contraindications for low-protein diets (e.g. acute alcoholic hepatitis), terminal disease (e.g. advanced hepatocarcinoma, hepatorenal syndrome), clinical instability requiring respiratory or cardiovascular support and neurological comorbidities (e.g. prior cerebrovascular disease, dementia) or intake of toxins (e.g. alcohol toxicity, benzodiazepines) that would make the assessment of mental state difcult. The study was approved by the Institutional Review Board. Informed consent was initially given by next of kin and later conrmed by the patient. Insertion of a nasogastric tube was justied by the need to provide adequate nutritional support in patients that could show a worse outcome if insufciently nourished [9].

accordance with the recommendations of the ESPEN [8]. They received 0 g of protein for the rst 3 days, then the amount of protein was increased progressively every 3 days (12, 24 and 48 g) up to 1.2 g/kg/day for the last 2 days. The normal protein group (group B) received 1.2 g/kg/day from the rst day. Both groups received the same amount of calories to control for this variable. Prior studies in alcoholic liver disease have shown a poorer outcome for patients that received low amount of protein and calories [5]; it is not possible to differentiate which one of the two factors determined the clinical outcome. Due to difculties in estimating dry body weight (without ascites or edema), ideal body weight was used for all calculations. Ideal body weight was calculated from height, gender and frame size, according to standard values of the Spanish population. Gastric retention was assessed every 6 h, and the velocity of the pump was adjusted according to the amount of retention in order to minimize the risk of aspiration. Metoclorpramide and omeprazole were systematically administered to improve enteric tolerance and to prevent gastrointestinal bleeding. Hepatic encephalopathy was treated similarly in all patients. Correction of precipitating factors consisted of the administration of broad-spectrum antibiotics when infection was suspected or conrmed and intravenous administration of saline infusions when signs of dehydration were present. Initially, an enema of lactulose was administered, irrespective of the precipitating factor, and was followed by treatment with neomycin (3 g/day during the rst 3 days, followed by 1 g/day thereafter). We chose treatment with neomycin instead of lactulose because we were concerned about the difculties in differentiating the effects of lactulose from the possible development of diarrhea complicating enteric nutrition. Furthermore, both drugs have been shown to be equally effective [12]. All patients received supportive therapy and special nursing care was provided for patients with depressed consciousness. Concomitant drugs were allowed as needed if they did not interfere with the assessment of mental status. For the duration of the study the patients did not receive any oral diet apart from the enteric formula.

2.3. Objectives
The objective of the study was to assess whether there would be signicant differences in the outcome of hepatic encephalopathy between both groups of therapy over a period of 14 days following inclusion into the study. Due to the lack of standardized measures to assess outcome, we decided to compare the degree of hepatic encephalopathy between each group at each day during the study period. The hypothesis of the study was that there would not be signicant differences in the outcome of hepatic encephalopathy, but those treated with a low-protein diet would experience a negative protein balance.

2.4. Outcomes 2.2. Interventions

All patients followed a systematic protocol in order to assess criteria of eligibility and search for precipitating factors. The protocol included the following assessment: prior history, standard blood and urinary tests, qualitative determination of benzodiacepines in blood, chest X-ray, paracentesis if ascites, blood cultures, urinary cultures, fecal exam in search for melenas and assessment of focal neurological signs. Those with advanced encephalopathy (Glasgow , 13) or with focal neurological signs underwent computed tomography of the brain. After the initial workup, those who fullled entry criteria were invited to participate in the study. Acute alcoholic hepatitis was excluded according to clinical data (heavy alcohol abuse, jaundice, leukocytosis, fever, tender hepatomegaly) [10]; liver biopsies were not performed. The interval of time between admission in the hospital and initiation of the enteric formula was in all cases less than 24 h (group A: 14.2 ^ 2.7 h; group B: 13.5 ^ 3.1 h). During this period of time the patients completed the clinical workup, a nasogastric tube was inserted to deliver the enteric diet, and therapeutic measures for hepatic encephalopathy were initiated. Patients were randomized to two groups of treatment: A) a low-protein group and B) a normal protein group. Enteric nutrition was delivered by a nasogastric tube for 2 weeks with the aid of a peristaltic pump that was adjusted to deliver the complete amount (< 1500 ml) for 24 h (from 4 pm to 4 pm). The enteric formula contained 30 kcal/kg/day and a 3:1 ratio of glycids to lipids. The low-protein group (group A) followed a progressive increase in the dose of protein in the diet, because this regimen is common in many centers, has been recommended by several experts [2,11] and is in Hepatic encephalopathy was assessed, quantifying the mental state with a scale that was an adaptation of the West Haven criteria, according to the recommendations of a consensus conference [13]. Due to the lack of a clear denition of the different stages, we decided to dene them as shown in Table 1. The same evaluator (F. S. or F. C.) assessed the same patient for the duration of the study at the same time every day (4 pm). We did not perform neuropsychological tests, such as Trail A or Symbol Digit, because to interpret them correctly the patient needs to be collaborative and show normal consciousness. Protein synthesis and degradation were evaluated with the glycine-N15 infusion method [14] at day 2 and at day 14. Essentially, patients received a 50-mg bolus dose of glycine-N15 followed by a continuous infusion (4.475 1027 mol/kg21/h21) for 24 h. Urinary samples were taken at baseline and after 12 h of infusion, and N15 urinary enrichment was determined with an isotope ratio mass spectrometer. Standard laboratory blood tests and anthropometric indexes were determined at day 1 and day 14.

