A randomized pilot study in JHEP Reports (2026, PMID 41831607) produced the first head-to-head comparison of a very low energy ketogenic diet versus a Mediterranean diet in biopsy-confirmed MASLD, and the margin was striking: MRI liver fat fell a relative 77 percent in the VLED group versus 14 percent in the Mediterranean group at 12 weeks, with p under 0.01.
The design was tight for a pilot. Twenty-five adults with histologically confirmed MASLD and BMI between 27 and 35 were randomized, 14 to the VLED at about 753 kcal per day and 11 to the Mediterranean plan at about 2,139 kcal per day, with the primary endpoint, change in MRI liver fat fraction, read at week 12 and a 24-week monitoring window that included a repeat biopsy.
The weight numbers explain the liver numbers. Total body weight loss at week 12 was 13 percent in the VLED arm versus 4 percent in the Mediterranean arm, and the steatosis reduction tracked the caloric gap, which is the same weight-loss-first pattern this field keeps reporting, only sharper because the VLED produced a much deeper deficit.
The 24-week story is the part that answers the “what happens after” question. From week 13, the VLED group received low-dose semaglutide, 0.5 mg weekly, for weight maintenance, and at week 24 they had kept a 14 percent total weight loss from baseline, while the Mediterranean group sat at 3 percent, and liver histology improved in both groups with a larger drop in the VLED arm, NAS score minus 2 versus minus 1.
The authors’ conclusion is direct and, in this field, slightly contrarian: the very low energy diet is widely available, easily accessible, and should be more commonly considered for MASLD patients with overweight or obesity, a sentence aimed at clinicians who have treated the Mediterranean diet as the default prescription and skipped the harder conversation about deeper caloric restriction.
Three honest caveats keep the headline honest. The sample is 25 people, a pilot by definition; the MRI measurement was a relative reduction, which flatters small baseline values; and the VLED-plus-semaglutide sequence means the 24-week durability combines two interventions, so the maintenance result cannot be credited to the diet alone.
For readers of this site, the food-level translation is the practical layer. The Mediterranean arm was not a failure, it improved steatosis 14 percent with a modest 4 percent weight loss, and its food pattern, olive oil, fish, vegetables, whole grains, is the plate this database rates friendly row after row, while the VLED arm won by producing a deficit the Mediterranean plan at 2,139 kcal could not match.
The ketogenic-mechanism question is worth separating from the calorie question. The study design cannot tell a ketosis effect apart from a deficit effect, because the VLED is both very low calorie and very low carbohydrate, and the literature does not currently separate those threads cleanly, which is why the honest reading is “deep deficit helped,” not “keys to a liver cure.”
The safety frame is exactly where a food-rating site must stand. A 753-kcal diet is a medically supervised intervention with defined refeeding and monitoring requirements, gallstone risk, lean-mass loss, and micronutrient gaps all need clinical management, and the paper itself frames it as a therapy for overweight and obese patients, not a self-selected plan.
The semaglutide bridge deserves its own paragraph because it is the newest piece of the maintenance story. Low-dose semaglutide at 0.5 mg weekly, far below the 2.4 mg MASH dose approved in Japan this June, was used purely for weight maintenance after the diet phase, and the 14 percent 24-week weight loss with that minimal dose is itself a signal that the hardest part of any diet, keeping the weight off, may have a pharmacological assistant.
For a reader choosing between the two plans in the title, the pilot offers a decision rule rather than a winner. The Mediterranean pattern is the sustainable, rated-friendly long-term frame, and the VLED is a short, supervised, deep-deficit phase that some patients can tolerate with clinical support, and the pragmatic sequence used in trials like this one is exactly that order: deep phase first, maintenance plan second.
The histology finding is the endpoint that matters most and is reported most cautiously. NAS improvement of 2 points versus 1 point at 24 weeks in a 25-person pilot is directional, not definitive, but it is the first signal in this comparison that the deeper diet strategy may change tissue, not just imaging, which is the bar that matters for disease progression.
Access and cost belong in the honest ledger. A supervised VLED requires dietitian time, monitoring visits, and often meal-replacement products, while a Mediterranean pattern runs on ordinary groceries, and the trial’s “widely available” claim means clinically available, not free, which is a real difference for readers deciding what to ask their own care team about.
The takeaway sentence for this site’s audience: the 2026 pilot confirms that deeper caloric deficit moves liver fat faster than any specific food pattern, that the Mediterranean plate remains the durable rated-friendly frame, and that both arms worked, which is the same message as this month’s larger three-diet trial, where weight loss mattered more than the diet label.
Who the study does not speak for matters as much as who it speaks for. The inclusion band, BMI 27 to 35, leaves out lean MASLD, the substantial subgroup with normal weight and fatty liver, for whom an 800-kcal ketogenic phase would be inappropriate, and it leaves out cirrhosis, where aggressive weight loss carries its own risks, so the paper’s message is aimed at a defined clinical population, not the whole fatty liver community.
What did the pilot trial compare?
Twenty-five adults with biopsy-confirmed MASLD and BMI 27-35 were randomized to 12 weeks of a very low energy ketogenic diet (about 753 kcal/day) or a Mediterranean diet (about 2,139 kcal/day). MRI liver fat fell 77% (relative) in the VLED group versus 14% in the Mediterranean group, p<0.01.
Was the weight loss maintained?
Yes. From week 13 the VLED group received low-dose semaglutide (0.5 mg weekly) for maintenance. At 24 weeks total body weight loss was 14% from baseline in the VLED-plus-semaglutide group versus 3% in the Mediterranean group, and liver histology improved more in the VLED arm (NAS -2 vs -1).
Is a very low energy diet something to try at home?
No. The study diet is a medically supervised intervention, and the authors present it as a clinical option for overweight and obese MASLD patients. Anyone considering it needs a doctor and a dietitian involved; this site does not translate a trial protocol into a do-it-yourself plan.