GLP-1 Hair Loss and Muscle Loss: What Was Measured

Shedding hair and losing lean tissue are both real, both show up in the trial records, and both are mostly consequences of dropping weight fast rather than of the drug acting on hair or on muscle. That is the reassuring half. The unreassuring half is that it changes nothing about what is happening to you: a mechanism you did not expect is still an outcome you have. Neither effect is a reason to change a dose on your own, and the lean-tissue question in particular is unsettled in ways the confident posts about it do not admit.

Hair: the labels put a number on it

This is not an internet rumor, and it does not need to be treated as one. Hair loss appears in the counted adverse reaction tables of both obesity labels. In the Wegovy adult weight-management trials it was reported by 3.3% of people on the 2.4 mg injection against 1% on placebo; at the newer 7.2 mg strength the figures were 5.8% against 1.0%.^[1] Zepbound's pooled trials put it at 1% on placebo and between 4% and 5% across the three tirzepatide doses.^[2]

Break those out by sex and the picture sharpens considerably. Wegovy's own detail section reports 4% among women against 0.9% among men at 2.4 mg, and 8.4% against 0.2% at 7.2 mg.^[1] Zepbound reports 7.1% among women against 0.5% among men, with 1.3% and zero respectively on placebo.^[2] An averaged headline number understates this substantially for one group and overstates it for the other.

The two diabetes-indication labels handle it differently, which is worth knowing if you are comparing pages. Mounjaro and Ozempic list alopecia only in their postmarketing sections, where reports arrive voluntarily from a population of unknown size and no rate can be calculated at all.^[3][4] Absence of a percentage there is a reporting artifact, not evidence of a lower risk.

Telogen effluvium, and why it shows up late

The pattern being described has a name. Telogen effluvium is diffuse shedding that happens when an unusual share of follicles are pushed out of their growth phase and into their resting phase at once. Follicles normally cycle through a growth phase lasting two to five years, which holds about 90% of scalp hair; a brief transition of three to six weeks; and a resting phase of three to five months, holding about 10%.^[5] Push a batch of follicles across that boundary early and they all shed together at the end of the resting phase.

Which is why the timing feels wrong to almost everyone it happens to. Shedding typically begins two to three months after whatever triggered it, and drug-associated telogen shedding tends to start around twelve weeks in.^[5] By then the fastest weight loss is often behind you and the natural suspicion falls on whatever changed most recently. The follicles are reporting on a decision made a season ago.

Sudden weight loss is a named precipitant in the dermatology literature, alongside childbirth, surgery, fever and severe illness.^[6] So are caloric restriction and marked protein deficiency.^[5] None of those involve a receptor in a hair follicle. They involve a body that has abruptly reprioritized.

What the hair studies actually are

Here is where we have to be careful, because the volume of writing on this subject has badly outrun the volume of evidence under it. A 2026 systematic review screened 133 studies and found 24 that qualified as primary research on GLP-1 medication and hair loss.^[7] Its findings line up with the labels: semaglutide and tirzepatide carried the highest reported incidence and the strongest pharmacovigilance signals, telogen effluvium and androgenetic alopecia were the predominant subtypes where a subtype was recorded at all, women appeared disproportionately affected, and tirzepatide — the agent producing the most weight loss — was the one most often linked to telogen effluvium specifically.^[7] The semaglutide signal looked dose-dependent, with doses below 2 mg weekly rarely implicated.

That review's own conclusion is that causality is not established and that prospective randomized work is needed. A separate scoping review is blunter about why. It found nine qualifying studies, noted that most lacked dermatological diagnostic confirmation, and reported that only one described the clinical pattern of the hair loss it counted.^[8] It also notes more than a thousand spontaneous reports in the FDA's adverse event reporting system, which is a signal-generating database rather than a rate.

Lean mass: the words are doing the arguing

Switch to the second question and the first problem is vocabulary. “Muscle loss” and “lean mass loss” get used as if they were synonyms in almost every discussion of these drugs. They are not, and the gap between them is where a lot of alarming arithmetic gets manufactured.

• Lean soft tissue is what a DXA scanner actually measures: everything that is neither fat nor bone. Skeletal muscle, organs, connective tissue, and the water in all of it.
• Fat-free mass is lean soft tissue plus bone mineral content. It is a different number from lean soft tissue, and in one calorie-restriction study bone mineral content rose while weight fell, so the drop in fat-free mass came out smaller than the drop in lean soft tissue.^[11]
• Skeletal muscle is the thing people actually mean, and measuring it properly takes computed tomography or magnetic resonance imaging rather than DXA.
• Muscle function — strength, gait speed, the ability to get out of a chair — is a fourth thing again, and it is the one that determines whether any of the above matters to a life.

