Losing weight can help a struggling liver, but a new study suggests the number on the scale isn’t the whole story. Two people can lose the exact same amount of weight and walk away with very different livers, depending on what they ate to get there.
That’s the takeaway from a new clinical trial that pitted three popular diets against each other: a very-low-carbohydrate ketogenic diet, a Mediterranean diet, and a very-low-fat, plant-forward diet. Researchers had 42 adults with obesity, prediabetes, and fatty liver disease lose about 10 percent of their body weight on one of the three eating plans. Researchers provided all the food directly to participants to ensure they stuck to their assigned diet. The results, published in Cell Metabolism, show that when it comes to a struggling liver, not all weight loss delivers the same drop in liver fat.
All three diets improved how well muscle handled insulin by about the same amount. But the keto diet pulled far ahead for the liver, cutting stored fat and improving blood sugar control by a wider margin than the other two plans. The results point past pounds lost, toward what happens inside the body once the weight is gone, and why the food on someone’s plate might decide how much the liver recovers.
How the Keto Diet Study Worked
Researchers at Washington University School of Medicine recruited 55 adults with what the study calls “metabolically unhealthy obesity,” meaning they carried excess weight along with prediabetes and a buildup of fat in the liver. Of those, 42 people completed the study: 14 on the ketogenic diet, 14 on the Mediterranean diet, and 14 on the low-fat, plant-forward diet. Participants were on average in their early 40s, and each group included nine women and five men.
To ensure people followed their assigned diet instead of just reporting what they ate, the research team’s kitchen prepared and delivered every meal and snack. Participants met weekly with a dietitian, and people stuck to their assigned diet more than 95 percent of the time in every group. That level of control is rare in nutrition research, which usually relies on people accurately remembering and reporting their own eating habits.
In this trial, the ketogenic diet got about 4 percent of its calories from carbs, 73 percent from fat, and 23 percent from protein. The Mediterranean diet ran about 50 percent carbs, 35 percent fat, and 15 percent protein. The low-fat, plant-forward diet was roughly 70 percent carbs, 15 percent fat, and 15 percent protein. Each participant stayed on their assigned diet until they lost close to 10 percent of their starting weight, which took roughly four to five months depending on the group. Their weight was then held steady for a few weeks before final testing.
To measure how well the body handled insulin, researchers used a test that infuses small, controlled amounts of insulin and sugar into the bloodstream while closely tracking blood levels. Researchers measured liver fat with a specialized MRI scan. Researchers also drew blood every hour for a full day, both before and after weight loss, to track how sugar, fat, and hormone levels moved throughout the day, including after meals.
What the Keto Diet Did to the Liver and Blood Sugar
Muscle responded the same way no matter the diet. All three plans improved how well muscle tissue used insulin by about the same amount, roughly 50 percent. Had the study stopped there, it might have looked like any weight-loss diet works equally well.
But the liver told a different story. The ketogenic diet produced a much bigger improvement in how well the liver responded to insulin than the other two diets. Liver fat fell from around 19 percent to about 6 percent in the keto group, a drop of roughly two-thirds. The Mediterranean and low-fat groups also improved, dropping to roughly 10 percent and 9.6 percent, but the change was smaller.
Blood sugar markers moved in the same direction. A long-term blood sugar marker that reflects average levels over three months dropped by 0.6 percentage points in the keto group, compared with a smaller 0.1-point drop in the Mediterranean group and a 0.4-point drop in the low-fat group. Fasting insulin fell more sharply in the keto group, too. Over a full day, the keto group showed a 20 percent drop in blood sugar exposure, compared with an 8 percent drop in the other two groups, and insulin exposure fell by 74 percent in the keto group compared with 44 percent and 27 percent in the Mediterranean and low-fat groups.
Prediabetes went away entirely, meaning blood sugar returned to a normal range, in half of the keto group, compared with 29 percent of the Mediterranean group and just 7 percent of the low-fat group.
Not everything favored the keto diet, however. Levels of “bad” cholesterol (LDL) and a similar marker that tracks cholesterol-carrying particles in the blood did not differ meaningfully between the three diets. Blood pressure improvements were also similar across groups. Even though the keto diet in this study contained a lot of saturated fat, it didn’t produce worse cholesterol numbers than the other diets, and it actually showed advantages in a few markers linked to heart disease risk, including certain fat particles in the blood and a marker tied to blood clot risk.
Researchers said the shift in insulin and glucagon could help explain some of the liver benefits. On the ketogenic diet, a hormone called glucagon rose while insulin fell sharply, changing the balance between the two. That shift is believed to slow the liver’s own fat production and push it to burn fat instead of storing it. On the higher-carbohydrate diets, that hormone shift didn’t happen the same way.
