Keto diet and liver fat: why protein did the heavy lifting

About one in three adults carries extra fat in the liver, and many do not know it. It often shows up as blood markers outside the normal range, such as HbA1c, fasting glucose or apoB. People then ask what to eat beyond simply losing weight. A new trial from Sam Klein's team at Washington University offers a clear answer, and host Simon Hill and a dietitian guest break it down on The Proof.

Why this trial is unusually strong

Most diet studies give advice and hope people follow it. In a meta analysis of low carb trials aiming for under 50 grams of carbohydrate a day, some participants ended up eating up to 180 grams. This study avoided that problem in three ways:

• All food was provided, so adherence was far higher than in free living trials
• Weight loss was matched at about 10 percent, which removes weight change as a confounder
• Biomarkers checked adherence, using urinary nitrogen for protein intake and ketones for carbohydrate restriction

It compared three diets: a ketogenic style diet, a Mediterranean diet and a plant forward diet.

The headline result

The keto style diet, with 23 percent of energy from saturated fat, cut liver fat by 67 percent, more than the other two diets. Headlines called ketogenic the best diet for fatty liver. The guest disagrees. Ketones averaged only about 0.5 mmol per liter, near the bottom of what a true ketogenic diet reaches, which can be 1.5 or higher. A better description is a high protein, low carbohydrate diet, and protein looks like the key driver.

Why protein may protect the liver

No one knows the exact mechanism, but there are clues:

• High protein, low carb meals lower liver fat even without weight loss
• Lower glucose and insulin plus higher glucagon may push the liver to burn fat instead of making it from carbohydrate
• In overfeeding studies with fat or fructose, adding amino acids such as branched chain amino acids protected the liver from fat buildup
• Special amino acid blends lowered liver fat more than equal amounts of regular protein, suggesting some amino acids act as signals

A Norwegian trial of about 200 adults over a year supports this. Protein was about 16 to 17 percent of energy in all three diets, weight loss was similar, and liver fat measured by CT did not differ despite different carbohydrate intakes. That points to protein, not carbs alone.

What about saturated fat?

This single trial cannot tell us much about saturated fat. The protein and carbohydrate effect may be so strong that saturated fat becomes irrelevant here. Other evidence is clearer. In overfeeding and isocaloric trials, polyunsaturated fat gives somewhat less liver fat than saturated fat, though the effect is smaller than protein's.

ApoB actually fell in all groups, including the keto style group. A likely explanation is that weight loss, high protein and carbohydrate control lowered liver fat and triglycerides, so the liver exported fewer VLDL and LDL particles. That would offset saturated fat's tendency to down regulate the LDL receptor. The guest still sees no nutritional reason to add butter, cream or coconut oil, and would favor polyunsaturated and monounsaturated fats.

What the trial did not show

• Vascular function, such as flow mediated dilation, was not reported
• Glucose tolerance did not differ between diets, and weight loss itself probably explains most of the improvement
• PREDIMED showed cardiovascular benefits from a Mediterranean diet with nuts through blood pressure and vessel function

Liver fat, insulin resistance and fiber

Mendelian randomization studies suggest liver fat contributes to insulin resistance, and insulin resistance then makes more liver fat, a vicious cycle. Whether or not it is the cause, extra liver fat is a sign of unhealthy metabolism, much like a larger waist, and it can progress to fibrosis and raise the risk of some liver cancers. In prediabetes, first phase insulin secretion is already reduced by 30 to 60 percent, and when people lose about 10 percent of their weight, glucose control often normalizes. Fiber helps as well. Viscous and fermentable fibers modestly lower liver fat and improve glucose tolerance by slowing absorption and producing short chain fatty acids, although protein still has the bigger effect on liver fat.

How to eat this way

• Protein: about 35 grams per meal is a good start, or raise protein by about 30 percent
• Sources: chicken, fish, tofu and legumes, which also add phytonutrients
• Carbohydrates: keep portions small, and a little white rice or potato is fine alongside protein and legumes
• Fats: use olive, sunflower or rapeseed oil, and eat about 25 grams of nuts a day
• Vegetables and fiber: plenty, with about 30 grams of fiber a day
• Seed oils: human studies do not show that sunflower or rapeseed oil increase inflammation

Monitor weight, apoB, HbA1c and fasting glucose with your doctor if you can.

Conclusion

Do not read this as proof that keto is best. High protein and lower carbohydrates reduced liver fat, and pairing that with heart healthy fats and plants gives the best of each diet.