Optimized strategy among diet, exercise, and pharmacological interventions for nonalcoholic fatty liver disease: A network meta-analysis of randomized controlled trials

Obes Rev. 2024 Jun;25(6):e13727. doi: 10.1111/obr.13727. Epub 2024 Mar 21.

Abstract

Background: Emerging treatment methods, including exercise, diet, and drugs, for nonalcoholic fatty liver disease have been proposed. However, the differences in their efficacy have not been determined. We aimed to compare the effects of these treatments excluding surgery via a systematic review and network meta-analysis of randomized controlled trials.

Data source: The data sources included PubMed, Embase, Web of Science and Cochrane up to February 1st, 2023. The endpoints consisted of body mass index (BMI), serum markers of metabolism and liver injury markers, liver fat content, and stiffness.

Results: A total of 174 studies with 10,183 patients were included in this meta-analysis. In terms of improving BMI, Pan-agonist of peroxisome proliferator-activated receptors (PPAR) is the best treatment with the highest SUCRA (surface under the cumulative ranking) of 84.8% (mean = -3.40, 95% CI -5.55, -1.24) by the comparative effectiveness ranking. GLP-1 (glucagon-like peptide-1) has the best effect in improving the liver fat content based on the MRI-PDFF, steatosis score (SUCRA 99.7%, mean = -2.19, 95% CI -2.90, -1.48) and ballooning score (SUCRA 61.2%, mean = -0.82, 95% CI -4.46, 2.83).

Conclusions: Pan-agonist of PPAR was the most efficacious regimen in lowering BMIs, whereas GLP-1R agonists achieved the highest efficacy of steatosis improvement in this network meta-analysis.

Keywords: efficacy; network meta‐analysis; non‐alcoholic fatty liver disease; pharmaceuticals.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't
  • Systematic Review
  • Review

MeSH terms

  • Diet
  • Exercise Therapy / methods
  • Exercise*
  • Humans
  • Network Meta-Analysis*
  • Non-alcoholic Fatty Liver Disease* / drug therapy
  • Randomized Controlled Trials as Topic*