The Role of Nutritional Therapy in Alcoholic Liver Disease

Christopher M. Griffith, M.D., and Steven Schenker, M.D.

Alcoholic liver disease (ALD) evolves through various stages, and malnutrition correlates with the severity of ALD. Poor nutrition is caused both by the substitution of calories from alcohol for calories from food and by the malabsorption and maldigestion of various nutrients attributed to ALD. The only established therapy for ALD consists of abstinence from alcohol. Sufficient nutritional repletion coupled with appropriate supportive treatment modalities may be effective in reducing complications associated with ALD—particularly infection. Nutrition makes a significant positive contribution in the treatment of ALD, especially in selected malnourished patients. KEY WORDS: Ethanol metabolism; heavy alcohol use; alcoholic liver disease; alcoholic fatty liver; alcoholic hepatitis; fibrosis; alcoholic cirrhosis; gut; infection; anorexia; hepatic encephalopathy; nutrition; malnutrition; nutritional therapy; S-Adenosylmethionine (SAM); antioxidants; milk thistle; silymaryin; phosphatidylcholine; dietary fat; vitamins

he study of malnutrition in alcoholic hepatitis) and only 20 percent Diversity of Liver Injury patients with alcoholic liver dis­ will progress to scarring of the liver (i.e., Tease (ALD) is based on several cirrhosis) (McCullough and O’Connor Alcoholic liver injury is known to general concepts and observations. 1998). Clearly, other risk factors, evolve through various stages. Patients Researchers and clinicians previously including genetic predisposition, obesity, exhibit a variety of clinical symptoms believed that malnutrition was the concomitant viral hepatitis infection, and signs in liver histology (as reviewed primary cause of liver injury in ALD and poor nutrition, may contribute in Dasarathy and McCullough 2003). rather than the consequence of excessive variably to the development of ALD. As previously mentioned, almost every­ alcohol consumption. This view was Indeed, in a large study of hospital­ one with heavy alcohol consumption based on the prevalence of malnutrition ized patients with varying severity of develops fatty liver. This often is a rather benign disorder and considered in alcoholics and those with clinical ALD, malnutrition (especially the type reversible upon cessation of alcohol evidence of liver (i.e., hepatic) dysfunction caused by deficient protein and calories) resulting from alcohol consumption intake. In some cases, continued drink­ was closely associated (although not ing may result in the development of (Mendenhall et al. 1984). It now is necessarily causal) with the severity of widely accepted that the quantity and alcoholic hepatitis, which may—and liver injury (Mendenhall et al. 1984). duration of alcohol consumption are the often does—end in severe clinical dis­ All patients with clinical evidence of principal agents in the development of ease. The severity of disease and liver alcoholic liver injury. This is based on ALD (regardless of severity) exhibited dysfunction correlates with an increas­ animal and human data showing that some features of malnutrition. With ing short-term mortality (as reviewed ALD can develop in well-nourished regard to the possible value of nutritional in Dasarathy and McCullough 2003). individuals who consume large amounts therapy, it would seem logical that It is in this acutely diseased group that of alcohol (Mezey 1991). However, a patients with more severe deficits would optimal nutritional therapy might have great deal of variability exists regarding benefit more, although convincing proof the most impact. Continued drinking the individual development of progres­ of this, to our knowledge, is not available. in this group of patients may lead to sive alcoholic liver injury. Although This article reviews the various forms more than 90 percent of people with of liver injury; the basis for and role of CHRISTOPHER M. GRIFFITH, M.D., is a excessive alcohol consumption will malnutrition in ALD, including the fellow of Gastroenterology and STEVEN develop fatty liver (defined as greater harmful effects of the products of alcohol SCHENKER, M.D., is professor of than 5 percent fat in the liver), only up metabolism; evidence for the benefits Medicine and Pharmacology, both at the to 35 percent will develop inflammation of nutrition; and special considerations University of Texas Health Science Center of the liver caused by alcohol (i.e., for nutritional therapy in ALD. at San Antonio, San Antonio, Texas.

