Is Obesity Really a Muscle Disease?

In 1989, six healthy young men in Denmark agreed to have one leg exercised while the other sat still. Four hours later, researchers threaded catheters into both femoral veins, infused insulin, and measured how much glucose each leg pulled from the blood. The exercised leg took up more glucose than the resting one, even though both legs belonged to the same man, carried the same amount of muscle, and saw the same insulin at the same concentration (00:00). The only thing that differed was what the tissue had done recently, not how much of it there was.

That result sits at the center of this episode of the Barbell Medicine Podcast, hosted by physician and powerlifter Dr. Jordan Feigenbaum. His target is Muscle Centric Medicine, the framework built by physician Dr. Gabrielle Lyon, who argues that obesity is, in her words, a disease of the muscle, and that people with obesity are simply undermuscled (00:47). Feigenbaum agrees with two of her premises: muscle is the body's largest site of insulin-driven glucose disposal, and muscle insulin resistance shows up years before blood sugar looks abnormal (01:10). What he disputes is the conclusion that follows, that building more muscle mass fixes the underlying problem.

The doors in the wall

Glucose cannot cross a cell membrane on its own. It needs a transporter, a piece of protein sitting in the cell wall that lets it through. Most of a muscle cell's transporters are not hanging in that wall at rest; more than nine in ten sit stored inside the cell, waiting to be installed (05:18). Insulin can install them. So can muscle contraction, through a separate pathway that does not need insulin's permission (05:52). That is what happened to the Danish leg: the same transporters, installed by exercise instead of insulin.

This distinction undoes the arithmetic behind "more muscle equals more glucose disposal." Under a lab test called a clamp, 80 to 90 percent of infused glucose ends up in skeletal muscle (06:39), the figure Lyon's case leans on. But the physiologist who produced that number, endocrinologist Dr. Ralph DeFronzo, also found that under ordinary eating conditions, the gut and liver claim 30 to 40 percent of glucose before muscle ever sees it (07:30). A clamp isolates muscle; breakfast does not. And because training increases the number of transporters a cell owns without necessarily increasing the cell's size (08:16), a muscle can get much better at handling sugar while staying exactly the same size. The effect also expires: insulin sensitivity gained from a single workout was still present at 48 hours but gone by five days (13:40).

"A large well built muscle that hasn't contracted in five days is, by this measure, an untrained one," says Dr. Jordan Feigenbaum [14:02].

The strongest evidence against the muscle-mass theory comes from the Sarcopenia Definitions and Outcomes Consortium, which pooled 18,800 community-dwelling older adults and tested which body-composition measures predicted falls, fractures, mobility loss, and death. Grip strength and gait speed predicted all four outcomes. Lean mass measured by a DEXA scan, the standard imaging test for body composition, predicted none of them (21:20). "Knowing how much lean mass somebody had told you nothing about whether they fall, break something, lose mobility, or die," Feigenbaum says (21:44). The panel that convened to define sarcopenia, a word that literally means muscle loss, ended up recommending that the muscle-mass measurement be removed from its own definition.

Odder still, people with obesity carry more absolute muscle mass than lean people do, and they are stronger in raw force, across ages from adolescence through old age (22:48). What they have less of is muscle quality, meaning force produced per kilogram of tissue, because fat infiltrating the muscle, a condition called myosteatosis, degrades performance without shrinking the tissue (18:33). The population supposedly suffering from a muscle deficiency has, on average, extra muscle. What it has is muscle that works worse.

Location, type, and whether the boxes are moving

Why does fat inside muscle sometimes wreck insulin sensitivity and sometimes not? Endurance athletes carry about as much fat inside their muscle cells as people with type 2 diabetes, yet they are among the most insulin-sensitive people ever measured (28:07). The variable is not how much fat sits there. It is how fast that fat is broken down and rebuilt: athletes turn over their muscle fat about 2.5 times faster than sedentary men (29:22). Feigenbaum's analogy runs through the whole episode: sedentary muscle is boxes sitting untouched on a staircase, trained muscle is boxes being carried up and down all day.

The liver tells a similar story. Among 133 people who had their livers scanned and then clamped, having any liver fat at all raised insulin resistance, but among people who already had liver fat, having 6 percent or 25 percent made no difference (31:46). What seems to matter is the type of fat molecule. Triglyceride, the storage form reported on lab tests, is largely inert. Diacylglycerol, an intermediate step in building triglyceride, is a signaling molecule that can jam the insulin receptor. Block the enzyme that finishes that conversion in a rat's liver and total liver fat drops, yet the animal becomes more insulin resistant, because the half-built molecule piles up and interferes with signaling (32:55). Less fat, worse metabolic health, a result that only makes sense once type, not just amount, enters the picture.

Feigenbaum applies the same lens to GLP-1 drugs, short for glucagon-like peptide-1 receptor agonists, the class that includes semaglutide and tirzepatide. Trial data show people losing roughly 40 percent of their total weight loss as lean mass on DEXA scans (56:49), a figure that has fueled worry about muscle wasting. But a DEXA scan's lean category includes water, glycogen, organ tissue, and roughly 15 percent of what used to be fat tissue itself, none of which is contractile muscle (57:41). In the year-long SEMILEAN trial, people taking semaglutide without exercising still gained about 4.5 kilograms of grip strength (59:49), a functional gain the scan could only register as loss.

"The scan is counting that success as a loss," says Dr. Jordan Feigenbaum [59:02].

None of this argues against lifting weights. Feigenbaum trains and coaches lifters for a living, and he repeats that resistance training lowers mortality, blood pressure, and hemoglobin A1c regardless of whether Muscle Centric Medicine's theory holds up (19:01). His point is narrower: the amount of muscle on a scan is the wrong target. What matters is what the tissue does, where fat ends up, what molecular form it takes, and how fast any of it moves.

That reframing lands on a bigger diagnostic problem. Researchers sorted 40,000 American adults by body mass index, or BMI, and separately checked for real metabolic markers such as blood pressure, blood sugar, and triglycerides. More than 30 percent of people with a normal BMI were metabolically unhealthy, and about 29 percent of people classified with obesity were metabolically fine (80:28). By the study's own estimate, 75 million American adults are mislabeled either way (81:02). Feigenbaum compares obesity, as currently named, to jaundice, a visible sign rather than a mechanism. Hypertension is named for the pressure that does the damage, so doctors treat the pressure. Obesity is named for how a body looks from the outside, and for decades medicine has been treating that reflection instead of what sits behind it.

His proposed replacement is not a number a bathroom scale can give. It is three questions: where is the fat sitting, what molecular form is it in, and how fast is it moving. Muscle mass answers none of them. Muscle function, measured through strength and gait speed, gets closer. Exercise remains the most reliable lever anyone has found, because it improves location, type, and flux at the same time, without requiring a single extra kilogram of tissue.