Glucagon is supposed to tell your liver to burn fat. So why do people with newly diagnosed type 2 diabetes so often have elevated glucagon and fatty liver together, when one should prevent the other? A new 2026 paper in Diabetes Care by Huttasch and colleagues set out to answer that question. doi.org/10.2337/dc25-3077. I’ll describe this study but let me begin with some background to set the stage.
Glucagon: the liver’s on switch
Glucagon is a hormone made by the pancreas, the counterpart to insulin. While insulin lowers blood sugar, glucagon raises it by telling the liver to release stored glucose. In a healthy liver, glucagon also drives fat burning and tells the liver to break down amino acids, stripping their nitrogen for disposal as urea and using the carbon backbone for fuel. Glucagon is the liver’s metabolic “on switch,” the signal that tells it to release glucose, burn fat, and break down amino acids.
From one hormone to two
In type 2 diabetes, glucagon is chronically elevated, pushing glucose out of the liver even when blood sugar is already too high. Before the 1970s, diabetes research had focused almost entirely on insulin. Roger Unger, an endocrinologist at UT Southwestern, changed that by demonstrating that glucagon excess alongside insulin deficiency drives high blood sugar. Unger and Lelio Orci named the idea the bihormonal hypothesis (Lancet, 1975), arguing that diabetes is a two-hormone disease where insulin fails to lower blood sugar while glucagon actively drives it higher. (For a comprehensive history of glucagon research from its discovery to the present, see Wewer Albrechtsen et al., 100 Years of Glucagon and 100 More, Diabetologia 2023.)
The Glucagon–Fatty Liver Paradox
The puzzle was why fatty liver (now called metabolic dysfunction-associated steatotic liver disease, or MASLD) so frequently accompanies this glucagon excess, when glucagon’s job is to prevent fat from building up in the liver.
Retatrutide, a drug that activates the glucagon receptor alongside GLP-1 and GIP receptors, cleared liver fat in a phase 2a trial (Sanyal et al., Nature Medicine, 2024). If a drug that amplifies glucagon’s signal can clear the liver of fat, why can’t the body’s own elevated glucagon do the same?
The amino acid explanation
Nicolai Wewer Albrechtsen and colleagues (including Jens Juul Holst) proposed the liver-alpha cell axis in 2018 (Diabetologia), arguing that fat accumulation in the liver impairs its ability to break down amino acids. The unprocessed amino acids pile up in the blood, reflexively stimulate the pancreas to make more glucagon, and the resulting glucose output creates a self-reinforcing loop. In this model, amino acids are the link between liver fat and glucagon excess.
Where the evidence fell short
The liver-alpha cell axis fit the available data, but the evidence came from people who had lived with diabetes for years, often with advanced liver disease. By that point, so much has changed (medications, weight, worsening insulin resistance, progressive liver damage) that it is impossible to tell whether liver fat is actually driving the glucagon excess or whether both simply got worse together over time.
Testing the theory in new onset disease
To test the theory before years of disease, treatment, and metabolic decline blur the picture, Huttasch et al. enrolled fifty people within 12 months of a type 2 diabetes diagnosis alongside 50 healthy controls matched for age, sex, and body weight. Huttasch’s team used gold-standard MRI for liver fat quantification, insulin clamp studies to directly measure how well the body responds to insulin, and a timed meal test where blood was drawn every 30 minutes to track glucagon alongside amino acids and fatty acids.
What Huttasch found
The central prediction of the liver-alpha cell axis, that amino acids mediate the link between liver fat and glucagon, did not hold. Amino acids, including alanine (the molecule the liver-alpha cell axis relied on), failed to statistically explain the glucagon excess, confirmed across 3,000 statistical simulations, and neither did fatty acids in the blood. Controlling for amino acid levels should have reduced the apparent glucagon-liver fat relationship but it didn’t.
Fatty liver was the independent driver of elevated fasting glucagon, holding even after accounting for diabetes status, insulin resistance, and abdominal fat. The long-standing assumption was that glucagon elevation in type 2 diabetes was a systemic consequence of the disease’s metabolic disruption. But when the researchers separated out who had liver fat and who didn’t, liver fat drove the excess glucagon. Having type 2 diabetes without liver fat did not produce the same glucagon excess.
After eating, the link between liver fat and glucagon existed only in the type 2 diabetes group, not in healthy controls, even healthy controls who also had some liver fat. In a healthy liver, fat accumulation does not appear to alter the post-meal glucagon response. In people newly diagnosed with type 2 diabetes, the more liver fat they had, the higher glucagon climbed after eating. The diabetic liver responds to fat accumulation differently, in a way that amplifies the post-meal glucagon response, and this altered response is already present at diagnosis.
People with both type 2 diabetes and fatty liver had 47% higher post-meal glucagon than those with type 2 diabetes alone. Liver fat amplifies the glucagon excess, and the size of the effect confirms it as an active driver.
A liver that stops listening
If liver fat is driving the glucagon excess rather than amino acids relaying it, the mechanism is more likely to be direct: the liver has become resistant to glucagon’s signal, a concept known as hepatic glucagon resistance. The liver receives glucagon’s message but doesn’t act on it, not burning fat, and not breaking down amino acids. In response to increased hepatic fat, glucagon rises higher.
Caveats
The data does not prove glucagon resistance as a mechanism, only that the pattern is consistent with it. The coexistence of high glucagon and high liver fat in early diabetes may also reflect a broader disruption of how the liver handles energy, with both being downstream consequences of the same underlying process rather than one causing the other.
The study’s key limitation is that all participants were assessed at a single point in time. The findings show that liver fat and glucagon excess travel together in early type 2 diabetes, but cannot prove that liver fat causes the glucagon excess or that one developed before the other. Only longitudinal studies following people from before diagnosis could establish the direction of causation.
The harder question
Even without proof of causation, the simplest explanation for why glucagon and liver fat coexist from the very start of type 2 diabetes is that the liver has already begun ignoring glucagon’s signal. Drugs that amplify glucagon’s signal, including tirzepatide (Mounjaro) and agents like retatrutide, are pushing harder on a system that may already be partially resistant at diagnosis. Whether that limits their effectiveness in people with type 2 diabetes, compared with those without it, is the question this paper raises but cannot yet answer.
Michael Mindrum, MD