The Muscle Question Was Studied in the Wrong Mice: Semaglutide Hits Aged Muscle Differently in Males and Females
RapAdmin
Semaglutide produces dramatic weight loss, but how much of that loss comes from functional muscle remains contested, and nearly all rodent work to date has used young males. Researchers at the University of Utah treated 21-month-old obese mice, roughly equivalent to humans in their sixties, with daily semaglutide for three weeks. Fat loss was large in both sexes. The muscle picture split sharply by sex. Aged females lost up to 14 percent of quadriceps mass, a proportionally greater loss than their whole-body lean mass decline, yet their force production per unit area held steady. Aged males preserved muscle mass almost entirely but lost 19 percent of maximal force in one fast-twitch muscle. Critically, obesity did not inflate muscle mass in these older animals the way it does in young ones, so any loss dropped them below age-matched lean controls.
For two years the argument about GLP-1 drugs and muscle has run in circles. Clinical trials report that a quarter to forty percent of the weight lost on semaglutide is lean mass, which sounds alarming. Skeptics counter that lean mass measured by DEXA is a crude proxy that lumps skeletal muscle together with liver, gut, spleen, and body water, so a falling number does not necessarily mean a shrinking quadriceps. Mouse studies, where investigators can simply dissect and weigh each muscle, mostly supported the skeptics. The problem is that those mice were young.
This group, which published one of the reassuring young-mouse papers last year, went back and repeated the experiment in old animals. Male and female C57BL/6J mice were fattened on a high-fat diet from eighteen months of age, then given daily semaglutide at twenty-one months, an age the authors equate to a human in their sixties and deliberately chosen to sit before the onset of frank sarcopenia in this strain. The design was otherwise identical to the young-mouse study, which makes the comparison unusually clean.
Fat loss was not the issue. It was large, dose dependent, and similar in both sexes. What changed with age was where the rest of the weight came from, and whether there was any cushion underneath.
In young obese mice, high-fat feeding itself builds muscle. The animals are carrying more load and eating more calories, so their muscles grow. When semaglutide then trimmed a few percent off, the muscles simply returned to the size of lean controls, and nothing looked lost. In these aged mice that padding was absent. High-fat feeding did not enlarge a single limb muscle. So when semaglutide took mass off, the animals ended up below their age-matched lean counterparts rather than back at par. That single observation is the paper’s most consequential point, and it does not depend on any sex comparison.
The sex findings are the attention-grabbing part, and they run in opposite directions. Aged females showed the muscle wasting: quadriceps down 14.2 percent, gastrocnemius down 9.4 percent, with individual fiber cross-sectional areas shifted smaller. Several muscles lost a larger percentage than whole-body lean mass did, which inverts the pattern seen in young animals. Yet when those muscles were mounted on a force transducer, they pulled as hard per unit area as controls. Aged males showed the mirror image: muscle masses essentially intact, but maximal specific force in the extensor digitorum longus down 19.2 percent, a strength deficit in a muscle that had not shrunk at all.
That dissociation is biologically interesting and statistically fragile. Smaller muscles that work fine, and normal-sized muscles that do not, are both plausible aging phenotypes, but with five to ten animals per group and borderline p values, either result could soften on replication. The authors are appropriately cautious, calling for clinical trials that use magnetic resonance imaging or D3-creatinine dilution and that deliberately enroll older adults. Given that a 7.2 milligram semaglutide dose is now approved, roughly matching the high dose used here, that call is timely.
Actionable Insights
Effect sizes here are large by conventional standards but rest on very small groups, so treat them as signals rather than settled magnitudes.
The strongest signal is about reserve, not drug toxicity. In old mice, obesity no longer inflates muscle mass, so weight loss cuts into baseline rather than into surplus. The practical translation: if you are over sixty and starting a GLP-1, you probably do not have the muscle buffer a forty-year-old has.
Dose appeared to matter in males. At the low dose, male quadriceps and lean mass were barely touched (lean mass down 2.3 percent, Cohen’s d of 0.41, which is a small effect). At the high dose, lean loss tripled to 6.6 percent (d of 1.07, a large effect). Females showed no such dose separation, losing roughly 8.5 percent of lean mass at both doses.
A Cohen’s d of 1.0 means the average treated animal sits a full standard deviation below the average control, outside roughly 84 percent of the control distribution. That is a real difference, not a statistical artifact.
Two monitoring implications. First, track function, not only the scale: the male strength loss occurred in a muscle that had not shrunk, so mass measurements would have missed it entirely. Second, appetite suppression was concentrated in week one and food intake normalized by week two, which identifies the early phase as the window where protein intake and resistance loading most need protection.
Context and Source
- Open Access Paper: Aging exacerbates muscle loss induced by semaglutide treatment in female mice.
- Institution: Center of Metabolic Health, Department of Nutrition and Integrative Physiology, Molecular Medicine Program, Department of Pathology, and Department of Biochemistry, University of Utah, Salt Lake City, Utah, United States.
- Journal: None. bioRxiv is a preprint repository.
- Impact Evaluation: None. bioRxiv is a preprint repository.









