Exercise-based Interventions for Nonalcoholic Fatty Liver Disease: A Meta-analysis and Meta-regression

BACKGROUND & AIMS: The burden of nonalcoholic fatty liver disease (NAFLD) is increasing worldwide. We performed a meta-analysis to determine the effectiveness of exercise-based lifestyle interventions on liver-specific end points in populations with NAFLD and underlying metabolic disorders such as obesity, type 2 diabetes, or metabolic syndrome. METHODS: We searched PubMed-MEDLINE, Embase, and the Cochrane Central register through October 21, 2015 for randomized trials of exercise-based lifestyle interventions on end points such as intrahepatic lipid content and blood levels of alanine and aspartate aminotransferases. Effect sizes are reported as standardized mean difference and weighted mean difference values. To investigate heterogeneity, we performed sensitivity and meta-regression analyses. Results were reported according to the PRISMA statement. RESULTS: We analyzed data from 28 trials. Physical activity, independently from diet change, was associated with a significant reduction in intrahepatic lipid content (standardized mean difference, -0.69; 95% confidence interval [CI], -0.90 to -0.48) and with reductions in alanine aminotransferase (weighted mean difference, -3.30 IU/L; 95% CI, 5.57 to -1.04) and aspartate aminotransferase (weighted mean difference, -4.85 IU/L; 95% CI, -8.68 to -1.02). By meta-regression, we found individuals with increasing body mass index to be increasingly more likely to benefit from the intervention (beta coefficient = -0.10; P = 037). We recorded no effect modification by variables related to the intensity of the intervention. CONCLUSIONS: In a meta-analysis of randomized trials, we found strong evidence that physical activity reduces intrahepatic lipid content and markers of hepatocellular injury in patients with NAFLD. This effect correlated with baseline body mass index.