296 Alcohol Research & Health Nutritional Therapy in ALD

the development of excess scar tissue in Decreased Food Intake decreased intestinal enzyme activity the liver (i.e., fibrosis) and the subse­ (i.e., lactase) required for carbohydrate People with ALD will substitute calories quent anatomical changes of cirrhosis. digestion and absorption, and/or from food with calories from alcohol. It Cirrhosis is considered a late stage of decreased absorption from increased has been shown in patients with ALD swelling of the gut. The latter is attrib­ the disease, clinically manifested by that calories from alcohol may con­ progressive liver dysfunction with asso­ uted to increased pressure in the drain­ tribute more than 50 percent of their ing vein (i.e., portal vein) and lym­ ciated yellowing of skin and whites total calories, with calories from pro­ of the eyes (i.e., jaundice)—caused by phatic vessels connecting the intestines tein comprising only 6 to 10 percent with the liver. The changes in intestinal decreased liver clearance of bilirubin, (Mendenhall et al. 1984). In addition, fluid accumulation in the abdomen digestion and absorption appear to the proportion of calories from alcohol reverse once the patient ceases drinking (i.e., ascites), and impaired brain func­ appears to increase, whereas those from tion caused by the accumulation of and starts to follow a normal diet, sug­ food decrease with escalating liver dys­ gesting that nutritional replacement ammonia in the brain tissues (i.e., function (Mezey 1991). This increase encephalopathy) (as reviewed in may be of special benefit (Mezey 1991). in alcohol calories and decrease in food Finally, the preferential metabolism Dasarathy and McCullough 2003). calories may be partially explained by These three disorders may occur of alcohol by the liver alters the meta­ the diversion of funds from the pur­ bolism of sugars (i.e., carbohydrates), separately or often in association with chase of food to that of alcohol; how­ each other. In many instances, these fats (i.e., lipids), and proteins (i.e., ever, hospitalized patients with ALD amino acids/nitrogen) (Mezey 1991). stages of liver injury may be difficult to given adequate access to nutrition still distinguish either clinically or by labo­ Abnormal breakdown of fat results in demonstrate decreased ingestion of the formation of triglycerides that are ratory measures of liver dysfunction (as nonalcohol calories (Schenker and reviewed in Dasarathy and McCullough deposited in the liver, manifesting as Halff 1993). This decreased desire for fatty liver (Schenker and Halff 1993). 2003). Prior studies regarding nutri­ food (i.e., anorexia) also correlates with tional therapy in ALD often did not Altered fat metabolism also may propa­ the severity of liver injury (Hirsh et al. gate fibrosis by increasing collagen for­ differentiate between these various types 1999; Mendenhall et al. 1995). Anorexia of liver disease, complicating researchers’ mation (Schenker and Halff 1993). is pervasive in ALD and is a key reason The altered functional mass of the liver abilities to assess the true benefit of for decreased dietary intake of nonalcohol from these increased deposits of fat and nutritional therapy. This has been a prob­ calories (Mezey 1991). collagen may result in decreased stores lem in the assessment of such data in the of vitamins and carbohydrates. Glycogen, past as well as for the present authors. Poor Absorption and Digestion of the storage form of carbohydrates in Nutrients the liver, serves as an energy reserve for periods of increased energy need. Basis for Malnutrition in Another factor contributing to poor Inadequate glycogen reserves cause the ALD nutrition in patients with ALD is the body to make use of other metabolic malabsorption and maldigestion of var­ pathways for energy, such as the break­ The signs and symptoms of nutritional ious nutrients from the gut (Mezey down of muscle (Schenker and Halff deficits in ALD patients have been well 1991; Schenker and Halff 1993). This 1993). This may explain the increased characterized (Mendenhall et al. 1984; may relate to the impaired output of muscle wasting, increased nitrogen Mezey 1991; Schenker and Halff 1993; bile from the liver, resulting in decreased excretion in the stool, and negative Nompleggi and Bonkovsky 1994; absorption of fat and fat-soluble vita­ nitrogen balance seen in patients with Hirsh et al. 1999) and include muscle mins. The possibility of concomitant ALD. It is important to note that in wasting, decreased lean body mass, var­ pancreatitis causing decreased output healthy individuals these alternate ious vitamin deficiencies, and decreased of enzymes necessary for absorption of pathways of energy usually are found measurable serum proteins. A complete fats and proteins also can occur with only after periods of prolonged fasting description of specific nutritional deficits alcohol abuse. Moreover, there may or starvation. Patients with ALD, how­ is beyond the scope of this review; be a direct effect of alcohol on the gut ever, may begin to use alternative path­ however, it is important to consider the itself. Alcohol has been demonstrated ways after only an overnight fast, sug­ factors that contribute to malnutrition to decrease the intestinal absorption gesting that the frequency of feeding is in individuals with ALD, as they may of amino acids and various vitamins, as important as the type of feeding have an influence on the administration particularly thiamine, folate, and B12 (McCullough and O’Connor 1998). of nutritional therapy. There are many (Schenker and Halff 1993). Malabsorp­ The basis for nutritional deficits is reasons for the deficits and abnormalities tion also may occur through mechani­ summarized in Table 1. This subject that occur as a result of malnutrition. cal alterations in the gut. This may be also has been extensively discussed in These are outlined below and include attributed to increased intestinal another review (Schenker and Halff decreased dietary intake and poor swelling (i.e., edema) from a lack of 1993), which can be referenced for absorption and digestion of nutrients. structural proteins in the gut wall, more details.

Vol. 29, No. 4, 2006 297 Role of Malnutrition in tion of various signaling chemicals (i.e., mins (e.g., vitamin E) that may offer a ALD cytokines), which may further con­ protective effect as antioxidants tribute to tissue injury (as reviewed in (Schenker and Halff 1993). It is not precisely known how alcohol Dasarathy and McCullough 2003). causes liver damage. The net effect of The liver has a built-in defense Increased Risk of Infection nutrition on the development of ALD against harmful oxidation in gluta­ may involve multiple factors, including thione, a compound that assists in the Patients with ALD are at increased risk free-radical damage and increased risk removal of the toxic byproducts of of infection (Schenker and Halff 1993; of infection. alcohol metabolism. Glutathione avail­ Hirsh et al. 1999) partly because alco­ ability depends on the presence of cer­ hol directly suppresses the immune sys­ Free-Radical Damage tain amino acids that may be deficient tem but also because the altered pro­ in patients with ALD (Schenker and tein metabolism observed in ALD It has been observed that one of the Halff 1993). One of these amino acids, results in decreased circulating antibod­ toxic byproducts of alcohol metabolism S-adenosylmethionine (SAM), serves as ies needed to fight infection. The (i.e., free radicals) may cause damage to a precursor for glutathione. Evidence intestinal system also works as a barrier the liver. Free-radical damage also exists that SAM is deficient in patients to prevent bacteria from inside the gut occurs as a result of oxidation of lipids with ALD (Schenker and Halff 1993). from crossing the intestinal wall and in cellular membranes and in the inter­ Furthermore, membrane integrity causing infection. In addition, the cells nal constituents (i.e., mitochondria) of depends on the availability of SAM, in of the gut secrete an antibody that is the cell. This improper oxidation of fat addition to an ample supply of phos­ unique to the gut and assists with can, in turn, lead to the increased fat pholipids. SAM is involved in the pro­ fighting infection (Schenker and Halff deposition and fibrosis, described pre­ cessing of phospholipids needed for cell 1993). There is evidence that the gut’s viously. Cell damage from this membrane repair (Schenker and Halff role as a barrier to infection decreases improper oxidation also results in an 1993). It also has been observed that with poor nutrition, as well as with inflammatory response and the genera- patients with ALD are deficient in vita­ excess alcohol intake (Schenker and Halff 1993). Research in animals has shown that improved nutrition results Table 1 Basis for Nutritional Deficits in Alcoholic Liver Disease (ALD) in decreased translocation of bacterial organisms across the gut and a subse­ Decreased caloric intake Results quent decrease in bacterial infections (Casafont et al. 1997). Generally, it is • Anorexia • Decreased calories, vitamins, and accepted that patients with advanced • Decreased ingestion of nutrients available for utilization ALD have an increased risk of morbidity non-alcohol calories and mortality with surgical procedures. • Increases with severity of ALD This increased risk may be related to higher infection risk and poor wound Decreased intestinal absorption/ healing (Schenker and Halff 1993). digestion of nutrients Considering that chronic ALD, after • Decreased bile excretion • Decreased fat digestion and cessation of drinking, is one of the absorption of fat-soluble vitamins more common indications for liver • Decreased pancreatic function • Decreased pancreatic enzymes transplant (McCullough and O’Connor for fat and protein digestion 1998), it is reasonable to suggest that • Altered intestinal integrity • Decreased amino acid and improved nutrition in patients with vitamin absorption and digestion • Decreased intestinal enzymes • Decreased carbohydrate digestion ALD can improve the outcomes of sur­ gical procedures by decreasing infection Decreased processing and storage and improving wound healing. of nutrients In brief summary, the decreased ability to process hepatic fat, as well as • Preferential metabolism of alcohol • Abnormal processing of fats and sugars the lack of key proteins and amino • Abnormal oxidation of fat • Fatty liver and increased acids that may decrease the liver’s abil­ collagen production ity to neutralize the effects of free radi­ • Decreased functional liver mass • Decreased energy stores and utilization of alternative pathways normally cals generated by alcohol metabolism, reserved for fasting (abnormal muscle may, in turn, result in damage to cell breakdown and abnormal oxidation of fats) membranes and promote inflammation and cell death (i.e., necrosis). Such SOURCE: Modified with permission from Mezey 1991 and Schenker and Halff 1993. events also may lead to fibrosis and even cirrhosis. Thus, theoretically,

