Can a very low-carb diet reduce fatty liver more than other eating plans? A small randomized trial suggests that it can produce a larger short-term reduction in liver fat when people achieve similar weight loss. But the findings are more nuanced than a headline declaring one diet the winner.
The study involved adults with obesity, prediabetes and excess liver fat, not a population with established type 2 diabetes. Weight loss improved important metabolic measures on all three diets. The very-low-carbohydrate ketogenic approach produced additional liver-related benefits, but the trial did not establish long-term safety or show that it prevents cirrhosis, heart attacks or diabetes.
What did the new fatty liver trial compare?
Researchers at Washington University School of Medicine randomly assigned 55 adults to a very-low-carbohydrate ketogenic diet, a Mediterranean diet or a very-low-fat plant-forward diet. Forty-two completed the study and were analyzed, with 14 in each group. The research was published online in Cell Metabolism on 27 August 2026.
Participants lost approximately 10% of their starting weight over about five months. Crucially, the team supplied all study food and provided weekly dietitian support. Matching weight loss helped the researchers examine whether dietary composition made an additional difference. These were carefully supported eating plans, not simply instructions to try a popular diet at home. Read the original randomized trial.
The results: all three diets helped, but liver fat fell most with keto
| Eating plan | Weight loss | Average relative reduction in liver fat |
|---|---|---|
| Very-low-carbohydrate ketogenic | About 10% | 67% |
| Mediterranean | About 10% | 45% |
| Very-low-fat plant-forward | About 10% | 45% |
The liver also became more responsive to insulin in all groups, with a greater improvement in the ketogenic group. Muscle insulin sensitivity improved by roughly 50% across the three diets. Average glucose measured over 24 hours fell by approximately 20% with the ketogenic diet and 8% with each of the other diets. WashU Medicine summarizes the findings.
The most useful interpretation is therefore two-part: weight loss helped across different eating patterns, and the composition of the diet appeared to matter for some liver and glucose measures. The Mediterranean and plant-forward approaches did not fail; they also produced substantial improvements.
Why liver fat and insulin sensitivity matter
Fatty liver means that excess fat has accumulated in the liver. When it occurs alongside metabolic risk factors, the condition is often called metabolic dysfunction-associated steatotic liver disease, or MASLD. It frequently overlaps with obesity, insulin resistance and type 2 diabetes.
Insulin sensitivity describes how effectively the body responds to insulin. Better sensitivity can help the body handle glucose. However, less liver fat is not the same as less liver scarring. A liver-fat scan and a fibrosis assessment answer different questions, so an improvement in one measurement should not be treated as proof that advanced liver disease has reversed.
European liver, diabetes and obesity guidelines already recommend supported, sustained weight loss for people with MASLD and overweight. They also emphasize dietary quality, physical activity and assessment of fibrosis risk, particularly in people with type 2 diabetes. The new trial adds a focused diet comparison; it does not replace that broader care plan. See the EASL-EASD-EASO MASLD guideline.
What this small trial cannot tell us
- Whether the benefits last. Several months of metabolic testing cannot establish years of health benefit or safety.
- How well the diets work without intensive support. Provided meals and weekly dietitian visits differ from shopping, cooking and managing food costs independently.
- Whether everyone would respond similarly. Only 42 of 55 randomized participants completed the study. The findings apply most directly to the particular population studied.
- Whether carbohydrate alone explains the difference. The diets differed in protein and fat composition as well as carbohydrate. This was not an experiment changing just one nutrient while holding everything else constant.
- Whether clinical complications are prevented. The study measured metabolic outcomes. It was not designed to determine which diet prevents heart attacks, cirrhosis or future diabetes.
LDL cholesterol and apolipoprotein B did not differ significantly between groups, but that does not establish equal long-term cardiovascular safety. Small studies are useful for identifying signals worth testing; they cannot settle every question about a dietary pattern. Trial methods and results.
Low-carb and ketogenic diets are not interchangeable
A lower-carbohydrate eating plan can still include fruit, pulses, whole grains and dairy in suitable portions. A ketogenic plan is substantially more restrictive. Diabetes UK describes a low-carb diet as less than 130 grams of carbohydrate a day, but the amount, food choices and support required vary.
That distinction matters when discussing this research: reducing sugary drinks or changing the portion of rice at dinner is not the same intervention as a tightly controlled ketogenic diet. The label alone also says little about fibre, nutritional adequacy or whether a plan fits someone’s life. Diabetes UK’s low-carb guidance explains these practical considerations. Our guide to low-carb versus keto for insulin resistance explores the differences further.
Medication safety comes before a major diet change
For people who already have diabetes, reducing carbohydrate substantially can change medication needs. Insulin and sulfonylureas can cause low blood glucose if carbohydrate intake falls without an appropriate treatment review. A clinician may need to adjust the treatment plan and glucose monitoring. Do not make those dose changes independently. Diabetes UK explains the medication precautions.
There is a separate concern with SGLT2 inhibitors, such as dapagliflozin and empagliflozin: ketogenic eating can increase the risk of diabetic ketoacidosis. The UK Kidney Association advises people taking these medicines against following a ketogenic diet. This is a reason to discuss the diet with the prescribing team, not to stop a medicine on your own. Read the SGLT2 inhibitor safety guideline.
Questions to take to your next appointment
A productive discussion starts with your health priorities and the support available, rather than a demand to follow the strictest diet. Consider asking:
- Do I have evidence of fatty liver, and do I need assessment for liver scarring?
- Which eating approach fits my medications, health conditions, food preferences and budget?
- Would a registered dietitian help me make a nutritionally complete plan?
- What should we monitor, and when should we review progress?
- If carbohydrate intake changes, what is my written medication and low-glucose safety plan?
Needing support is not a failure. The participants in this trial received a high level of practical help. Sustainable progress in everyday life should not be judged against a research setting that supplies every meal.
Frequently asked questions
Does this prove keto cures fatty liver?
No. The trial found a larger reduction in liver fat with the very-low-carb diet under controlled conditions. It did not demonstrate a cure, reversal of cirrhosis or protection against long-term complications.
Should people with type 2 diabetes switch diets?
Not on the basis of this trial alone. Participants had prediabetes, and medication safety and individual circumstances matter. For the broader comparison, read our low-carb versus low-fat guide for type 2 diabetes.
Can a Mediterranean diet still be a reasonable choice?
Yes. It improved metabolic measures in this study and remains consistent with MASLD dietary guidance. Choosing a plan involves overall nutritional quality, safety and the ability to sustain it, not just one trial result.
The bottom line
This study offers a credible reason to investigate very-low-carbohydrate diets further for fatty liver and prediabetes. It also reinforces a less dramatic but useful message: several eating patterns can support meaningful metabolic improvements. The next step is an individualized conversation about safe, sustainable care, not a universal instruction to go keto.
This article provides general medical education, not a diagnosis or a personal diet or medication plan.
Sources
- Petersen MC, Smith GI, Farabi SS, et al. Effect of diet macronutrient content on the cardiometabolic response to weight loss: A randomized clinical trial. Cell Metabolism. Published online 27 August 2026.
- WashU Medicine: Low-carb diet delivers added liver benefits beyond weight loss. 27 August 2026.
- EASL-EASD-EASO Clinical Practice Guidelines on MASLD. 2024.
- Diabetes UK: Low-carb diet and meal plan.
- Diabetes UK: Low-carb diets for type 2 diabetes remission.
- UK Kidney Association SGLT2 inhibitor guideline, 2023 update.