Type 2 diabetes is usually described as a problem of blood sugar. That is accurate, but it misses where much of the problem actually lives: in the muscles. When you eat, your skeletal muscles take up most of the glucose from your blood. When they stop responding properly to insulin — the hormone that tells them to absorb that sugar — blood glucose stays high. Scientists have known this for decades, but not exactly why muscle stops listening to insulin.
A study published in Nature Communications on 19 October 2026 offers a new and surprisingly specific answer. Led by senior author Dr. John P. Kirwan of Pennington Biomedical Research Center and the OU Health Harold Hamm Diabetes Center, with first author Elizabeth Heintz, a doctoral candidate at Pennington, it points to a little-known protein called Miro1. In people with type 2 diabetes, Miro1 builds up in skeletal muscle. In laboratory and animal models, removing it restored the muscle's response to insulin. And in a randomised trial in people, exercise lowered Miro1 levels while improving insulin sensitivity. No Miro1 drug exists yet, but the finding connects muscle, mitochondria and exercise into a single, testable story.
Why muscle matters so much in type 2 diabetes
To understand why this matters, start with what insulin resistance actually is. Insulin is a hormone made by the pancreas. After a meal, it acts like a key: it signals cells, especially muscle cells, to open their doors and take glucose in from the blood. In type 2 diabetes, those doors open less and less readily. The pancreas compensates at first by making more insulin, but over time it cannot keep up, and blood sugar climbs.
Muscle is the central player here in a way that surprises most people. Skeletal muscle accounts for roughly 85 percent of the body's glucose disposal at rest, according to the new paper. When insulin resistance sets in, the largest glucose sink in the body is the one that has stopped working. That is why researchers call skeletal muscle insulin resistance a primary defect in type 2 diabetes, not a side effect. It is also why exercise is so powerful: working muscle takes up glucose even when insulin signalling is impaired, and regular training improves the signalling itself.
Yet the molecular chain of events inside an insulin-resistant muscle cell has remained only partly mapped. Suspects have included fat inside muscle cells, inflammation, and failing mitochondria, the tiny structures that produce a cell's energy. The new study adds a fresh suspect at the intersection of mitochondria and insulin signalling.
Meet Miro1: a protein on the surface of mitochondria
Miro1, short for Mitochondrial Rho GTPase 1, is a protein anchored to the outer membrane of mitochondria. It is an old protein in evolutionary terms, found in organisms from yeast to humans. Inside cells, Miro1 helps control mitochondrial dynamics — the splitting, fusing and moving of mitochondria that keeps the cell's energy network healthy — including anchoring a protein called DRP1, which drives mitochondrial fission.
What makes Miro1 newly interesting is its relationship with AKT, one of the main messengers in the insulin signalling chain. When insulin binds to a muscle cell, AKT is activated, helping move glucose transporters — molecular doors called GLUT4 — to the cell surface so glucose can enter. In healthy muscle, insulin promotes an interaction between AKT and Miro1 at the mitochondrial surface. In diabetes, that interaction breaks down: AKT signalling at the mitochondria becomes defective, and Miro1 accumulates instead of being kept in check, leaving a muscle cell whose mitochondria are out of balance and whose response to insulin is blunted.
What the study actually found
The team built its case in stages, moving from people to animals to cells — the strongest evidence a single study can offer short of a drug trial. First came a cross-sectional clinical study. The team took fasted skeletal muscle biopsies from three groups — healthy-weight people, people with obesity but without diabetes, and people with obesity and type 2 diabetes — and measured insulin sensitivity with the gold-standard hyperinsulinaemic-euglycaemic clamp, which infuses insulin while holding blood sugar steady.
The pattern was striking. Miro1 accumulated in the skeletal muscle mitochondria of the patients with type 2 diabetes, but not in the people with obesity alone. The more Miro1 a person's muscle contained, the worse their insulin sensitivity tended to be — a modest but statistically significant negative correlation, driven largely by the diabetes group. The same accumulation showed up in mouse models of type 2 diabetes.
Association, however, is not causation. A protein can pile up in sick tissue simply because the tissue is sick. So the team tested whether Miro1 actually causes the problem. In mice engineered so that Miro1 was deleted specifically from skeletal muscle, and in muscle cells grown in the laboratory, removing the protein improved the muscle's oxidative capacity — its ability to burn fuel and produce energy — and restored insulin sensitivity. As Kirwan put it in the University of Oklahoma's announcement, the team showed "that Miro1 accumulation impairs the muscle's response to insulin, that removing it causally restores that response in our models, and that exercise reverses its buildup in people."