298 Alcohol Research & Health Nutritional Therapy in ALD

improved nutrition could ameliorate feeding tube, or intravenous line) as well various substances. Kearns and colleagues these adverse events, enhance hepatic as compliance and length of treatment. (1992) demonstrated improved clear­ regeneration, and decrease the risk of Patient compliance in the studies ance of antipyrine,1 whereas Bonkovsky infection, which is a common compli­ cited generally is not estimated and is and colleagues (1991a,b) showed no cation of advanced ALD and a leading therefore difficult to assess. Compliance change in antipyrine clearance but did cause of patient mortality. is clearly likely to be better in patients report improved results from the galac­ given food intravenously or via a feed­ tose elimination test2 with improved ing tube. It also may be better in those Evidence for Benefits of with less severe ALD because they have nutrition. Only two studies (Diehl et Nutrition a lesser degree of anorexia. al. 1985; Achord 1987) evaluated the role of nutritional therapy on histology of the liver in ALD. There was a As shown in Table 2, there have been Research Findings about 15 studies (14 controlled and 1 decrease in fat in one study (Diehl et al. pilot study) evaluating the benefits of Even with these stipulations, a number 1985) and fibrosis in the other (Achord nutritional therapy in ALD. These of conclusions emerge from the studies 1987) but no improvement in inflam­ studies have assessed the possible benefit reviewed. In most studies, nutritional mation or necrosis (Diehl et al. 1985). of optimal nutrition in a variety of ways. status appears to have been improved They have evaluated the effects on in patients given proper dietary intake, nutritional status, liver tests, liver histol­ especially with regard to nitrogen bal­ Effect on Mortality ogy, and, most importantly, mortality. ance. The best evidence of this was in Despite improvement in nutritional the studies by Mezey and colleagues parameters and tests of liver “function” Variations in Studies (1991) and Bonkovsky and colleagues (1991a,b). One study (Naveau et al. in most studies, the majority did not The studies reviewed here have wide 1986) found no improvement in nutri­ demonstrate a change in mortality. variations, which make meaningful tional parameters. A better correlation Two studies did report improved sur­ comparison difficult. For instance, there with improvement in nutritional status vival with nutritional therapy. Cabre is a disparity in the severity of liver dis­ was seen in those studies with more and colleagues (1990) studied patients ease among study participants. Two stud­ specific parameters of nutrition, such as with cirrhosis, the majority of whom ies (Diehl et al. 1985; Nasrallah and visceral proteins (which are reactive to had alcoholic cirrhosis, and demon­ Galambos 1980) had no deaths and one physiological changes in the body and strated a decreased mortality rate during had only 5 percent mortality (Naveau therefore are valuable in identifying hospitalization with nutritional supple­ et al. 1986). When considering that malnutrition), nitrogen balance, triceps mentation through a feeding tube. mortality ranged from 12 to 47 percent skinfold thickness, and creatinine– Nasrallah and Galambos (1980) were in the other studies, this clearly suggests height index (a ratio of a patient’s 24­ able to show improved survival in patients a less severe form of disease in the for­ hour excretion of the waste product mer groups. Additionally, the studies creatinine [which is related to muscle with alcoholic hepatitis with adminis­ have variable patient populations with mass] and the expected normal creati­ tration of intravenous amino acids. regard to stage of ALD. Four studies nine excretion) (Mezey et al. 1991; Although a consistent improvement (Marchesini et al. 1990; Kearns et al. Bonkovsky et al. 1991a,b). in survival has not been demonstrated 1992; Cabre et al. 1990; Hirsch et al. in individual studies, the average com­ 1993) primarily addressed alcoholic Liver Tests posite mortality in 10 studies (Diehl et cirrhosis, whereas 11 (Diehl et al. 1985; al. 1985; Nasrallah and Galambos 1980; Nasrallah and Galambos 1980; Naveau The bulk of the data available suggests Naveau et al. 1986; Calvey et al. 1985; et al. 1986; Calvey et al. 1985; Cabre et that improved nutrition is reflected in Mezey et al. 1991; Achord 1987; al. 2000; Mezey et al. 1991; Achord 1987; specific improvements in liver tests such Mendenhall et al. 1985; Bonkovsky et Mendenhall et al. 1985; Bonkovsky et as serum albumin (a protein found in al. 1991a,b; Simon and Galambos al. 1991a; Bonkovsky et al. 1991b; blood plasma), bilirubin (produced by Simon and Galambos 1988) focused the breakdown of hemoglobin and 1988) involving amino acid therapy in on patients with alcoholic hepatitis. processed in the liver), transaminases alcoholic hepatitis was 9.6 percent in However, a large number of patients in (enzymes involved in breaking down those treated compared with 17 percent the studies involving alcoholic hepatitis amino acids), and blood clotting (i.e., in control subjects, suggesting a trend also had underlying cirrhosis. Moreover, coagulation). A few studies (Kearns et toward significantly improved mortal­ many of these studies utilized only a al. 1992; Mezey et al. 1991; Bonkovsky ity (McCullough and O’Connor 1998). small number of patients, making mean­ et al. 1991a,b) even demonstrated ingful statistical analysis difficult. The enhanced liver status as measured by 1 studies also differed in methods of more direct tests of liver function that Antipyrine metabolism is used as an index of liver function. nutrient administration (e.g., oral, involve hepatic uptake and clearance of 2 The galactose elimination test measures liver function.

