Impact of Nutritional Supplementation in Patients with Alcoholic Cirrhosis and Alcoholic Hepatitis: A Meta-Analysis of Randomized Placebo Controlled Trials

Abstract

Purpose: Whether nutritional supplementation (NS) is beneficial in patients with alcoholic liver disease (ALD) is controversial. We performed a meta-analysis of randomized studies to examine benefits of NS in ALD. Methods: Electronic databases (1980- 2010) were searched to include randomized controlled trials (RCT) on NS in alcoholic cirrhosis (AC) and/or alcoholic hepatitis (AH), and extract data on short-term mortality (STM) at 28 days, long term mortality (LTM) at 1 year, proportion of patients with resolution of encephalopathy or ascites, improvement in serum albumin, AST, or total bilirubin. Pooled data on categorical variables was reported as odds ratios (OR) with 95% confidence interval (CI), and continuous variables as differences in means (pretreatment - post treatment) with 95% CI. Data were considered signifi cant if the 95% CI on OR did not cross one, and on mean difference, did not cross 0. Random effects model was used for analyzing the pooled data. I2 and Egger's tests were used for assessing heterogeneity and publication bias, respectively. Results: A total of 19 RCT were included; 13 in AH patients (n=804), and six in AC (n=818). Seven RCT examined parenteral, and 12 enteral nutrition, of which five used branched chain amino acids (BCAA). RCTs differed with respect to patient cohorts, intervention details, and treatment duration. Pooled data showed trend for lower STM with NS for all studies: 0.87 [0.69-1.09], but not when analyzed separately for AH: 0.85 [0.52-1.38], or AC: 0.87 [0.67-1.13], or with BCAA supplementation: 0.93 [0.71-1.21]. However, STM improved significantly in nine RCTs, with difference between treatment and placebo groups of >1000 calories and/or 40 g protein/d: 0.64 [0.41-0.99]. There were trends for improved LTM: 0.69 [0.46-1.05] and encephalopathy 0.75 [0.51-1.11], without improvement in ascites 1.06 [0.44-2.56] in favor of supplementation. NS was associated with increase in mean serum albumin 2.08 [1.05 to 3.1] and decrease in AST 1.02 [0.18-1.86] and serum bilirubin 1.5 [0.49 - 2.5]. There were no adverse effects of NS. Conclusion: Nutritional supplementation in ALD patients is associated with improved liver tests and a trend for improvement in encephalopathy. A subgroup of patients with anorexia and markedly reduced intake derive potential survival benefit with supplementation. Because of strong medical rationale and good safety profile of nutritional supplementation, it should be considered a standard of care for ALD patients, especially in those with poor appetite and intake. Further large prospective studies are suggested to examine benefits of nutritional supplementation in patients with alcoholic liver disease.