2.5. Sample size

Due to the lack of prior studies, the sample size was calculated on the assumption that the detection of a clinically relevant effect would require the inclusion of 20 patients that survive the episode of hepatic encephalopathy. The inclusion of this number of patients in a randomized study demonstrated the superiority of lactulose over placebo [15]. Although the efcacy of lactulose has been questioned, a large clinical experience

40

rdoba et al. / Journal of Hepatology 41 (2004) 3843 J. Co treatment. The physicians in charge of the patients (J.C., F.C., F.S.) did not know treatment allocation. The pharmacist (P.S.), who was also responsible for the preparation of the enteric formula, was not informed about the evolution of patients. To favor blinding of assessment, the enteric products were prepared to have a similar appearance and were kept in dark, coded bags that were distributed by the pharmacy.

Table 1 Stages of hepatic encephalopathy (adapted from Ref. [13]) Stage 0 Neurological manifestations Alert and attentive (in time and space) without signs of encephalopathy (neither dysarthria, ataxia, apping tremor or obvious decrease in the speed of mental processing) Alert and attentive, but with at least one of the following signs: dysarthria, ataxia, apping tremor or obvious decrease in the speed of mental processing Awake but inattentive: disoriented, somnolent, easy to distract, unable to perform easy mental tests (addition, subtraction, remember a list of numbers). Patients speech is easy to understand Marked somnolence or psychomotor agitation. Speech is difcult to understand Coma. The patient does not speak and does not follow simple commands (such as raising an arm or opening the mouth)

2.7. Statistical methods

Although several variables followed a normal distribution, due to the small number of patients the results are expressed as median (rst quartile). The non-parametric Mann Whitney rank sum test was used for comparisons between both groups of treatment. Calculations were performed with the Sigma Stat v3.0 statistic software.

3 4

3. Results The enrolment period started on March 2001 and nished on November 2002. The ow of participants through each stage of the trial is shown in Fig. 1. During this period 62 patients were considered eligible; of them, 30 patients were randomized. The reasons for exclusion were: acute alcoholic hepatitis n 6, gastrointestinal bleeding n 10, hepatocellular carcinoma n 3, previous intake of benzodiacepines n 2, neurological comorbidities n 3, respiratory failure requiring ventilatory support n 2, refusal to participate n 2, and other n 4. The baseline demographic and clinical characteristics of patients included in the study or those that nished

supports its use. The analysis was intentionally directed to assess patients that survived to 14 days of therapy to avoid the bias that could be introduced by including patients with terminal liver failure.

2.6. Randomization
Allocation was performed following a randomization scheme controlled by the Hospital pharmacy. The allocation sequence was generated by a computer and kept concealed until the end of the study. Enrolment was nished after the inclusion of 20 patients who completed the 14 days of

Table 2 Clinical characteristics of patients included in the study Low protein n 10 Demographics Age (years) Gender (male/female) Etiology of cirrhosis Alcohol Hepatitis C Other Decompensation Prior encephalopathy Ascites at admission Precipitating factor Infection Disturbance of electrolytes Unknown Blood tests Prothrombin activity (%) Bilirubin (mg/dl) Albumin (mg/dl) Ammonia (mcM/l) Anthropometrical data Body weight (kg) TSF (mm) MAC (cm) MAMC (cm) Malnutrition Normal protein n 10

