Keto diet offers better liver health alongside weight loss

Clinical takeaway: For patients with obesity, prediabetes, and fatty liver, a ketogenic approach delivered the largest improvements in liver fat, hepatic insulin sensitivity, and glucose control. 

Obesity is often accompanied by prediabetes and excess liver fat, with a path that leads toward type 2 diabetes, cirrhosis, and cardiovascular disease unless interrupted. Weight loss remains the cornerstone of treatment because it improves insulin activity across organs and can push metabolic disease into remission. Clinicians advise it and prescribe to support it with little consideration of the best diet along the way. 

Popular diets differ radically in composition, from very low carbohydrate plans to those drawing most of their energy from carbohydrates, and prior comparisons have been muddied by uneven adherence and unequal weight loss between groups, leaving open whether any diet confers advantages of its own. A recent randomized trial held weight loss constant across three popular diets, measuring what happened to metabolic health. 

Weight loss on the very low carbohydrate ketogenic diet cut liver fat by 67%, versus 45% on both the Mediterranean and very low fat plant-forward diets. Hepatic insulin sensitivity, one of the trial's two primary endpoints, rose two to three times more in the ketogenic group than in the others. Prediabetes remitted in half of ketogenic diet participants, compared with 29% on the Mediterranean diet and 7% on the plant-forward plan. 

A surprise ran against expectation for a diet drawing 23% of its energy from saturated fat: LDL cholesterol, apolipoprotein B, and 24-hour triglycerides showed no differences among groups, and HDL cholesterol tended to rise only in the ketogenic group. Blood glucose control followed the liver pattern. Glycated hemoglobin and 24-hour glucose fell most on the ketogenic diet, with 24-hour glucose down 20% from baseline against 8% on the other two plans, and 24-hour insulin down 74% against 44% on the Mediterranean and 27% on the plant-forward diet. 

Some benefits held steady regardless of the diet plan implemented. Muscle insulin sensitivity, the other primary endpoint, climbed roughly 50% in every group, and body fat and oral glucose tolerance improved without a diet effect, with blood pressure tending lower across groups, evidence that the weight loss itself was the key. 

Researchers randomized 55 adults with metabolically unhealthy obesity, defined as obesity plus prediabetes plus hepatic steatosis, to one of the three diets, which drew 4%, 50%, and 70% of energy from carbohydrate, respectively; 42 completed the study, 14 per group. All food was provided as prepared meals, participants met weekly with a study dietitian, and logged adherence exceeded 95%. Calorie targets were tailored to produce roughly 10% weight loss, reached over about five months, with testing repeated after a short weight stabilization period. 

For clinicians, the finding reframes the diet conversation with metabolically unhealthy patients: the choice of plan is not just an adherence question but a lever on liver and glycemic outcomes, and carbohydrate restriction pulls that lever hardest. The result cannot say whether a less severe carbohydrate cut without ketosis would do the same, or whether the benefits require the weight loss phase rather than starting keto during maintenance. The research team's next question is mechanistic. They are turning to the fundamental drivers of weight loss's metabolic benefits, including the effects of GLP-1 medicines, whose users increasingly face exactly this diet-composition decision. 

"We know weight loss can improve metabolic health in people who are obese," said Samuel Klein, MD, Danforth Professor of Medicine and Nutritional Science, Washington University School of Medicine, and the study's senior author. "Our data suggest that if you have high liver fat content and prediabetes, reducing carbohydrate intake can have important therapeutic effects on metabolism beyond just weight loss alone." 

Source: Petersen MC, et al. (2026 Aug 27) Cell Metab. Effect of diet macronutrient content on the cardiometabolic response to weight loss: A randomized clinical trial