Vol. 29, No. 4, 2006 299 Table 2 Studies of Nutrition and Alcoholic Liver Disease (ALD)

Refer– ALD # Route Time Type of diet Liver Liver his- Nutritional Mort- Comments ence stage Pts (days) tests tology parameters ality

Cabre AH 71 Enteral 28 Control: 2,000 kcal/ Improved Not Improved No Followed for 1 et al. (80% vs. 1g protein/kg/day in both assessed albumin in over- year; deaths ear­ 2000 AC) steroids oral diet + 40 mg/day groups both; no all lier in NGT group prednisolone (actual change in effect (7 vs. 23 days); intake not stated) anthro- deaths in steroid Treatment: 2,000 pometrics group later; kcal/72 g protein deaths in steroid 31% as BCAA via group from infec- NGT tion

Menden- AH 57 Oral 30 Historic controls: No dif- Not Improved No Study not hal and 2,500 kcal/day oral ference assessed in treated effect randomized; et al. enteral diet (actual ~2,300 group improved nutri­ 1985 kcal/day) tion and nutrition- Treatment: 2,500 al parameters; kcal oral diet + 2,200 mortality 17 to kcal/day BCAA sup- 21% plement

Calvey AH 64 Oral 21 Control: 1,800 to No dif- Not Improved No 23% with HE not et al. (some 2,400 kcal/70 to 100 ference assessed but no effect affected by diet; 1985 IV) g protein oral diet difference improved cuta- Treatment: control between neous energy; diet + 2,000 kcal groups improved nitro- supplement with 65 g gen balance; 32 conventional protein to 43% mortality or 65 g protein as BCAA

Nasral- AH 35 Oral 28 Control: 3,000 kcal/ Improved Not Not Improved Significant improve­ lah and and IV 100 g protein oral in treated assessed assessed; (p <0.02) ment in liver tests; Galam- diet (actual intake group improved improved nitro­ bos ~1,400 kcal/day) nitrogen gen balance; 1980 Treatment: control balance 22% mortality in diet + 70 to 85 g AA control subjects; IV (actual oral intake small patient 1,500 kcal) number

Diehl AH 15 Oral 30 Control: unrestrict- No dif- Decreased Improved N/A Improved histology et al. and IV ed diet (~3,000 ference fat but not in treated and nutrition; no 1985 kcal/day actual inflamma- group mortality; patients intake) tion or less severely ill; Treatment: control necrosis in small patient diet + 53 g AA + 130 g treated number glucose IV group

Achord AH 28 Oral 21 Control: 2,700 Improved Decreased Not No Low oral intake in 1987 and IV kcal/day oral diet in treated hyaline in assessed effect both; improved (actual intake group treated functional liver ~1,200 kcal/day) (galac- group mass; improved Treatment: control tose elim- liver tests; 7 to diet + 42.5 g AA/day ination 21% mortality; IV (actual oral ~1,200 improved) small patient kcal) number not assessed

AA = amino acids; AC = alcoholic cirrhosis; AH = alcoholic hepatitis; BCAA = branched chain amino acids; FT = feeding tube; HE = hepatic encephalopathy; IBW = ideal body weight; IV = intravenous; NGT = nasogastric tube; for comparison * = no control for comparison.

300 Alcohol Research & Health Nutritional Therapy in ALD

Table 2 Studies of Nutrition and Alcoholic Liver Disease (ALD) continued

Refer– ALD # Route Time Type of diet Liver Liver Nutritional Mort­ Comments ence stage Pts (days) tests histology parameters ality

Simon AH 34 Oral 28 Control: 2,400 No dif­ Not Not No Improved liver and and IV kcal/100 g protein ference assessed assessed effect tests in severe Galam­ oral diet + 3 cans in mod­ AH; stratification bos Ensure (actual erate but into moderate 1988 intake not recorded) improved and severe AH Treatment: control in severe decreased num­ diet + 70 g AA, 100 hepatitis ber for analysis g glucose, 50 g lipid in treat­ per day IV ed group

Mezey AH 54 Oral 30 Control: 2,000 kcal Improved Not Improved No Improvement in et al. and IV orally + 130 g glu­ in treat­ assessed in treated effect liver tests and 1991 cose IV (actual oral ed group group nutritional param­ intake ~1,400 kcal; (galac­ eters; 1 month total ~1,950) tose survival 19% con­ Treatment: control elimina­ trols and 21% diet + 52 g AA/130 g tion) treated; 2-year glucose IV; (oral survival 40% in intake ~1,300 kcal; both total ~2,300 kcal)

Bonkov­ AH 39 Oral 21 Control: oral 30 kcal Trend Not Improved No No mortality dur­ sky and IV and 1 g protein/kg toward assessed; in treated effect ing study; et al. IBW/day (actual intake improve­ decreased group improved nutri­ 1991 ~1,800 kcal/day) ment in liver vol­ tion parameters; Treatment: control treated ume in trend toward diet + 70 g AA and group treated improvement in 100 g glucose IV/day group liver tests (~2,300 kcal oral + IV daily)

Naveau AC 40 Oral 28 Control: 40 kcal/kg Improved Not No signifi­ No Low difference in et al. (23% and IV and 200 mg nitro- in treated assessed cant effect oral intake in treat­ 1986 AH) gen/kg/day orally group improve­ ment vs. control; (actual intake ment in only 5% total mor­ ~2,100 kcal/day) treated tality; no improve­ Treatment: control group ment in nutrition diet as IV + offered parameters; some oral diet; (actual oral improvement in intake ~1,100 kcal) liver tests

Cabre AC 35 Oral 23 to Control: oral 2,200 Improved Not Improved Impro­ Improved albumin et al. vs. 25 kcal and 70 to 80 g in treat­ assessed albumin in ved and Child’s score in 1990 enteral protein/day; (40 g ed group treated treatment; 12% protein/day if HE); (improv­ group; no mortality in treated (actual intake ed Child’s difference vs. 47% in controls; ~1,320 kcal) score) in anthro­ no overall improve­ Treatment: 2,115 pometrics ment in nutrition kcal and 71 g protein parameters as BCAA via FT

AA = amino acids; AC = alcoholic cirrhosis; AH = alcoholic hepatitis; BCAA = branched chain amino acids; FT = feeding tube; HE = hepatic encephalopathy; IBW = ideal body weight; IV = intravenous; NGT = nasogastric tube; for comparison * = no control for comparison.

