Search for whether these medicines cause hair loss and the results split into two camps: pages that say yes and cite forum posts, and pages that say it is really just the weight loss and cite nothing. The FDA labels answer it more precisely than either, and the answer is stranger than both.
Every figure below was read on 2026-09-10 from labels published through openFDA: Zepbound effective 2026-08-28, Wegovy effective 2024-04-23, Mounjaro effective 2026-07-29, Ozempic effective 2026-01-30, Saxenda effective 2026-02-25. This page reports what those documents record and does not advise anyone on treatment; that belongs with a prescriber, as our editorial standards set out.
The same molecule, counted twice and not counted twice
Tirzepatide is sold as Zepbound for weight management and as Mounjaro for type 2 diabetes. Semaglutide is sold as Wegovy for weight management and as Ozempic for type 2 diabetes. Same molecules, same manufacturers, four separate labels.
| Label | Molecule | Indication | How hair loss appears |
|---|---|---|---|
| Zepbound | tirzepatide | weight management | Counted: 1% placebo, 5% / 4% / 5% at 5, 10 and 15 mg, plus its own labelled subsection |
| Wegovy | semaglutide | weight management | Counted: 1% placebo, 3% at 2.4 mg in adults; 0% placebo, 4% in adolescents |
| Mounjaro | tirzepatide | type 2 diabetes | Postmarketing only, listed as alopecia, no frequency |
| Ozempic | semaglutide | type 2 diabetes | Postmarketing only, listed as alopecia, no frequency |
| Saxenda | liraglutide | weight management | Postmarketing only, listed as alopecia, no frequency |
The two columns that matter are the last two rows against the first two. A postmarketing entry and a counted adverse reaction are not the same claim. A counted reaction has a number, a denominator, and a placebo column standing next to it. A postmarketing entry means reports arrived after approval and the manufacturer added the term to a list; there is no denominator, no comparison group, and no way to know whether it happens more often than it would have anyway.
So a reader who checks the Ozempic label for hair loss finds a word on a list. A reader who checks the Wegovy label for the same molecule finds 3% against 1%.
Why the split is probably not arbitrary
The obvious suspicion is that this is a paperwork artefact — the diabetes labels are older, or the trials were smaller, or nobody asked. The Zepbound label undercuts that suspicion in one sentence, and it is a sentence the manufacturer had no commercial reason to write:
Hair loss adverse reactions in ZEPBOUND-treated patients were associated with weight reduction.
That is Lilly's own reading of its own trial data, printed in the approved label. It says the reaction tracks the weight change rather than the drug as such — which is exactly what would produce the pattern in the table above, because the weight-management products are prescribed at doses that produce far larger reductions than the diabetes ones.
Saxenda is the test of that idea, and it is the reason it is in the table. Saxenda is liraglutide, a weight-management product, in the same drug class, with a weight-management indication — but it produces substantially smaller weight reductions than Wegovy or Zepbound. If the driver were the indication, or the fact of being a weight-loss drug, Saxenda's label would count hair loss. It does not. Alopecia sits in its postmarketing list beside cutaneous amyloidosis, with no frequency attached.
Three products with a weight indication, and the two that produce the largest reductions are the two that count it. That is consistent with the manufacturer's stated association, and it is as far as the published documents allow anyone to go. It is not proof. These are different molecules from different companies with different trial programmes, and nobody has run the study that would separate weight change from drug exposure. Telogen effluvium — shedding that follows a physiological stress such as rapid weight loss, surgery or illness — is a well-described phenomenon that long predates this drug class, but the labels do not name it and this page will not put a mechanism in their mouths.
The sex difference, and the thing hiding inside it
The Zepbound label pools its two main trials and reports hair loss by sex:
| Female | Male | |
|---|---|---|
| Zepbound | 7.1% | 0.5% |
| Placebo | 1.3% | 0% |
Fourteen times as many women as men reported it on the drug. That is the number that gets quoted.
The number that does not get quoted is the placebo row. The imbalance is present there too — 1.3% against 0% — among people who received no tirzepatide at all. Some part of a fourteen-fold gap is therefore about who notices shedding, who considers it worth reporting to a trial investigator, and who has hair long enough to see it come out. The treatment gap is far larger than the placebo gap, so reporting behaviour is very unlikely to be the whole of it. But a reader comparing 7.1% against 0.5% and concluding the drug affects women's hair fourteen times more is reading past the control column, which is the same mistake as reading a nausea rate without its placebo.
What almost nobody who quit was quitting over
One line in the Zepbound label deserves more attention than the incidence figures:
No ZEPBOUND-treated patients and one placebo-treated patient discontinued study treatment due to hair loss.
Across two placebo-controlled trials with 2,519 tirzepatide-treated participants, the number who left the study because of hair loss was zero. The only person who did was on placebo.
That does not make the reaction trivial to the people who had it, and a trial is not the world: participants in a supervised weight-management study finishing a defined protocol are not in the same position as someone paying monthly for an indefinite prescription. But it is the closest thing in either label to a measure of how the reaction weighed against the benefit for the people experiencing it, and the answer in that setting was that it did not tip the balance for anyone.
The gap in the record
Here is what the labels do not contain, on either product: how long the hair loss lasted, whether it resolved during treatment, whether it resolved after stopping, whether it recurred on rechallenge, whether it was diffuse shedding or patterned, and whether anything affected it.
And here is why those gaps have not been filled. Searching ClinicalTrials.gov on 2026-09-10 for registered studies matching semaglutide and alopecia returns zero. Not a small number — none.
Put beside the rest of this beat, that absence is the finding. Nine registered studies are currently testing drugs designed to counteract the muscle loss that accompanies this class, enrolling more than 2,400 participants between them, as our muscle loss page documents. A reaction that appears in two approved labels with a measured incidence, a dose breakdown and a sex breakdown has attracted no registered study at all. The industry has decided which of the two body-composition side effects is worth spending money on, and hair is not it.
Anyone searching for how long this lasts is not finding an answer because the answer has not been produced. That is a more useful thing to know than another page of anecdotes.
Sources
- FDA label for Zepbound (tirzepatide) injection, effective 2026-08-28, via the openFDA drug label endpoint, read 2026-09-10 — incidence table, hair loss subsection, sex breakdown, discontinuation line.
- FDA label for Wegovy (semaglutide) injection, effective 2024-04-23, same source and date — adult and adolescent incidence tables.
- FDA labels for Mounjaro (effective 2026-07-29), Ozempic (effective 2026-01-30) and Saxenda (effective 2026-02-25), same source and date — postmarketing sections.
- ClinicalTrials.gov API, study count for the term semaglutide and alopecia, queried 2026-09-10, with a control query of known scale run in the same session.