What the boxed warning actually says
Every semaglutide product (Ozempic, Wegovy, Rybelsus) and every tirzepatide product (Mounjaro, Zepbound), plus the long-acting GLP-1s liraglutide and dulaglutide, carries the same FDA boxed warning — the strongest warning the FDA can require. Reading it carefully matters, because the wording is deliberately cautious rather than conclusive.
The Ozempic label (updated 2025) states, word for word: "In rodents, semaglutide causes dose-dependent and treatment-duration-dependent thyroid C-cell tumors at clinically relevant exposures. It is unknown whether OZEMPIC causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans as the human relevance of semaglutide-induced rodent thyroid C-cell tumors has not been determined."
Two phrases carry the weight: "in rodents" and "it is unknown ... in humans." The warning documents a real, reproducible finding in animals and then states plainly that scientists have not established whether the same thing happens in people. It is a precaution built on animal data, not a report of tumors observed in human trial participants.
Why it's based on rodents — and why that may not translate to people
In two-year carcinogenicity studies, semaglutide produced statistically significant increases in thyroid C-cell adenomas and carcinomas in both mice and rats, and the effect grew with higher doses and longer exposure (FDA Ozempic label, Nonclinical Toxicology section). That is a robust animal signal, which is why the FDA acted on it.
The catch is a biological difference between species. Thyroid C-cells in rodents carry far more GLP-1 receptors than human C-cells do; the level of GLP-1 receptor expression in the thyroid is species-specific, and it is much higher in rodents than in humans and other primates (Diabetes Care, 2025; Thyroid, 2025). Because the drug acts on those receptors, a species with many more of them may respond very differently. That is precisely why the FDA labels the human relevance as "not determined" rather than assuming the rodent result carries over.
The FDA also notes that cases of MTC have been reported after liraglutide (another GLP-1) reached the market, but that these post-marketing reports are "insufficient to establish or exclude a causal relationship" between MTC and GLP-1 use in humans.
Who absolutely should not take a GLP-1: the contraindication
The boxed warning translates into one firm rule. GLP-1 medications are contraindicated — meaning they should not be used — in anyone with a personal or family history of medullary thyroid carcinoma (MTC), or anyone with Multiple Endocrine Neoplasia syndrome type 2 (MEN 2), a rare inherited condition that dramatically raises MTC risk (FDA labels for Ozempic, Wegovy, Mounjaro, Zepbound).
This is why reputable telehealth intake forms and clinicians ask about your and your family's history of thyroid cancer before prescribing. It is a screening question, not a formality. If either applies to you, a licensed clinician will steer you toward a different weight or diabetes treatment.
The label also tells clinicians to counsel patients to watch for and report possible symptoms of a thyroid tumor: a lump or mass in the neck, trouble swallowing (dysphagia), shortness of breath (dyspnea), or persistent hoarseness. Noticing any of these is a reason to contact a clinician — not a reason to panic, since these symptoms have many benign causes.
- Absolute contraindication: personal history of medullary thyroid carcinoma (MTC)
- Absolute contraindication: family history of MTC (a close blood relative)
- Absolute contraindication: Multiple Endocrine Neoplasia syndrome type 2 (MEN 2)
- Report to a clinician: a neck lump/mass, trouble swallowing, shortness of breath, or persistent hoarseness
What the human evidence shows so far
Since the warning is based on animals, researchers have spent years looking for a thyroid-cancer signal in actual patients. The reassuring news is that the largest, best-designed studies have not found a clear increase — though none can yet speak to very long-term use.
A nationwide Scandinavian cohort of more than 145,000 GLP-1 users (BMJ, 2024) found no increased overall thyroid cancer risk (hazard ratio 0.93, 95% CI 0.66-1.31), and a neutral, if imprecise, estimate for medullary thyroid cancer specifically (HR 1.19, 95% CI 0.37-3.86). An international multi-country cohort (Thyroid, 2025) and a US retrospective cohort of about 460,000 GLP-1 users (Diabetes Care, 2025, LEGEND-T2DM) likewise found no increased risk of thyroid tumors versus other diabetes drugs. A US Medicare study (BMJ Open Diabetes Research & Care, 2025) reached the same reassuring conclusion for three-year risk.
The evidence is not perfectly uniform. A French nested case-control study (Bezin et al., Diabetes Care, 2023) reported a higher risk (hazard ratio 1.46 for all thyroid cancer; 1.78 for MTC). However, other researchers have flagged likely confounding and detection bias in that analysis — for example, risk appearing too early to plausibly be caused by the drug. After reviewing the evidence, the European Medicines Agency concluded that the available data do not support a causal association. The honest summary: current human data are largely reassuring for short-to-medium-term use, but thyroid cancer is rare and slow-growing, so longer follow-up is still needed.
A different issue: GLP-1s do not treat hypothyroidism
There's a common mix-up worth clearing up. The boxed warning is about C-cell tumors — it has nothing to do with an underactive thyroid (hypothyroidism), Hashimoto's thyroiditis, or being on levothyroxine (Synthroid). GLP-1 medications do not treat hypothyroidism, do not replace thyroid hormone, and are not a thyroid medication in any sense.
Ordinary, well-managed hypothyroidism is not the same as the contraindication. Millions of people take levothyroxine, and having a common thyroid condition like Hashimoto's or a simple goiter is generally not, by itself, the reason GLP-1s are contraindicated — that specific bar is MTC and MEN 2. That said, your full history belongs in a conversation with a licensed clinician, who is the one to decide.
It's also true that significant weight loss can change how much thyroid hormone the body needs, so people on levothyroxine sometimes need their dose re-checked over time. That is a monitoring point for a clinician, not something to adjust on your own.
The levothyroxine timing interaction
Because GLP-1 medications slow gastric emptying (that's part of how they curb appetite), they can affect how oral medications — including levothyroxine — are absorbed. This is a pharmacokinetic interaction, not a claim that GLP-1s harm the thyroid.
The clearest data come from oral semaglutide (Rybelsus). In a controlled study of healthy volunteers, taking levothyroxine alongside oral semaglutide raised total T4 (thyroid hormone) exposure by about 33%, while peak concentration was unaffected; the proposed mechanism is delayed gastric emptying increasing levothyroxine absorption (Hauge et al., 2021; drug interaction monographs). The FDA-approved oral semaglutide labeling advises that clinicians consider monitoring thyroid parameters when the two are used together.
Practically, this is a reason your prescriber may check your TSH after you start or change a GLP-1, and it's why the timing of levothyroxine relative to other medications matters. But how and when to take your thyroid medication — and whether any dose change is needed — is a decision for your prescribing clinician and pharmacist, not something to change based on a web page. Bring both medications up at your next visit.
Putting the risk in perspective (without dismissing it)
The right frame is neither "this drug causes thyroid cancer" nor "the warning is meaningless." The accurate frame is: there is a real animal signal, an undetermined human relevance, a firm contraindication for the small group at genetically higher risk, and human data so far that are largely reassuring but not final.
The FDA itself notes that routine screening for MTC in GLP-1 users — measuring serum calcitonin or doing thyroid ultrasounds — is of "uncertain value" and may lead to unnecessary procedures, because those tests have low specificity against a high background rate of benign thyroid findings. In other words, the current guidance is not to mass-screen everyone on a GLP-1, but to screen out the contraindicated group up front and stay alert to symptoms.
If you're considering a GLP-1 for weight loss or diabetes, the single most useful step is a complete, honest medical intake with a licensed clinician — including your family's thyroid history. Every legitimate GLP-1 telehealth provider builds that screening into the process.