67 (55) 4/6 5 4 1 9 5 4 2 4 51 (41) 1.7 (1.2) 2.2 (1.9) 92 (66) 60 (58) 8 (7) 25 (23) 22 (21) 5

68 (58) 7/3 4 4 2 7 3 5 1 4 57 (46) 2.2 (1.5) 2.5 (2.4) 108 (76) 69 9 26 21 (61) (9) (24) (20) 8

Results in medians (rst quartile). There were non-statistically signicant differences between both groups. TSF: Triceps skin fold thickness; MAC: Midarm circumference; MAMC: Midarm muscle circumference. Malnutrition (severe): TSF, MAC or MAMC , 5th percentile.

 rdoba et al. / Journal of Hepatology 41 (2004) 3843 J. Co

41

At the end of the study, both groups showed similar values of plasma ammonia (group A: 59 (34), group B: 61 (46) mcm/l), prothrombin activity (group A: 51 (39), group B: 59 (47) %), bilirubin (group A: 2 (1.1), group B: 1.6 (0.9) mg/dl) and albumin (group A: 2 (1.9), group B: 1.6 (0.9) mg/dl).

4. Discussion The main nding of this study is that there is no major benet of limiting protein intake on the evolution of episodic hepatic encephalopathy, while the administration of a low-protein diet exacerbates protein breakdown. In the short term, protein restriction did not have major consequences on liver function or plasma ammonia, as indicated by the evolution of blood tests. However, favoring protein breakdown for more prolonged periods may have detrimental consequences on the nutritional status and may result in adverse consequences. Our results suggest that patients with episodic hepatic encephalopathy can safely receive a normal protein diet. The current recommendation advocates limiting protein restriction to short periods of time and early initiation of a normal protein diet [1,8]. We designed the study according to this recommendation. Patients in group A underwent protein restriction followed by progressive increments, while those in group B followed a normal protein diet from the beginning. The lack of difference in the evolution of encephalopathy between both groups indicates that tolerance to a normal protein diet is higher than previously thought [4]. In practice, most of the patients can be treated with intravenous solutions for the rst days. The majority would then be able to reinitiate oral intake. Our data support that at that time patients can receive a diet with a normal amount of protein. In case of difculties in regaining oral intake, naso-gastric feeding should then be considered. Ammonia has been classically recognized as the main culprit for the development of hepatic encephalopathy [16]. Protein restriction was promoted as a simple way to decrease the nitrogenous load, which could lead to lower values of plasma ammonia [17]. However, this approach does not take into consideration the effects of limiting protein intake on protein metabolism. Kinetic studies in stable cirrhotics show that skeletal muscle removes large amounts of ammonia from circulation [18]. In muscle, ammonia is trapped and released as the ammonia carrier glutamine. In splanchnic viscera, glutamine is deaminated, resulting in ammonia production in portal drained viscera. An increase in protein breakdown causes a release of amino acids from the skeletal muscle. In cirrhotics, this nitrogenous load could cause an increase in plasma ammonia due to a decrease in the capacity to synthesize urea [19]. Most stable cirrhotics exhibit increased protein requirements that appear to be secondary to increased protein breakdown [20,21] and may explain why they are prone to

Fig. 1. Flow-chart of participants through each stage of the trial. Reasons for not meeting inclusion criteria are reported in the section of results.

treatment (Table 2) were similar in both groups. All patients received the allocated intervention. During follow-up 10 patients nished treatment before the end of the study. The reasons for discontinuation were equally distributed among both groups of treatment and corresponded to: death n 8, gastrointestinal bleeding n 1 or voluntary abandon n 1. None of the patients developed aspiration pneumonia. Some patients required repeated introduction of nasogastric tubes due to unconscious removal; none of them developed signicant gastric retention. The administration of the enteric feeding along the 14-day period of treatment was over 80% of the prescribed dose in all the cases. Twenty patients, 10 in each group, completed the 14 days of therapy. All patients nishing the study were discharged alive from the hospital. There were no signicant differences in the course of hepatic encephalopathy between the two groups of treatment, neither among all patients enrolled in the study (Fig. 2) nor those who nished the 14 days of treatment (Fig. 3). The analysis of protein metabolism was limited to those patients completing the study. During the second day on enteric nutrition, protein breakdown was exacerbated in the low protein group (Fig. 4; group A: 4.1 (3.6), group B: 2.5 (2.4) g protein kg/day; p 0:04), but without differences in protein synthesis (group A: 3.5 (3.1), group B: 3.1 (3.0) g protein kg/day). However, at the end of the study, when both groups received the same amount of protein, there were no differences either in protein synthesis (group A: 4.8 (3.4), group B: 3.9 (3.2) g protein kg/day) or breakdown (group A: 4.8 (3.1), group B: 3.7 (3.1) g protein kg/day).