The distinction is not pedantry. Lean soft tissue is mostly water by weight, so a scan taken early in a calorie deficit records fluid shifts and glycogen depletion as lean tissue lost, when very little of it is protein. Anyone quoting a lean-mass figure from a short study without saying which of the four things above they measured is not giving you information you can use.

What the body-composition substudies measured

The cleanest single dataset comes from SURMOUNT-1, where 160 of the 2,539 participants were scanned by DXA at the start and again at week 72.^[12] Total weight in the tirzepatide arm came down 21.3%. Fat mass came down 33.9%, lean mass 10.9%. The placebo arm registered 5.3%, 8.2% and 2.6% on those same three measures.

Now the number that matters. Roughly three quarters of what came off was fat and roughly one quarter was lean tissue — and that ratio was the same in the placebo group.^[12] It held across sex, across age bands, and across tertiles of how much people lost. The drug moved how much weight went. It did not detectably move what the weight was made of.

The labels make the same point in a line each. Under pharmacodynamics, Wegovy and Zepbound both say the drug brings weight down with more of the reduction coming from fat than from lean tissue.^[1][2] That is a claim about proportion, and the substudy is what it rests on.

Where the picture gets less tidy

A 2026 systematic review in Annals of Internal Medicine pulled together 35 studies of incretin therapies and body composition, and it is the most useful thing published on this question so far — partly because it declines to reassure.^[13] The studies had a median duration of 26 weeks and a median of 78 participants. Only ten of the 35 had prespecified body composition as a primary outcome. Fewer than half were rated at low risk of bias.

Against a prespecified benchmark of about 25% of weight loss for DXA- or impedance-measured indices, the median across incretin groups was 28.3% once every measurement method was counted together, with 65% above the benchmark. Among the studies that used DXA or impedance alone, the median was about 29%, and 67% came in above it.^[13] That is a worse result than SURMOUNT-1's clean 25%, and it deserves to be reported rather than buried.

Two things in the same review pull the other way, and both belong in the same paragraph as the finding above. Nearly half of the non-drug interventions that produced weight loss also exceeded their benchmarks, which is the strongest available evidence that this travels with weight loss rather than with incretins.^[13] And not one of the 35 studies reported an objective measure of physical function.^[13] Nobody measured whether anyone got weaker. That is a striking gap in a literature this large.

It is also why the honest answer to “does semaglutide cause muscle loss” is: losing weight causes lean tissue loss, these drugs cause a great deal of weight loss, and the consequences for strength and function have not been measured. The supplement most often reached for in response is creatine, and the evidence for creatine on a GLP-1 has the same shape as the problem it is meant to solve: no trial has tested it in anyone taking one.

The pooled answer, and the comparison that reframes it

In 2026 a systematic review pooled 20 randomized trials covering 15,782 participants, restricted to trials that measured body composition by DXA or MRI, and asked what share of the weight lost was lean mass.^[16] It is the largest answer available, and it is more useful than any single substudy.

Read the last row against the first three. Losing weight through diet and exercise alone cost a comparable proportion of lean mass — 26.2%, sitting between tirzepatide and liraglutide, and below semaglutide.^[16] The implication is not that lean mass does not matter. It is that shedding some of it is what losing weight does, by any method, and that a claim framing this as something these drugs do to you needs to explain why the lifestyle arm looks the same.

The population where the question is sharpest

For a 38-year-old with a functioning strength reserve, a proportional lean-mass reduction alongside a large fat reduction is not obviously a bad trade. For a 74-year-old it might be. Sarcopenic obesity — high body fat combined with low muscle mass — is common in older adults, and a 2025 review argues that the risk in this group is compounded by something structural rather than pharmacological: a large share of people stop these drugs within a year and a substantial share restart them, and repeated weight cycling can regain fat faster than it regains muscle.^[14]

That is a review article setting out a concern, not a trial reporting an outcome, and we are labeling it as one. It is the clearest statement we found of a question the trial literature has not answered, and it is a reasonable thing to raise with a prescriber if you are older, if you are already frail, or if this is not your first course of one of these drugs.

What we are not going to tell you

What we can report, attributed, is that the trials never tested these drugs on their own. Both obesity labels indicate them for use alongside a calorie-reduced diet and more physical activity.^[1][2] The pivotal semaglutide trial went further and gave a lifestyle intervention to both arms, drug and placebo.^[15] The weight-loss figures everyone quotes were produced under those conditions. Whether your own program includes any of that is a question for whoever prescribed it.

Related reading: how long GLP-1 side effects last, pancreatitis and gallbladder risk, what is known about tirzepatide over the long term, and side effects reported on compounded tirzepatide. Sellers we track, each with a dated price, are on the compounded semaglutide board.