Does semaglutide or tirzepatide make you lose muscle, not just fat? Yes. Some do. The real question is how much, and what you did about it while the weight was coming off.
Maria is not one patient. She is three, folded into one person for this piece, because the same conversation happens in my office most weeks. (Composite patient, built from several real visits. No identifying details are real.) She is forty-six. She started semaglutide in April and was down twenty-two pounds by October. Her jeans don’t fit anymore. Neither does the blazer she bought two years ago for a conference. She came in last week, not to celebrate that. She came in because she’d carried a bag of water-softener salt up from her garage and her legs shook the whole way up the stairs. “I’m smaller,” she told me. “I don’t feel stronger. I think I’m actually weaker.” She was right to notice. She wasn’t imagining it.
Here is what the medication is actually doing to a body like hers, past the number on the scale.
The best data on this comes from the drug trials themselves, not a forum post or someone’s before-and-after photo. In the STEP 1 trial’s body-composition substudy, people on semaglutide lost 15.0% of their body weight over 68 weeks, against 3.6% on placebo. Fat mass dropped 19.3%. But total lean body mass — muscle, organ tissue, the water that lives inside cells — dropped too, by 9.7% (Wilding et al., 2021). Run that ratio out and roughly 39% of everything these patients lost was lean mass, not fat. That is not a side effect buried in the fine print. It is in the paper’s own numbers.
Tirzepatide’s data, from the SURMOUNT-1 substudy, tells a similar story with a slightly better ratio. At 72 weeks, body weight fell 21.3%, fat mass fell 33.9%, and lean mass fell 10.9%, against placebo’s 5.3%, 8.2%, and 2.6% (Aronne et al., 2024). Close to three-quarters of the weight lost there was fat. The other quarter was lean tissue. And some separate observational comparisons since then have suggested tirzepatide may cost a bit more lean mass than semaglutide at matched time points. That comes from routine-care records, not a head-to-head trial. I’d treat it as a lead worth watching, not a settled fact.
So. Not a myth. Not nothing, either.
A quarter to nearly forty percent of the weight these medications take off is muscle, fluid, and the tissue that keeps your metabolism and your grip strength where they belong. For someone starting from a healthy reserve of muscle, that trade is worth making for the much larger gain. For someone who was already thin-framed, sedentary, or well past menopause when they started, that same percentage can tip them toward something closer to frailty than the strong, lighter body they pictured on day one.
This is where I stop reading a chart and tell you where I actually watch this play out. I compete in bodybuilding. I’ve cut weight on purpose, under controlled conditions, with training built specifically to hold onto muscle while the scale drops. What Maria went through was the uncontrolled version of that same physiology. Minus the training, minus the extra protein, minus anyone checking her strength along the way. Nobody had told her any of that mattered. Nobody had measured anything but her weight.
What actually protects lean mass here is not complicated, even if it is not easy either. Resistance training, two or three real sessions a week against real weight, not a few bands and some range-of-motion work, is the one intervention with the most consistent support across the body-composition data on these medications. Protein intake has to go up while total calories go down, because a calorie deficit without enough protein tells the body it is safe to burn muscle for fuel instead of fat. And somebody needs to be tracking something other than the number on the scale. A tape measure at the waist and the arm. A grip-strength check, which takes thirty seconds and tells you more than a month of weigh-ins. A bioimpedance reading or a DEXA scan, if one is available. Weight loss with no plan for any of that is flying blind on the exact thing that will decide how this patient feels in ten years, not just how she looks in six months.
None of this is an argument against semaglutide or tirzepatide. The fat-mass reductions documented in both trials are real, and for most patients starting this treatment, the metabolic benefit outweighs the lean-mass cost by a wide margin. It is an argument against starting either one with no plan for the muscle you are also going to lose. A GLP-1 prescription with no protein target and no resistance-training plan attached to it is half a treatment, not a whole one.
Maria is back in a gym now. Twice a week, with a trainer watching her form, not her reflection. Her weight loss slowed down a little once the training went back in. Her legs do not shake on the stairs anymore. That is the actual goal here. Not the number on the scale. What the number was made of.