42

rdoba et al. / Journal of Hepatology 41 (2004) 3843 J. Co

Fig. 4. Protein metabolism (medians) estimated with the glycine-N15 method in patients who nished the study and followed the low-protein diet (group A) or the normal protein diet (group B). Protein breakdown at day 2 was higher in the low protein diet group p 5 0:04. There were no statistically signicant differences between baseline and nal results in either groups. The gure shows the normal range observed in a previous group of healthy individuals [14] as a reference.

Fig. 2. Evolution of hepatic encephalopathy in all patients randomized to follow the low-protein diet (upper panel) or the normal protein diet (lower panel). The asterisks identify those patients who died during the study.

Fig. 3. Stage of hepatic encephalopathy (box plot: median, 10th90th percentile, 25th75th percentile, dashed line: mean) at inclusion (day 0), day 7 and end of the study (day 14) in the patients that nished the study (per-protocol analysis), grouped according to treatment. There were no statistical differences between the low-protein diet (white boxes) and the normal protein diet (gray boxes).

rapid protein wasting in periods of metabolic stress, fasting or reduced caloric intake. Whole-body protein turnover has not been studied before in cirrhotic patients with encephalopathy. In spite of limitations of the method [22], our data are in accordance with results from prior studies [20,21]. We observed high values of protein synthesis in patients with hepatic encephalopathy. Those fed a low-protein diet exhibited higher protein breakdown. The differences in protein breakdown can be explained by an increase in substrate requirements to maintain critical metabolic processes such as protein synthesis. Those differences disappeared at the end of the study, when the same group of patients received a diet with normal protein content (1.2 g/kg/d). The main drawback of this study is small sample size; hence, a larger study is required to validate our ndings. However, this kind of study is difcult to conduct, as shown by the lack of similar studies in the past. In spite of sample size limitations, our data are more solid than those provided by uncontrolled observations. Our results are in accordance with the current view that maintaining an adequate nutritional status is benecial for patients with cirrhosis [23,24]. This appears to be especially critical for those who are to undergo liver transplantation [25]. Furthermore, avoiding protein breakdown may decrease the production of ammonia derived from amino acid release from skeletal muscle. Thus, administering enough protein to attain protein requirements could improve hepatic encephalopathy. In conclusion, the rst controlled study randomizing cirrhotic patients with hepatic encephalopathy to receive different amounts of protein suggests that low-protein diets should be abandoned. Contrary to current recommendations [1,8], many centers still prescribe low-protein diets to cirrhotic patients who have had prior episodes of encephalopathy [26], a

 rdoba et al. / Journal of Hepatology 41 (2004) 3843 J. Co

43

practice that has been severely criticized [27]. The present study may help to implement these recommendations.
[14]

Acknowledgements Study supported by FIS 99/809 and Instituto de Salud Carlos III (projects G03/155 and C03/02). Dr. Francisco Castro was supported by an educational grant of Instituto Danone.
[15]

[16] [17]