There are caveats. The human part of the study was cross-sectional, a snapshot in time, so it cannot by itself show that Miro1 rises before diabetes develops. The team also notes that interfering with Miro1 could affect muscle contraction. Prospective studies following people as they move from obesity to overt diabetes are still needed.
The exercise connection: a familiar prescription gains a mechanism
Perhaps the most immediately interesting part of the study is what it says about exercise. Doctors have recommended physical activity for type 2 diabetes for generations, and the evidence that it works is overwhelming. But the precise molecular reasons it works are still being worked out. This study adds a new one.
The team ran a randomised controlled trial, registered as NCT02977442, in sedentary adults with obesity and type 2 diabetes. One group received standard medical care; the other completed a 12-week supervised aerobic exercise programme — five sessions a week at moderate-to-vigorous intensity. After the programme, the exercisers had lower Miro1 levels in their skeletal muscle, better insulin sensitivity, more GLUT4 transporters moving to the muscle cell surface, and improved muscle energy-producing capacity. The control group showed no such changes.
Be precise about what this proves. The trial shows that exercise lowers Miro1 and improves insulin sensitivity at the same time. It does not prove the drop in Miro1 is the reason sensitivity improved — exercise changes dozens of things in muscle at once. But the same protein that accumulates in diabetes, and whose removal restores insulin response in models, also falls when people exercise and get better, making Miro1 a plausible part of the explanation.
That picture also fits a broader shift in diabetes care, where the biggest advances have been new medicines — most visibly the GLP-1 drugs. Our report on how retatrutide compared with Ozempic in a head-to-head diabetes trial shows how fast that drug race is moving, and our coverage of what happens to weight after GLP-1 drugs are stopped shows its limits. The Miro1 study points in a different direction: toward the muscle itself, and toward treatments that might one day restore the body's own insulin response rather than working around it.
What this could mean — and what it does not mean
The honest version of this story has two halves. The first is genuinely exciting. Insulin resistance has been described as "frustratingly unclear" in its molecular causes — Kirwan's own words — and Miro1 is a new, specific piece of the puzzle. It sits on the outside of mitochondria, making it more accessible to a future drug than a protein buried deep inside the cell. It has a clear mechanism, a clear link to insulin signalling through AKT, and evidence from humans, mice and cells pointing in the same direction. The research was funded by five institutes of the US National Institutes of Health, including the National Institute of Diabetes and Digestive and Kidney Diseases and the National Institute on Aging.
The second half is the caution. No Miro1 drug exists, and none is in human trials. The causal evidence — removing Miro1 restores insulin response — comes from mice and cells, not people. The human evidence shows that Miro1 is elevated in diabetes and falls with exercise, but that is an association, and it will take years to learn whether lowering Miro1 in people is safe and effective. Because Miro1 helps mitochondria move and divide, interfering with it could affect how muscles contract. Any future treatment would have to help insulin signalling without disrupting the muscle's basic energy machinery — a narrow path many promising targets have failed to walk.
What patients can take from this today
For anyone living with type 2 diabetes, the practical message of this study is not "wait for a Miro1 drug" — it is the opposite. The one intervention already proven to lower Miro1 and improve insulin sensitivity in people is exercise, and it is available now. The trial used a supervised programme of five aerobic sessions a week, but the broader evidence is consistent: regular movement, at whatever level a person can sustain, improves blood sugar control. Small, everyday factors matter too, as our look at cleaner bedroom air and steadier blood sugar showed — the environment around us shapes metabolic health in ways researchers are only beginning to map.
Nobody should change their treatment, start a new exercise regime, or stop a medicine because of a single study — especially one whose drug implications are years away. Exercise programmes need to fit a person's health and medical history, so people with diabetes should talk to their doctor before making significant changes. What this research offers right now is understanding: a clearer picture of why muscle stops responding to insulin, and why the advice to keep moving is grounded in real biology rather than vague encouragement. When a walk lowers your blood sugar, something measurable is happening inside your muscle cells — and Miro1 may be part of the reason why.
Next, researchers must follow people over time to see whether Miro1 rises before diabetes appears, test whether it can be lowered safely, and learn whether its levels could one day identify who might benefit most from a treatment aimed at it. Each question will take years. But the first step — identifying the protein at all — is done.
This article is for general information only and is not medical advice. Type 2 diabetes is a serious condition, and any decisions about exercise, diet or medication should be made with a qualified healthcare professional who knows your individual circumstances.