The short answer is yes, a little, and probably not for the reason people assume. A 2026 systematic review and meta-analysis pooled 17 studies covering 1,091,743 patient-exposures and found the odds of any non-scarring alopecia raised at 1.40, with a confidence interval from 1.33 to 1.48 and no measurable heterogeneity between the cohort studies [1].
That is a real signal and a small one. It is also an average taken across very different conditions that happen to share the word “alopecia”, and splitting them apart is where the useful answer turns out to live.
Which kind of hair loss actually goes up
| Type of hair loss | Pooled odds at 12 months | Raised? |
|---|---|---|
| Any non-scarring alopecia | 1.40 (95% CI 1.33–1.48) | Yes |
| Telogen effluvium | 1.76 (95% CI 1.34–2.32) | Yes |
| Androgenetic alopecia | 1.64 (95% CI 1.35–1.99) | Yes |
| Alopecia areata | 1.08 (95% CI 0.87–1.34) | No — interval crosses 1 |
Telogen effluvium is diffuse shedding that begins two to four months after a physiological stress, which is the detail that makes the timing recognizable to anyone it has happened to. Rapid weight loss is one of the classic triggers, alongside illness, surgery and childbirth — the same stress response behind the lean-mass changes in what is lost and what returns. Alopecia areata, by contrast, is autoimmune and patchy, and it has nothing to do with how much weight somebody has lost or how quickly they lost it. The first of those is raised here and the second is not, and that split is the closest thing this literature has to an explanation.
The clue that points at the weight, not the drug
The same matched cohort reports one observation that does more work than any odds ratio in it: among GLP-1 users, the people who developed hair loss had a lower follow-up BMI than the people who did not. The ones who lost more hair were the ones who lost more weight.
That is consistent with telogen effluvium being the mechanism, because on that reading the shedding tracks the speed and the size of the weight change rather than anything about the molecule itself. It is an association found inside an observational cohort rather than a randomized comparison, which makes it a strong hint about the mechanism rather than proof of it. Nobody has run a trial that holds weight loss constant and varies the drug, which is the experiment that would settle it. It is the same reasoning problem this site worked through in separating a drug effect from a weight effect.
Why you will still find sources saying the evidence conflicts
Because until very recently it genuinely did, and the pages saying so were describing the evidence accurately at the time they were written. A 2025 systematic review found only five primary studies covering 2,905 adults, and they disagreed with each other — some reported hair loss as an adverse event, others reported improvement and regrowth [3]. That was an honest description of a thin evidence base. The 2026 meta-analysis is roughly 375 times larger by exposures, which is what changed.
Pages written against that older review are not wrong about what it said, and it would be unfair to read them as careless. They are just describing the state of the field before the larger synthesis landed.
Does it matter which drug you are on?
On current evidence, not much. The matched cohort found higher odds of non-scarring alopecia with both semaglutide and tirzepatide compared with other GLP-1 drugs, and higher odds of telogen effluvium with tirzepatide across every follow-up window it examined [2]. The published summary reports those as directions rather than as odds ratios, so this site will not attach numbers to them.
Since semaglutide and tirzepatide are also the two that take off the most weight — the gap measured in the real-world comparison of the two — a difference between molecules and a difference in weight lost cannot be separated here either.