References
[1] Blei AT, Cordoba J. Hepatic encephalopathy. Am J Gastroenterol 2001;96:1968 1976. [2] Riordan SM, Williams R. Treatment of hepatic encephalopathy. N Engl J Med 1997;337:473 479. [3] Plauth M, Merli M, Kondrup J. Management of hepatic encephalopathy. N Eng J Med 1997;337:19211922. [4] Mullen KD, Weber Jr FL. Role of nutrition in hepatic encephalopathy. Semin Liver Dis 1991;11:292 304. [5] Kondrup J, Mueller MJ. Energy and protein requirements of patients with chronic liver disease. J Hepatol 1997;27:239247. [6] Merli M, Riggio O, Dally L. PINC. Does malnutrition affect survival in cirrhosis? Hepatology 1996;23:10411046. [7] Kearns PJ, Young H, Garcia G, Blaschke T, OHanlon G, Rinki M, et al. Accelerated improvement of alcoholic liver disease with enteral nutrition. Gastroenterology 1992;102:200205. [8] Plauth M, Merli M, Weimann A, Ferenci P, Mueller MJ. ESPEN guidelines for nutrition in liver disease and transplantation. Clin Nutr 1997;16:4355. [9] Cabre E, Gonzalez-Huix F, Abad-Lacruz A, Esteve M, Acero D, Fernandez-Banares F, et al. Effect of a total enteral nutrition on the short-term outcome of severly malnourished cirrhotics. Gastroenterology 1990;98:715 720. [10] Akriviadis E, Botla R, Briggs W, Han S, Reynolds T, Shakil O. Pentoxifylline improves short-term survival in severe acute alcoholic hepatitis: a double blind, placebo-controlled trial. Gastroenterology 2000;119:16371648. [11] Anonymous, Hepatic encephalopathy. In: Sherlock S, Dooley J, editors. Diseases of the liver and the biliary system, 9th ed. Oxford: Blackwell Scientic; 1993. p. 86101. [12] Atterbury CE, Maddrey WC, Conn HO. Neomycinsorbitol and lactulose in the treatment of acute portal-systemic encephalopathy. Am J Dig Dis 1978;23:398406. [13] Ferenci P, Lockwood AH, Mullen KD, Tarter RE, Weissenborn K, [18]

[19]

[20]

[21]

[22]

[23] [24]

[25]

[26]

[27]

Blei AT, et al. Hepatic encephalopathydenition, nomenclature, diagnosis, and quantication: nal report of the working party at the 11th World Congress of Gastroenterology, Vienna, 1998. Hepatology 2002;35:716 721. Lopez-Hellin J, Lopez-Lara M, Mercader S, Gemar E, Garcia-Arumi E, Sabin P, et al. Early curbing of protein hypercatabolism in postoperative patients by nutritional support with glucose plus amino acids, but not with glucose alone. Clin Nutr 1997;16:6773. Simmons F, Goldstein H, Boyle JD. A controlled clinical trial of lactulose in hepatic encephalopathy. Gastroenterology 1970;59: 827832. Butterworth RF. The neurobiology of hepatic encephalopathy. Semin Liver Dis 1996;16:235 244. Phillips GB, Schwartz R, Gabduzda GJ, Davidson CS. The syndrome of impending hepatic coma in patients with cirrhosis of the liver given certain nitrogenous substances. N Eng J Med 1952; 247:239 246. Damink SWMO, Jalan R, Redhead DN, Hayes PC, Deutz NE, Soeters PB. Interorgan ammonia and amino acid metabolism in metabolically stable patients with cirrhosis and a TIPSS. Hepatology 2002;36: 1163 1171. Hamberg O, Nielsen K, Vilstrup H. Effects of an increase in protein intake on hepatic efcacy for urea synthesis in healthy subjects and in patients with cirrhosis. J Hepatol 1992;14:237243. Kondrup J, Nielsen K, Juul A. Effect of long-term refeeding on protein metabolism in patients with cirrhosis of the liver. Br J Nutr 1997;77: 197212. Swart GR, van den Berg JWO, van Vuure JK, Rietveld T, Wattimena DL, Frenkel M. Minimum protein requirements in liver cirrhosis determined by nitrogen balance measurements at three levels of protein intake. Clin Nutr 1989;8:329336. Deutz NE, Wagenmakers AJM, Soeters PB. Discrepancy between muscle and whole body protein turnover. Curr Opin Clin Nutr Metab Care 1999;2:29 32. McCullough AJ. Malnutrition in liver disease. Liver Transplant 2000; 6:S85S96. Marchesini G, Bianchi G, Merli M, Amodio P, Panella C, Loguercio C, et al. Nutritional supplementation with branched-chain amino acids in advanced cirrhosis: a double-blind randomized trial. Gastroenterology 2003;124:17921801. Selberg O, Bottcher D, Tusch G, Pichlmayr R, Henkel E, Muller MJ. Identication of high- and low-risk patients before liver transplantation: a prospective cohort study of nutritional and metabolic parameters in 150 patients. Hepatology 1997;25:652657. Soulsby CT, Morgan MY. Dietary management of hepatic encephalopathy in cirrhotic patients: survey of current practice in United Kingdom. Br Med J. 1999;318:1391. Seymour CA, Whelan K. Dietary management of hepatic encephalopathy. Too many myths persist. Br Med J 1999;318:13641365.