ST. LOUIS — For the roughly 97 million American adults living with prediabetes, the question has never been how to lose weight; most diets handle that. The harder question is what happens to the liver, the pancreas, and the blood sugar during the loss. A clinical trial published last week offers the clearest comparative answer yet: the choice of diet matters far more than a simple calories-in, calories-out framework suggests.
Researchers at Washington University School of Medicine in St. Louis enrolled 42 adults with obesity, prediabetes, and metabolic dysfunction-associated steatotic liver disease, the clinical term for fatty liver tied to metabolic syndrome, and assigned them randomly to one of three eating approaches: ketogenic, Mediterranean, or plant-forward low-fat. All three groups lost roughly 10 percent of their body weight over four to five months. That was not the surprise.
What changed beneath that shared number was. Participants on the ketogenic diet saw liver fat fall by 67 percent. Those on the Mediterranean and low-fat regimens saw a 45 percent reduction. The gap, 22 percentage points of liver fat, held even after the investigators accounted for identical weight loss across groups, suggesting the macronutrient composition of the diet is doing something to the liver beyond what calorie restriction alone explains.
The findings, published August 27, 2026 in Cell Metabolism, add clinical-trial weight to a debate that has produced more heat than light in nutrition science for decades. For patients who arrive at a metabolic specialist’s office carrying two diagnoses at once, fatty liver and prediabetes, the trial suggests the starting point of their dietary intervention may matter as much as their adherence to it.
The prediabetes numbers carried the most clinical punch. Half of participants in the ketogenic group, 50 percent, achieved remission of prediabetes by the end of the trial, meaning their blood sugar returned to the normal range without medication. In the Mediterranean group, 29 percent reached remission. In the plant-forward low-fat group, just 7 percent did.
Dr. Max C. Petersen, an assistant professor in the John T. Milliken Department of Medicine at Washington University and the trial’s first author, found that the ketogenic diet produced its strongest results in the two areas most directly tied to carbohydrate restriction: liver fat and blood sugar regulation. The low-carbohydrate approach reduces insulin demand and limits the dietary substrate for liver fat synthesis in ways that lower-fat diets do not, even when total calories and weight loss are matched.
One concern the trial addressed directly was cardiovascular risk. Ketogenic diets, by virtue of their high fat content, have long carried warnings about elevating LDL cholesterol, the type associated with arterial plaque. In this trial, the researchers found no significant difference between groups in LDL or apolipoprotein B, a marker of atherogenic particle burden. That finding will not settle the long-running debate, but it challenges the assumption that the liver-specific benefits of a ketogenic diet necessarily come at a cardiovascular price.
Fatty liver disease has become one of the fastest-growing metabolic diagnoses in the United States, driven by the same dietary patterns and sedentary behavior linked to type 2 diabetes and cardiovascular disease. The condition, previously called nonalcoholic fatty liver disease, affects an estimated 25 to 30 percent of American adults. For most, there are currently no approved pharmaceutical treatments, a gap that has pushed diet and lifestyle intervention to the front of clinical management.
Washington University’s trial is among the few to randomize patients to specific diets rather than study self-reported eating habits, a design that strengthens its conclusions. Earlier observational studies linking low-carbohydrate diets to improved liver markers were difficult to interpret because people who choose ketogenic diets tend to differ from those who favor Mediterranean or low-fat approaches in ways that affect outcomes. A randomized trial can, in theory, isolate the diet effect itself.
The Washington University team noted in its announcement that all three diets were developed and provided to participants rather than simply recommended, controlling for adherence differences that often confound nutrition research.
The trial comes as scientists have begun mapping the cellular mechanisms that prevent fatty liver disease recovery at a molecular level, suggesting the clinical picture is more complex than once appreciated. Understanding which dietary interventions act most effectively on those pathways may reshape how metabolic specialists sequence treatments.
What the trial does not resolve is durability. Four to five months is enough time to see biochemical change; it is not enough time to confirm whether the improvements hold, whether patients can maintain the ketogenic diet’s restrictive profile beyond the trial window, or whether the gap between dietary approaches narrows or widens over years. The researchers have not yet reported a longer-term follow-up arm.
EurekAlert noted in its news release that the study was funded in part by the National Institutes of Health and by the Digestive Diseases Research Core Center at Washington University.
For clinicians managing patients with obesity, prediabetes, and fatty liver disease together, the trial adds data points but not a definitive answer. It does not tell them which patient will sustain the ketogenic diet long-term, which will find the Mediterranean approach more tolerable, or whether the liver fat advantage translates into better outcomes on harder endpoints, including liver cirrhosis, cardiovascular events, or progression to type 2 diabetes. Those questions are larger and will take longer to answer.
The results also add to accumulating evidence on how lifestyle interventions produce metabolic benefits through pathways that differ from pharmacological treatment, a distinction researchers believe may matter for treatment sequencing and patient selection.
The most important number from the trial may be the 50 percent. Half of the ketogenic participants reversed a diagnosis that most people carry for years before progressing to type 2 diabetes, and they did it through food, not a prescription. Whether that conversion rate holds at twelve months, or whether the metabolic gains quietly recede when the supervised trial ends and the structured meal plan disappears, is the question the Cell Metabolism paper leaves open.
Related Posts