Vol. 29, No. 4, 2006 301 Effect on ALD noted decreased hospitalization at 1 with alcoholic hepatitis were treated with both steroids and nutritional therapy, Other recent evidence suggests that nutri­ year in patients with alcoholic cirrhosis resulted in a mortality of 15 percent tional therapy may play a role in treat­ given supplemental nutrition. This at 1 year. This was a lower rate than ment for ALD. Cabre and colleagues decline also was attributed to a decreased expected (25 percent) in this group (2000) compared the use of steroids rate of infection. A follow-up, uncon­ historically from treatment with either with nutritional therapy in severe alco­ trolled study by the same authors modality alone. Controlled data are holic hepatitis and noted an improved demonstrated that improved nutritional needed to verify this finding. mortality in the steroid group during status in patients with alcoholic cirrho­ the first week, whereas the nutrition sis resulted in improved cell-mediated group had improved mortality over the immunity, perhaps explaining the reduced Protein Tolerance. A review of the long term. The latter appeared to be incidence of infection (Hirsh et al. 1999). composite data also suggests that attributed to decreased risk of infec­ An uncontrolled study by Alvarez and patients with ALD are able to tolerate tion. Hirsch and colleagues (1993) also colleagues (2004), in which 13 patients proteins (i.e., amino acids/nitrogen)

Table 2 Studies of Nutrition and Alcoholic Liver Disease (ALD) continued

Refer– ALD # Route Time Type of Diet Liver Liver Nutritional Mort- Comments ence stage Pts (days) tests histology parameters ality

Kearns AC 31 Oral 28 Control: standard Improved Not Improved No Improved nitro­ et al. with oral diet (actual not in treated assessed in treated effect gen balance and 1992 enteral stated) group group liver tests in treat­ support Treatment: control ment; 13% mor­ diet + 1.5 g/kg/day tality in treated vs. supplement as 27% in control at casein; (reported 1 week, equal 200% increase kcal mortality at 8 intake vs. control) weeks

Mar­ AC 64 Oral 90 Control: 15 days 45 Improved Not Improved Not Greater improve­ chesni to 65 g protein then in treat­ assessed in both, asses­ ment in nitrogen et al. oral diet + 0.175 ed group but greater sed balance in treat­ 1990 g/kg protein casein in treated ment; improved supplement group HE in treatment; Treatment: 15 days crossover at 3 of 45 to 65 g protein months with then regular diet + improvement in HE 0.24 g/kg protein as in those changed BCAA supplement to treatment

Hirsch AC 51 Oral 12 Control: oral diet + Improved Not Improved No Trend toward et al. mo placebo (actual in both assessed in treated effect improved mortality 1993 ~1,580 kcal intake) group in treatment; Treatment: control improvement due diet + 1,000 kcal to decreased and 34 g protein/day infection orally (actual ~2,470 kcal intake)

Alvarez AH 13 Enteral 22 Control: none Improved Not Not 15%* Pilot study, not et al. + (+/­ Treatment: 2,000 assessed assessed randomized, mor­ 2004 steroids 3.8) kcal and 72 g pro­ tality less than tein orally/day + 40 expected from mg prednisolone either historic treatment alone

AA = amino acids; AC = alcoholic cirrhosis; AH = alcoholic hepatitis; BCAA = branched chain amino acids; FT = feeding tube; HE = hepatic encephalopathy; IBW = ideal body weight; IV = intravenous; NGT = nasogastric tube; for comparison * = no control for comparison.

302 Alcohol Research & Health Nutritional Therapy in ALD

well (i.e., no encephalopathy), whether SAM ment of ALD, there has been interest given orally or by feeding tube. The use in the efficacy of antioxidants in the S-Adenosylmethionine (SAM) is impor­ of intravenous lines to provide nutrition treatment of ALD. Metadoxine, an tant in the synthesis of proteins and increases the risk of infection but also is antioxidant with few known side effects, polyamines as well as of glutathione well tolerated. Somewhat surprisingly, has been approved for use in the treat­ (Schenker and Halff 1993). As men­ even patients with severe ALD tolerate ment of ALD in countries other than tioned above, glutathione plays a major generous protein supplements without the United States. A controlled trial from role in the removal of damaging free precipitation or worsening of hepatic Spain randomly assigned 136 alcoholics radicals. The enzyme that produces encephalopathy. Protein is of theoreti­ with fatty liver to receive either meta­ SAM (SAM synthetase) is deficient in doxine or a placebo. After 3 months, cal concern in patients with impaired patients with alcoholic cirrhosis, leading liver function because the intestinal both groups showed an overall improve­ to decreased levels of SAM (Mato et al. ment in markers of liver function; how­ breakdown of protein results in ammo­ 1999). Administration of SAM poten­ nia, which is normally processed by the ever, the metadoxine group demonstrated tially could benefit ALD patients by more improvement and a more rapid liver to urea. The exact mechanism by improving levels of glutathione and which excess ammonia acts as a neuro­ response to therapy (Caballeria et al. decreasing the damage to liver cells caused 1998). The metadoxine group also had toxin is not completely understood. It by alcohol. SAM is relatively benign with generally is accepted, however, to play a a more substantial decrease in the few side effects (Rambaldi and Gluud amount of fat in the liver. Improvements pivotal role in hepatic encephalopathy 2001). Despite one study demonstrating (Albrecht and Norenberg 2006). In were noted in those who continued improved survival with long-term admin­ drinking, although the recovery was brain cells, ammonia is converted into istration of SAM (Mato et al. 1999), a the presumably nontoxic amino acid better in those who were abstinent. Cochrane systematic review of eight Although the results are promising, it glutamine (Albrecht and Norenberg randomized controlled trials did not 2006). A recent study proposes that appears that further studies involving demonstrate a survival advantage more advanced ALD are needed before glutamine actually may be neurotoxic, (Rambaldi and Gluud 2001). Therefore, as surplus glutamine is converted back recommending this as therapy for ALD this agent should not be used for treat­ patients. to ammonia and also exerts an individ­ ment, pending further randomized con­ ual effect in the development of hepatic trolled trials. Of note, the critical Cochrane encephalopathy and its cerebral com­ review only considered one of the eight Milk Thistle (Silymaryin) plications (i.e., brain edema) (Albrecht studies evaluated to be an adequate one. Milk thistle is a popular dietary herbal and Norenberg 2006; Ferenci 1994). supplement used by many patients As a consequence of impaired liver Antioxidants with chronic liver disease. Milk thistle function, other avenues are needed to primarily is composed of silymaryin dispose of ammonia. One method is As increased free radical formation (70 to 80 percent), which is its active conversion of ammonia to glutamine appears to play a role in the develop­ ingredient (Ball and Kowdley 2005). by muscle. A recent study in rats with acute liver failure and resulting ammonia elevations demonstrated increased glu­ Table 3 Goals of Nutritional Supplementation in Chronic Liver Disease tamine production by skeletal muscle. This suggests that preservation of muscle • Prevent or correct protein–calorie malnutrition mass via better nutrition may be bene­ ficial in preventing or ameliorating • Prevent or correct hepatic encephalopathy hepatic encephalopathy and its ensuing complications in patients with ALD • Aid hepatic healing and regeneration insofar as possible (Chatauret et al. 2006). The best type of nutritional therapy remains uncer­ • Improve quality of life tain; however, general guidelines and goals for nutritional support in patients • Prolong life and improve prognosis after liver transplantation with ALD have been proposed (Tables 3 and 4). • Control the costs and discomforts of therapy insofar as possible

• Avoid potential unwanted side effects of therapy, including encephalopathy, Special Considerations azotemia, electrolyte or water imbalance, aspiration, venous thrombosis or throm­ bophlebitis, and sepsis Several specific nutrients, outlined below, require further evaluation as SOURCE: Nompleggi and Bonkovsky 1994, with permission. nutritional therapy for ALD patients.

