Objectives
This study investigated the effects of different exercise interventions on intrahepatic lipid content and markers of glucose homeostasis and liver function in adults with and without nonalcoholic fatty liver disease (NAFLD).
Methods
A comprehensive search was conducted in PubMed, Scopus, Web of Science, and EMBASE using three primary keywords including "exercise training," "liver fat," and "randomization" from inception to August 2025. Eligible studies were those that compared exercise training with either nonexercise groups or other types of exercise training.
Results
Thirty-eight studies comprising 1880 participants were included. Exercise training was more effective for reducing intrahepatic lipids [SMD: -0.33, p = 0.001], confirmed through subgroup analyses based on age, health status, body mass index, intervention duration, specific health status, frequency of weekly exercise sessions, biological sex, steatosis classification, and weight change%. Based on the p-score rankings, the intervention that was likely to be the most effective for lowering intrahepatic lipids was high-intensity interval training (p = 0.95), followed by aerobic training (p = 0.77), combined aerobic and resistance training (p = 0.52), and resistance training (p = 0.10). Secondarily, exercise training reduced fasting blood glucose [WMD: -2.27 mg/dL, p = 0.007], fasting insulin [SMD: -0.16, p = 0.02], glycated hemoglobin [SMD: -0.13, p = 0.03], and liver function enzymes including alanine aminotransferase (ALT) [WMD: -3.72 U/L, p = 0.001] and aspartate aminotransferase (AST) [WMD: -3.51 U/L, p = 0.02] significantly more than nonexercise groups. However, there were no significant differences in insulin resistance.
Conclusions
These results provide evidence of the effects of different exercise interventions as part of a healthy lifestyle for reducing intrahepatic lipids and markers of glucose dysmetabolism. PROSPERO Registration: CRD42025639421.