Weight loss advice usually treats “how many calories” as the only variable that matters. A new randomized trial complicates that idea: three different diets produced almost identical weight loss, yet one of them dramatically outperformed the others at clearing fat from the liver and improving blood sugar control.
Key Takeaways
- In a randomized trial of 55 adults with obesity, prediabetes, and fatty liver disease, participants followed a ketogenic, Mediterranean, or low-fat plant-forward diet for 4 to 5 months.
- All three groups lost a similar amount of weight (about 10% of body weight), but the keto group saw liver fat drop by 67%, compared to 45% in the other two groups.
- About 50% of the keto group no longer met the criteria for prediabetes by the end of the trial, compared to 29% on the Mediterranean diet and just 7% on the low-fat diet.
- This was a small, short trial, and a registered dietitian not involved in the study cautions keto isn’t meant as a long-term diet, particularly for people with heart disease risk factors.
What the Researchers Actually Tested
Led by Dr. Max Petersen at Washington University School of Medicine in St. Louis, and published in Cell Metabolism, the trial randomly assigned participants to one of three diets: a ketogenic diet (about 4% carbohydrates, 73% fat), a Mediterranean diet (about 50% carbohydrates, 35% fat), or a low-fat, mostly plant-based diet (about 70% carbohydrates, 15% fat). All food was provided to participants rather than self-reported, a design feature that strengthens confidence in the results compared to studies relying on people to track their own eating.
What Changed Beyond the Scale
Muscle insulin sensitivity — how efficiently muscles take up blood sugar — improved by roughly 50% across all three groups, regardless of which diet they followed. Liver-specific results told a different story: liver insulin sensitivity improved two to three times more in the keto group than in the other two, and liver fat dropped by 67% on keto compared to 45% on the other diets, a difference the researchers linked to reduced fat production in the liver itself. Blood sugar and insulin measures also improved more broadly in the keto group, translating into a meaningfully higher prediabetes remission rate.
Why This Matters Beyond “Which Diet Is Best”
Dr. Petersen framed the findings as complementary to newer weight-loss approaches rather than as a case against them: “GLP-1-based medications have revolutionized the therapy of obesity. We now have a very effective pharmacotherapy for achieving significant weight loss. But in addition, our study shows that the specific composition of the diet can give you added benefits.” That’s the more interesting scientific point buried in the headline number — weight loss and metabolic health aren’t the exact same thing, and this trial suggests what you eat while losing weight may matter for specific outcomes like liver fat, independent of how much weight comes off.
What an Independent Dietitian Makes of It
Registered dietitian Ilana Muhlstein, who wasn’t involved in the study, called it validation that “calories are not equal” — a reminder that reducing weight management to a simple calorie deficit misses real differences in how diet composition affects the body. She noted the keto group didn’t show worsened cholesterol or blood fat measurements during this short study, which she considers notable, but was direct about her actual clinical recommendation: “Keto is not something I recommend for anyone long-term because of the high saturated fat, and usually it translates into a lower-fiber diet, which is not good for heart health.” Her suggested approach is more nuanced than “keto is better”: use it short-term for people specifically dealing with prediabetes or blood sugar irregularity, then transition toward a Mediterranean-style diet for sustained, long-term heart and brain health. She also flagged an important caveat — this approach may be worth more caution for people with a family history of heart disease or who are otherwise at elevated cardiovascular risk.
What This Doesn’t Prove Yet
The researchers themselves were upfront about the limits here: 55 participants over 4 to 5 months is a real, well-controlled trial, but it’s still small and short by the standards needed to draw firm long-term conclusions. It doesn’t tell us what happens after a year or more on any of these diets, or whether the liver benefits persist, reverse, or grow if keto is continued longer than a few months. The findings apply specifically to people with obesity, prediabetes, and existing fatty liver disease — not a general recommendation for everyone trying to lose weight.
What This Means for You
If you have prediabetes and fatty liver, this trial adds real, controlled evidence that a short-term ketogenic phase might offer a meaningful metabolic head start beyond what weight loss alone would provide — but it’s not positioned, even by the researchers or outside experts, as a diet to stay on indefinitely. Talking to a doctor or dietitian about a structured, time-limited approach — potentially keto first, transitioning to a more sustainable pattern like the Mediterranean diet — is a more evidence-aligned strategy than adopting any single diet permanently based on one trial.
Source: Petersen, M.C., Smith, G.I., et al. (2026). Effect of diet macronutrient content on the cardiometabolic response to weight loss: A randomized clinical trial. Cell Metabolism. DOI: 10.1016/j.cmet.2026.07.020.