ST. LOUIS — For patients told they have both fatty liver and prediabetes, the standard clinical advice has been consistent for years: lose weight, and lose it by whatever method you can sustain. A new trial published this week in Cell Metabolism challenges that framing with clinical precision. The method, it turns out, matters almost as much as the number on the scale.
Researchers at Washington University School of Medicine in St. Louis randomly assigned 55 adults with metabolically unhealthy obesity, prediabetes, and fatty liver disease to follow one of three popular diets for roughly four to five months: a ketogenic diet restricting carbohydrates to fewer than 30 grams per day; a Mediterranean diet that balanced carbohydrates, fats, and proteins; or a plant-forward, low-fat diet high in carbohydrates. To strip out the confounding effect of adherence that undermines most diet studies, participants received 100 percent of their food throughout the trial and met weekly with a study dietitian. All three groups lost approximately the same amount of weight, around 10 percent of their starting body mass.
After that, the results diverged sharply.
Participants following the ketogenic diet reduced liver fat by 67 percent. The Mediterranean and low-fat groups reduced liver fat by 45 percent. About half of those in the ketogenic group, 50 percent, no longer met the clinical criteria for prediabetes by the end of the trial. That compared with 29 percent in the Mediterranean group and 7 percent among those following the plant-forward diet.
“For patients with obesity, prediabetes and fatty liver disease, weight loss induced by a very low-carbohydrate diet provides additional therapeutic effects on glucose and lipid metabolism that should further help prevent the progression to more severe metabolic diseases than weight loss alone,” said Samuel Klein, MD, the Danforth Professor of Medicine and Nutritional Science at WashU, who led the study. Klein noted in its press release that the findings point toward carbohydrate restriction as a distinct therapeutic mechanism, not merely a path to weight loss.
The finding centers a question the field has debated for more than two decades: is ketogenic dieting doing something to the liver and the metabolic system that has nothing to do with the weight it removes? The researchers believe the answer is yes. When carbohydrates are sharply restricted, the body shifts its primary fuel source toward fat, which reduces the liver’s own production of triglycerides. That reduction accelerates the clearance of hepatic fat beyond what caloric restriction alone produces. For patients with metabolic-associated steatotic liver disease, the condition previously known as nonalcoholic fatty liver disease, that mechanism is clinically significant. It distinguishes a liver recovering from one that is stalled.
Fatty liver disease affects an estimated 80 to 100 million Americans and is the leading cause of liver-related illness in the United States, according to the National Institute of Diabetes and Digestive and Kidney Diseases, which has documented its progression in its guidance as a sequence from fat accumulation to inflammation to fibrosis and potentially cirrhosis. Weight loss interrupts that progression. The WashU trial suggests that the form weight loss takes determines how forcefully the interruption occurs.
The prediabetes remission numbers are equally striking. A 7 percent remission rate in the plant-forward group is not a failure in absolute terms, but it sits against 50 percent in the ketogenic group as a stark comparative signal. Researchers attribute the gap to insulin dynamics: a diet that sharply limits glucose intake suppresses insulin levels across the day, reducing the chronic demand on beta cells that have been taxed by years of glucose excess. The prediabetes remission data, published in Cell Metabolism alongside the liver fat findings, will likely shape clinical conversations about dietary intervention for patients at the threshold between prediabetes and type 2 diabetes.
The trial’s most transparent limitation is its duration. Four to five months is not long enough to know whether these outcomes persist. The study was also conducted under conditions no patient actually lives in: every meal provided, every week attended, every food decision removed. How a 50 percent prediabetes remission rate performs in real life, where adherence to a very low-carbohydrate diet over years is difficult for most people, remains an open question. Klein’s team acknowledged the need for longer-term adherence studies before keto can be designated a first-line intervention for this population.
The trial arrives as GLP-1 weight-loss drugs like semaglutide are rapidly expanding as the go-to pharmacological intervention for metabolic obesity. Early data on GLP-1 therapies also shows meaningful liver fat reduction and prediabetes reversal. The WashU trial does not compare keto against pharmacological treatment directly. But for patients who cannot access, afford, or tolerate those drugs, the trial provides the clearest comparison yet of which dietary approach does the most liver-specific metabolic work independent of weight. Research showing how meditation reduces inflammation at the gene level has similarly pointed toward non-pharmacological interventions reaching deeper into metabolic biology than the field assumed.
For the patient in a clinician’s office with rising blood sugar and an enlarged liver, who has been told to lose weight, the WashU data offers the closest thing the literature has produced to a controlled answer about which diet to choose. The weight will come off on any of the three. What the liver does next depends on what gets cut from the plate.
Related Posts
