A Bayesian network meta-analysis made available online January 12, 2026 in the Journal of Sport and Health Science finally puts numbers on the exercise question for fatty liver. The team pooled 24 randomized controlled trials with 961 adults with MASLD and modeled both which exercise type worked best and how much of it was enough. The answer to the second question is the part readers will want to write down.
On the type question, combined aerobic and resistance training ranked first for reducing hepatic steatosis, with the highest surface under the cumulative ranking curve at 0.81. The dose-response analysis sharpened the story: benefits were non-linear, rising as exercise increased, then flattening. The plateau arrived around 630 MET-minutes per week, and the peak benefit near 850, after which extra effort bought little.
The MET-minute number translates into ordinary life. At a moderate intensity of about 4.5 METs, 630 MET-minutes equals roughly 30 minutes five times a week, and the analysis says that lands in the cost-effective zone where most of the benefit has already arrived. Vigorous activity at 7.5 METs reaches the same dose in fewer sessions, which gives readers with tight schedules a choice rather than a single rule.
The minimum-dose finding is the sleeper detail. Overall, 460 MET-minutes per week cleared the threshold of clinically significant liver-fat reduction, but combined aerobic-resistance training reached that threshold at only 130 MET-minutes per week, less than a third of the general minimum. That means a modest, realistic program of walking plus a short resistance session can already move the liver, which is a genuinely encouraging number for beginners.
The analysis also ranked the other styles. High-intensity interval training showed progressive benefits and needed about 880 MET-minutes per week for optimal effect, moderate continuous training about 590, and resistance training alone needed to exceed 640 to join the club. The pattern says the combined approach is the most efficient, which is why it led the ranking rather than because the others fail.
The non-linear shape is the part that changes advice. Because the benefit flattens after the plateau, the sensible target is not as much as possible but the cost-effective zone, which protects against the all-or-nothing trap where a person skips everything because they cannot do the ideal. The analysis says the ideal is modest, which makes it reachable, and reachable is what keeps people doing it.
The independence from the scale is worth restating, because it lines up with earlier work covered on this site. Exercise lowers intrahepatic fat, fibrosis markers, and inflammation even when body weight barely moves, and the trials inside this meta-analysis show the liver responds to the movement itself. A reader discouraged by a stubborn scale still has a lever that works, and the lever starts at a very low dose.
The limits are the usual ones for pooled analysis. The 24 trials vary in intervention length, supervision, and adherence measurement, and MET-minute estimates are modeled from self-reported activity, which carries error. The network also compares styles indirectly, so the ranking is the best current synthesis rather than a final verdict, and head-to-head trials would firm it up further.
The practical starting plan is simple enough to begin tonight. A brisk 30-minute walk five days a week clears the 630 MET-minute cost-effective zone, and adding two short resistance sessions, body-weight squats, push-ups, bands, or light weights, pushes the combined benefit that ranked first. Beginners can start at 130 MET-minutes, roughly three short sessions, and build, which the analysis explicitly supports.
The plateau has a liberating side. A reader who cannot manage an hour a day is not failing; the curve says the marginal benefit above 630 MET-minutes is small, so the gap between a good program and a perfect one is mostly wasted effort. That reframe turns exercise from a performance standard into a dose to hit, which is easier on the psychology and just as honest to the data.
None of this replaces the clinician. The meta-analysis is about exercise dosing for steatosis, not a treatment plan for a person with confirmed fibrosis or other liver disease, and anyone with significant stiffness or decompensation needs the care team to set the activity plan. The exercise is the daily lever a reader controls; the elastography and the prescription belong to the doctor.
The exercise finding also pairs naturally with the fasting meta-analysis covered earlier today. Both are timing and dose questions, and the two interventions work through overlapping insulin and fat-oxidation pathways, which is why the combined lifestyle arm in trials tends to out-perform either alone. A reader using a modest eating window and a modest exercise dose is stacking the two best-evidenced levers at once.
The dose-response model also gives clinicians a concrete prescription to offer, which is the quiet advance of the paper. Instead of vague advice to exercise more, the clinic can now say 630 MET-minutes a week, and the patient can convert that into minutes and sessions. Numbers are stickier than intentions, and the analysis turns an intention into a target, which is exactly what the evidence-based conversation needed.
Starting an exercise plan is safe for most people, and the analysis’s low minimum dose lowers the bar further. A person with confirmed liver disease, heart conditions, or joint problems should clear the plan with the care team first, then start at the 130 MET-minute entry point and build gradually. The evidence rewards starting, not heroics, and a modest consistent program beats an intense abandoned one every time.
The exercise dose also works hardest when the plate cooperates. A reader hitting 630 MET-minutes a week while eating the friendly pattern this site rates gets the double effect, less substrate for the liver to turn to fat and a body better equipped to burn it. The meta-analysis is about movement, and the diet is the other half of the same lever, which is why the two appear together in the lifestyle arm of every MASLD trial.
Which exercise type lowered liver fat most?
Combined aerobic and resistance training ranked first, with the highest surface under the cumulative ranking curve at 0.81. It also reached the minimum clinically significant dose at just 130 MET-minutes per week, far below the 460 needed overall.
How much exercise is enough?
Benefits plateaued around 630 MET-minutes per week and peaked near 850. In plain terms, moderate activity for about 30 minutes five times a week lands in the cost-effective zone, and doing much more bought little extra liver benefit.
Does this replace diet?
No. Exercise and diet are separate levers, and the trial evidence shows exercise works even without major weight loss. The plate still decides the inputs the liver receives, and exercise decides how efficiently the body handles them, so the two belong together.