Vol. 29, No. 4, 2006 303 Silymaryin appears to have antioxidant Phosphatidylcholine taking a placebo (Lieber et al. 2003b). and toxin-reducing properties in the This may have been because of much liver and has been shown to reduce Phosphatidylcholine is a soybean extract lower than expected progression of liver fibrosis and enhance liver regener­ that also is a component of cell mem­ fibrosis in the control group attributable ation in animal studies (Lieber et al. branes. It theoretically is an attractive to decreased alcohol intake in patients 2003a). However, a systematic review therapy because of its low side effects, given supportive care. Thus, the pro­ and meta-analysis of studies involving antioxidant properties, and fibrosis- jected statistical analysis could not be milk thistle in patients with chronic reducing effects (Lieber et al. 2003b). conducted. Further data are needed liver disease, the majority of whom Studies with phosphatidylcholine in before this approach can be considered suffered from alcoholic cirrhosis, failed baboons demonstrated a reduction in as therapy in this patient population. to detect a benefit in liver histology hepatic fibrosis (Lieber et al. 1994). or mortality. The review did note that The positive results in animals led to a the use of milk thistle was safe and well long-term, multicenter study in humans Dietary Fat tolerated (Jacobs et al. 2002). Another with chronic alcohol-induced liver dis­ Fat has been suggested as a risk factor review on the use of milk thistle in ease. Patients with biopsy-proven early for the development and aggravation of ALD treatment concluded that studies fibrosis attributed to alcohol were treated ALD in animals and humans (Mezey on the use of milk thistle were too limited with phosphatidylcholine or a placebo 1998). More than 90 percent of heavy by poor quality and lack of standard­ for 2 years. Subsequent liver biopsies drinkers develop fatty liver (McCullough ization of the milk thistle preparation to after treatment failed to demonstrate a and O’Connor 1998). Continued alco­ recommend its use (Ball and Kowdley beneficial effect on progression of hep­ hol ingestion in the presence of a fatty 2005). atic fibrosis compared with patients liver can lead to the development of fibrosis and cirrhosis. Countries with a Table 4 Guidelines for Daily Dietary Feeding in Alcoholic Liver Disease (ALD) diet high in saturated fats have a lower incidence of ALD despite equivalent Protein = 1.0 to 1.5 kg body weight alcohol ingestion (McCullough and O’Connor 1998). Research with ani­ Total calories = 1.2 to 1.4 times resting energy expenditure with a minimum of mals also has shown that intake of 30 kcal/kg body weight increased saturated fat can reverse the changes of ALD despite continued • 50 to 55% as carbohydrate (preferably as complex carbohydrates) • 30 to 35% as fat (preferably high in unsaturated fat and with adequate essen­ alcohol intake (You et al. 2005). This tial fatty acids) may, in part, be attributed to less free- radical production from saturated fat. Nutrition should be given enterally by voluntary oral intake and/or by small-bore Moreover, a recent study (You et al. feeding tube 2005) concluded that a diet high in • PPN is second choice saturated fats and alcohol stimulates • TPN is last choice the release of adiponectin from fat cells. Salt and water intake should be adjusted for patient’s fluid volume and It appears that adiponectin is responsi­ electrolyte status ble, at least partially, for the protective effect of saturated fats in animal models Liberal multivitamins and minerals of ALD by its beneficial effect on hepatic fat metabolism and reduction in harm­ Specialized BCAA-enriched supplements not usually necessary as most ful cytokines. Diets high in saturated patients tolerate standard AA supplements fats, however, are well known to increase the risk of cardiovascular disease in • Reserve BCAA formulations for patients who cannot tolerate the necessary amount of standard AA (which maintain nitrogen balance) without precipitating humans. Thus, studies involving nutri­ encephalopathy tional supplementation of adiponectin • Avoid supplements providing only BCAA; they do not maintain nitrogen balance as a therapy for ALD need to be under­ • Conditionally essential AA as well as all essential AA are needed taken before a diet high in saturated • Conditionally essential AA are those that normally can be synthesized from fats can be recommended. other precursors but cannot be synthesized in cirrhotic patients. These include choline, cystine, taurine, and tyrosine Vitamins AA = amino acids; BCAA = branched chain amino acids; PPN = peripheral parenteral nutrition (nutrition is pro­ vided intravenously via a peripheral vein); TPN = total parenteral nutrition (nutrition is provided intravenously via Vitamin deficiencies are common a central vein, more concentrated than PPN). among people who consume excessive SOURCE: McCullough and O'Connor 1998, with permission. amounts of alcohol, even in the absence of liver disease. Deficits of vitamins A,

