Keto beat Mediterranean for the liver, even with identical weight loss

Weight loss is the standard advice for people whose obesity comes with prediabetes and a fatty liver, but a live question has hung over it: does the composition of the diet matter, or only the weight that comes off? Popular diets take radically different positions on carbohydrate and fat, and people choosing between them have had little direct evidence comparing their metabolic effects at the same amount of weight lost. This trial set out to answer exactly that, in the group with the most to gain: people with metabolically unhealthy obesity, meaning prediabetes plus excess fat stored in the liver.

Researchers randomly assigned adults with metabolically unhealthy obesity to one of three popular diets with markedly different macronutrient content: a very-low-carbohydrate ketogenic diet, a Mediterranean diet, or a very-low-fat plant-forward diet. Each group lost a moderate amount of weight, around 10% of body weight, so the diets could be compared at a matched weight loss. The team measured insulin sensitivity in both muscle and liver, fat stored inside the liver, the liver's own fat production, glycated haemoglobin, and plasma glucose and insulin tracked serially over 24 hours, alongside blood lipids including LDL cholesterol and apolipoprotein B.

Weight loss improved muscle insulin sensitivity by roughly 50% in all three groups, so on that front the diets were equals. The liver told a different story: hepatic insulin sensitivity rose two to three times more in the very-low-carbohydrate group than in the other two. Liver fat content, the liver's fat production, glycated haemoglobin, and 24-hour glucose and insulin all fell most on the ketogenic diet. Importantly, there were no differences between the diets in LDL cholesterol, apolipoprotein B or 24-hour triglycerides, so the liver advantage did not come with a measured lipid penalty in this trial.

This complicates the common view that weight loss itself does all the metabolic work and the diet used to achieve it is interchangeable. It suggests that for the liver specifically, cutting carbohydrate very low adds something beyond the weight lost, at least in people who start with prediabetes and hepatic steatosis. It also pushes back on the expectation that a ketogenic diet necessarily worsens LDL cholesterol, which did not differ here. Open questions remain about how long the advantage lasts, whether people can sustain such a restrictive diet outside a trial, and whether the findings extend to metabolically healthier people.

For a reader carrying extra weight alongside prediabetes or a fatty liver, this is a reason to think the route taken to weight loss may matter for liver health, not just the destination on the scale. It is equally a reason not to dismiss the other diets: muscle insulin sensitivity improved by about 50% regardless of which diet people followed, so the weight loss itself did plenty. This is one trial in a specific metabolic group, so it informs a conversation about diet choice rather than settling it, and it says nothing about ketogenic eating without weight loss.