Not all GLP-1 drugs carry the same hair-loss and hormone signals, six-drug data show

GLP-1 receptor agonists are often discussed as a single drug class, but a pharmacovigilance study published this month in Diabetes/Metabolism Research and Reviews finds meaningfully different safety signals across six individual drugs in the class, specifically for hair loss and reproductive or hormonal adverse events [s1].

The design

Researchers analyzed US FDA Adverse Event Reporting System (FAERS) data from the second quarter of 2022 through the second quarter of 2025, covering six GLP-1 receptor agonists: exenatide, lixisenatide, liraglutide, dulaglutide, semaglutide, and tirzepatide [s1]. Disproportionality analysis — comparing how often each drug appears alongside a given adverse event relative to statistical expectation — was conducted using crude and adjusted reporting odds ratios, with logistic regression controlling for potential confounding factors [s1]. Sensitivity analyses used positive and negative control drugs to validate that the signals detected were robust and specific [s1].

What it found

Across the dataset, researchers identified 1,276 alopecia-related and 759 reproductive or endocrine-related adverse event reports [s1]. Semaglutide showed significant positive associations with alopecia (adjusted reporting odds ratio 1.23, 95% CI 1.11–1.35) and with reproductive or hormonal disorders, including a striking signal for polycystic ovary syndrome (adjusted ROR 6.59, 95% CI 3.73–11.64) and for menstrual abnormalities [s1]. Dulaglutide and tirzepatide, by contrast, showed negative associations — meaning lower reporting odds than expected — for several reproductive outcomes, including dysmenorrhea, amenorrhea, and heavy menstrual bleeding [s1].

Reading the PCOS signal specifically

A sixfold reporting odds ratio for polycystic ovary syndrome associated with semaglutide is the most striking single number in this study, but it needs careful interpretation. PCOS is a chronic hormonal condition, not something that develops acutely after starting a medication in the way a drug-induced rash might — so a pharmacovigilance signal here more plausibly reflects either detection of pre-existing, previously undiagnosed PCOS being newly reported to prescribers once a patient started semaglutide (since PCOS and obesity frequently co-occur, and semaglutide is widely prescribed for weight management), rather than semaglutide itself causing new-onset PCOS, though the study's FAERS-based design cannot distinguish between these explanations.

What this does and doesn't establish

The study's authors describe their findings as identifying "agent-specific differences" that argue for personalized drug selection within the GLP-1 class, rather than treating the drugs as interchangeable [s1]. That's a reasonable takeaway given the sensitivity analyses using control drugs, which are a meaningful methodological strength — they help confirm the semaglutide signals aren't simply an artifact of semaglutide's much higher prescription volume relative to the other drugs studied.

Still, FAERS-based pharmacovigilance data carries the limitations inherent to any spontaneous-reporting system: reports are voluntary, not systematically collected from everyone taking each drug, and can't establish true incidence rates or prove that a drug caused a reported event rather than merely preceding it. The study also can't fully rule out reporting bias — if hair loss or PCOS has received more media attention in connection with semaglutide specifically (its brand names, Ozempic and Wegovy, are the most publicly recognized GLP-1 drugs), patients and clinicians might be more inclined to report and attribute related symptoms to it than to a less-discussed drug like dulaglutide.

Why this matters for prescribing

If the pattern found here holds up under further study, it would offer a more granular basis for choosing among GLP-1 drugs based on a patient's specific risk profile — for instance, favoring dulaglutide or tirzepatide over semaglutide in a patient with particular concern about hair loss or reproductive hormone effects — rather than treating the drug class as a single interchangeable option, which is closer to current prescribing practice.

What to watch

Whether prospective studies or clinical trial safety data corroborate these pharmacovigilance signals, and whether the mechanism behind semaglutide's specific alopecia and reproductive-hormone signals — as opposed to the other drugs studied — is ever established. This article is not medical advice.

Sources

1. Not All GLP-1 Receptor Agonists Are Alike: Real-World Evidence of Differential Endocrine and Dermatologic Safety — Diabetes/Metabolism Research and Reviews, 1 May 2026