304 Alcohol Research & Health Nutritional Therapy in ALD

B1 (thiamine), B6 (pyridoxine), and Conclusion tion in acute alcoholic hepatitis. American Journal folic acid are of particular concern, as of Gastroenterology 82(9):871–875, 1987. PMID: they not only are common but may It is obvious that nutrition plays some 3307390 exacerbate the detrimental effects of part in ALD given the prevalence of ALBRECHT, J., AND NORENBERG, M.D. Glutamine: alcohol. Other vitamins that often are malnutrition, especially of the protein– A Trojan horse in ammonia neurotoxicity. Hepatology 44(4): 788–794, 2006. PMID: 17006913 lacking in those with ALD include B2 calorie type. The malnutrition usually (riboflavin), B3 (niacin), C, D, and E. is associated with disease seen in hospi­ ALVAREZ, M.A.; CABRE E.; LORENZO-ZUNIGA, V.; Vitamin deficiencies in those with talized patients and correlates with the ET AL. Combining steroids with enteral nutrition: A better therapeutic strategy for severe alcoholic hep­ ALD are caused by a combination of severity of ALD. The primary, estab­ atitis? Results of a pilot study. European Journal of decreased intake, absorption, and stor­ lished therapy for ALD consists of Gastroenterology and Hepatology 16(12):1375–1380, age (Lieber 2003). In addition, alcohol abstinence from alcohol. Good nutrition 2004. PMID: 15618848 often interferes with the conversion of improves nitrogen balance, may improve BALL, K.R., AND KOWDLEY, K.V. A review of vitamins to a metabolically active form: liver tests, and may decrease hepatic fat Silybum marianum (milk thistle) as a treatment one example is vitamin A. The enzymes accumulation, but generally it does not for alcoholic liver disease. Journal of Clinical needed to convert vitamin A to its enhance survival. This suggests that Gastroenterology 39(6):520–528, 2005. PMID: active form are part of the same family adequate nutrition is beneficial when 15942440 of enzymes that metabolize alcohol administered with other forms of treat­ BONKOVSKY, H.L.; FIELLIN, D.A.; SMITH, G.S.; ET (Leevy and Moroianu 2005). The ment but is not sufficient therapy by AL. A randomized, controlled trial of treatment of competition between vitamin A and alcoholic hepatitis with parenteral nutrition and itself. It has been suggested that sufficient oxandrolone. I. Short-term effects on liver function. alcohol results in decreased levels of nutritional repletion coupled with other American Journal of Gastroenterology 86(9):1200– active vitamin A. treatment modalities may be effective 1208, 1991a. PMID: 1909085 Identification and restitution of in reducing complications associated BONKOVSKY, H.L.; SINGH, R.H.; JAFRI, I.H.; ET AL. vitamin deficiencies in patients with with ALD—particularly infection. A randomized, controlled trial of treatment of alco­ ALD is essential. This especially is true Optimal nutrition requires adequate holic hepatitis with parenteral nutrition and oxan­ regarding vitamin B1 (thiamine), because protein calories and vitamins. Ideally, drolone. II. Short-term effects on nitrogen metabolism, thiamine deficiency can cause acute the patient should receive adequate metabolic balance, and nutrition. American Journal Wernicke’s encephalopathy, a degenera­ of Gastroenterology 86(9):1209–1218, 1991b. nutrition orally or through a feeding PMID: 1909086 tive brain disease that potentially is rapidly tube. This may require a feeding tube CABALLERIA, J.; PARES, A.; BRU, C.; ET AL. Metadoxine reversible (Lieber 2003). Replacement that goes through the nose to the stom­ of vitamin A, however, is more compli­ accelerates fatty liver recovery in alcoholic patients: ach (i.e., nasogastric) or, if that is not Results of a randomized double-blind, placebo-con­ cated. Even standard recommended possible, intravenous nourishment. trol trial. Spanish Group for the Study of Alcoholic doses of vitamin A can become toxic Administration of nutritional therapy Fatty Liver. Journal of Hepatology 28(1):54–60, with continued alcohol consumption. (i.e., amino acids/nitrogen) is tolerated 1998. PMID: 9537864 Moreover, it is difficult to estimate vita­ well with little adverse effect in patients CABRE, E.; GONZALEZ-HUIX, F.; ABAD-LACRUZ, A.; min A levels because serum levels often with ALD. There apparently is no ET AL. Effect of total enteral nutrition on the short- do not reflect stores of vitamin A in the term outcome of severely malnourished cirrhotics: problem with the precipitation of hep­ body. Many of the remaining vitamins A randomized controlled trial. Gastroenterology atic encephalopathy. Specific nutrients (B complex, C, D, E, and folic acid) 98(3):715–720, 1990. PMID: 2105256 (i.e., phosphatidylcholine, SAM, and can be restored with amounts found in CABRE, E.; RODRIGUEZ-IGLESIAS, P.; CABALLERIA, saturated fat), although generally standard multivitamins (Lieber 2003). J.; ET AL. Short- and long-term outcome of severe Vitamin E deficiency and replacement innocuous and well tolerated, require alcohol-induced hepatitis treated with steroids or further evaluation. Overall, nutrition is enteral nutrition. Hepatology 32(1):36–42, 2000. is of special interest. As mentioned pre­ PMID: 10869286 viously, there is an increased formation not a panacea in the treatment of ALD, of free radicals in ALD (as reviewed in but it makes a significant positive con­ CALVEY, H.; DAVIS, M.; AND WILLIAMS, R. Dasarathy and McCullough 2003). tribution, especially in selected mal­ Controlled trial of nutritional supplementation, nourished patients. ■ with and without branched chain amino acid Vitamin E is an antioxidant that prevents enrichment, in treatment of acute alcoholic hepati­ the degradation of lipids by oxidation tis. Journal of Hepatology 1(2):141–151, 1985. PMID: 3932509 (i.e., lipid peroxidation) and free radi­ Financial Disclosure cal formation (Traber and Sies 1996). CASAFONT, F.; SANCHEZ, E.; MARTIN, L.; ET AL. It would appear logical that replace­ Influence of malnutrition on the prevalence of bac­ ment of vitamin E potentially would The authors declare that they have no terial translocation and spontaneous bacterial peri­ be beneficial in treating ALD; however, competing financial interests. tonitis in experimental cirrhosis in rats. Hepatology 25(6):1334–1337, 1997. PMID: 9185748 one study (Mezey et al. 2004) that ran­ domly assigned patients to receive vita­ CHATAURET, N.; DESJARDINS, P.; ZWINGMANN, C.; min E or a placebo did not demon­ References ET AL. Direct molecular and spectroscopic evidence for increased ammonia removal capacity of skeletal strate improvement in patients with ACHORD, J.L. A prospective randomized clinical muscle in acute liver failure. Journal of Hepatology mild to moderate alcoholic hepatitis. trial of peripheral amino acid-glucose supplementa­ 44:1083–1088, 2006. PMID: 16530878

Vol. 29, No. 4, 2006 305 DASARATHY, S., AND MCCULLOUGH, A.J. Alcohol LIEBER, C.S.; WEISS, D.G.; GROSZMANN, R.; ET AL. Seminars in Liver Disease 11(4):340–348, 1991. induced liver disease. In: Schiff, E.R.; Sorrell, M.F.; II. Veterans Affairs Cooperative Study of poly­ PMID: 1763339 and Maddrey, W.C., eds. Diseases of the Liver. 9th enylphosphatidylcholine in alcoholic liver disease. ed. Philadelphia: Lippincott, Williams & Wilkins, Alcoholism: Clinical and Experimental Research MEZEY, E.; CABALLERIA, J.; MITCHELL, M.; ET AL. 2003. p. 1019. 27(11):1765–1772, 2003b. PMID: 14634492 Effect of parenteral amino acid supplementation on short-term and long-term outcomes in severe DIEHL, A.M.; BOITNOTT, J.K.; HERLONG, H.F.; ET MARCHESINI, G.; DIOGUARDI, F.S.; BIANCHI, G.P.; alcoholic hepatitis: A randomized clinical trial. AL. Effect of parenteral amino acid supplementation ET AL. Long-term oral branched-chain amino acid Hepatology 14(6):1090–1096, 1991. PMID: 1959859 in alcoholic hepatitis. Hepatology 5(1):57–63, 1985. treatment in chronic hepatic encephalopathy: A PMID: 3917968 randomized double-blind casein-controlled trial. MEZEY, E.; POTTER, J.J.; RENNIE-TANKERSLEY, L.; The Italian Multicenter Study Group. Journal of ET AL. A randomized placebo controlled trial of FERENCI, P. Brain dysfunction in fulminant hepatic Hepatology 11(1):92–101, 1990. PMID: 2204661 vitamin E for alcoholic hepatitis. Journal of failure. Journal of Hepatology 21(4):487–490, 1994. Hepatology 40(1):40–46, 2004. PMID: 14672612 PMID: 7814792 MATO, J.M.; CAMARA, J.; FERNANDEZ DE PAZ, J.; ET AL. S-adenosylmethionine in alcoholic liver cir­ NASRALLAH, S.M., AND GALAMBOS, J.T. Amino HIRSCH, S.; BUNOUT, D.; DE LA MAZA, P.; ET AL. rhosis: A randomized, placebo-controlled, double- acid therapy of alcoholic hepatitis. Lancet 2(8927): Controlled trial on nutrition supplementation in blind, multicenter clinical trial. Journal of Hepatology 1276–1277, 1980. PMID: 6108449 outpatients with symptomatic alcoholic cirrhosis. 30(6):1081–1089, 1999. PMID: 10406187 Journal of Parenteral and Enteral Nutrition 17(2): NAVEAU, S.; PELLETIER, G.; POYNARD, T.; ET AL.A 119–124, 1993. PMID: 8455312 MCCULLOUGH, A.J., AND O’CONNOR, J.F. randomized clinical trial of supplementary par­ Alcoholic liver disease: Proposed recommendations enteral nutrition in jaundiced alcoholic cirrhotic HIRSCH, S.; DE LA MAZA, M.P.; GATTAS, V.; ET AL. for the American College of Gastroenterology. patients. Hepatology 6(2):270–274, 1986. PMID: Nutritional support in alcoholic cirrhotic patients American Journal of Gastroenterology 93(11):2022– 3082733 improves host defenses. Journal of the American 2036, 1998. PMID: 9820369 College of Nutrition 18(5):434–441, 1999. PMID: NOMPLEGGI, D.J., AND BONKOVSKY, H.L. 10511325 MENDENHALL, C.L.; ANDERSON, S.; WEESNER, Nutritional supplementation in chronic liver disease: R.E.; ET AL. Protein-calorie malnutrition associated An analytical review. Hepatology 19(2):518–533, JACOBS, B.P.; DENNEHY, C.; RAMIREZ, G.; ET AL. with alcoholic hepatitis. Veterans Administration 1994. PMID: 8294109 Milk thistle for the treatment of liver disease: A sys­ Cooperative Study Group on Alcoholic Hepatitis. tematic review and meta-analysis. American Journal American Journal of Medicine 76(2):211–222, RAMBALDI, A., AND GLUUD, C. S-adenosyl-L­ of Medicine 113(6):506–515, 2002. PMID: 12427501 1984. PMID: 6421159 methionine for alcoholic liver diseases. Cochrane Database System Reviews (4):CD002235, 2001. KEARNS, P.; YOUNG, H.; GARCIA, G.; ET AL. MENDENHALL, C.; BONGIOVANNI, G.; GOLDBERG, Accelerated improvement of alcoholic liver disease PMID: 11687153 S.; ET AL. VA Cooperative Study on Alcohol with enteral nutrition. Gastroenterology 102(1): Hepatitis. III: Changes in protein-calorie malnutri­ SCHENKER, S., AND HALFF G.A. Nutritional ther­ 200–205, 1992. PMID: 1727754 tion associated with 30 days of hospitalization with apy in alcoholic liver disease. Seminars in Liver LEEVY, C.M., AND MOROIANU, S.A. Nutritional and without enteral nutritional therapy. Journal of Disease 13(2):196–209, 1993. PMID: 8337604 aspects of alcoholic liver disease. Parenteral and Enteral Nutrition 9(5):590–596, Clinics in Liver SIMON, D., AND GALAMBOS, J.T. A randomized 1985. PMID: 3930765 Disease 9(1):67–81, 2005. PMID: 15763230 controlled study of peripheral parenteral nutrition LIEBER, C.S. Relationships between nutrition, alco­ MENDENHALL, C.; ROSELLE, G.A.; GARTSIDE, P.; in moderate and severe alcoholic hepatitis. Journal hol use, and liver disease. Alcohol Research & Health AND MORITZ, T. Relationship of protein calorie of Hepatology 7(2):200–207, 1988. PMID: 27(3):220–231, 2003. PMID: 15535450 malnutrition to alcoholic liver disease: A reexamina­ 3142949 tion of data from two Veterans Administration TRABER, M.G., AND SIES, H. Vitamin E in LIEBER, C.S.; LEO, M.A.; CAO,Q.; ET AL. Silymarin Cooperative Studies. Alcoholism: Clinical and humans: Demand and delivery. Annual Review of retards the progression of alcohol-induced hepatic Experimental Research 19(3):635–641, 1995. 16:321–347, 1996. PMID: 8839930 fibrosis in baboons. Journal of Clinical Gastroenterology PMID: 7573786 Nutrition 37(4):336–339, 2003a. PMID: 14506392 YOU, M.; CONSIDINE, R.V.; LEONE, T.C.; ET AL. MEZEY, E. Dietary fat and alcoholic liver disease. LIEBER, C.S.; ROBINS, S.J.; LI, J.; ET AL. Phosphatidyl­ Hepatology 28(4):901–905, 1998. PMID: 9755223 Role of adiponectin in the protective action of choline protects against fibrosis and cirrhosis in the dietary saturated fat against alcoholic fatty liver in baboon. Gastroenterology 106(1):152–159, 1994. MEZEY, E. Interaction between alcohol and nutri­ mice. Hepatology 42(3):568–577, 2005. PMID: PMID: 8276177 tion in the pathogenesis of alcoholic liver disease. 16108051

306 Alcohol Research & Health