GLP-1 side effects: the ones the trials measured
GLP-1 receptor agonists are among the most heavily studied drugs in modern medicine, which means the side effect question can be answered with trial numbers and label text rather than forum anecdotes, so this article walks through what was actually measured, what is rare but serious, and what changed on the labels in 2026.
The short answer
- Gastrointestinal effects dominate: nausea, diarrhoea, vomiting, constipation and abdominal pain. In the STEP weight-loss trials these were mostly mild to moderate, dose-related, and clustered around escalation rather than maintenance.
- A short list of serious events sits in the label warnings: acute pancreatitis, gallbladder disease, acute kidney injury from fluid loss, severe gastrointestinal reactions, hypersensitivity, worsening of existing diabetic retinopathy, a small rise in heart rate, and pulmonary aspiration during anaesthesia or deep sedation.
- The rodent thyroid tumour finding is real, and its relevance to humans is unresolved. Human studies disagree with each other.
- The suicidal ideation warning was removed from GLP-1 labels in 2026 after a class review found no increased risk, so older articles on that point are out of date.
- None of the trial safety data transfers automatically to a grey-market or “research use only” vial. That is a different product.
Most of the numbers below come from semaglutide, which has the largest published dataset. Where a figure is specific to one product or trial, it says so.
Why the side effects look the way they do
GLP-1 receptor agonists work by mimicking a gut hormone. They stimulate insulin release in a glucose-dependent way, suppress glucagon when it is inappropriately high, slow how fast the stomach empties, and reduce appetite through circuits in the hypothalamus and brainstem.
Three of those four actions happen in or around the gut. That is the whole explanation for the side effect profile. Weight loss comes from eating less, driven centrally, not from blocking absorption. Slowed gastric emptying is most pronounced early in treatment, which is exactly when nausea peaks. How peptides work in the body covers the general receptor logic; the semaglutide page carries the compound-specific detail.
Evidence tier: Approved. Semaglutide is approved for type 2 diabetes, chronic weight management, cardiovascular risk reduction, kidney outcomes in type 2 diabetes, and non-cirrhotic MASH with F2 to F3 fibrosis. The safety data below comes from that evidence base, not from extrapolation.
The common ones, with numbers
Nausea, diarrhoea, vomiting, constipation and abdominal pain are the headline group. A dedicated analysis of gastrointestinal tolerability across the STEP trials found these events were mostly mild to moderate and transient, and that weight loss was not driven by nausea (Wharton 2022). People who felt fine still lost weight.
Beyond the gut, the label and trial reports list gallstones, injection-site reactions, headache, fatigue, hair loss, and a heart-rate increase of a few beats per minute.
One less familiar entry is dysaesthesia, meaning skin sensitivity or tingling. In the STEP UP trial of semaglutide 7.2 mg weekly it appeared in 22.9% of participants, against 6.0% at 2.4 mg and 0.5% on placebo. That is a clear dose-response signal and a useful reminder that higher doses are not simply “more of the same”.
For a sense of real-world tolerability: in SELECT, which ran a mean of about 34 months, 16.6% of the semaglutide group stopped for adverse events versus 8.2% on placebo (Lincoff 2023).
The serious ones on the label
The US semaglutide label as revised in June 2026 lists the following as warnings: acute pancreatitis including necrotizing cases with postmarketing deaths, acute gallbladder disease, hypoglycaemia when combined with insulin or a sulfonylurea, acute kidney injury from volume depletion, severe gastrointestinal reactions, hypersensitivity including anaphylaxis and angioedema, diabetic retinopathy complications in type 2 diabetes, heart-rate increase, and pulmonary aspiration during general anaesthesia or deep sedation. Postmarketing reports include ileus and intestinal obstruction.
Two of these deserve unpacking.
Retinopathy. In SUSTAIN-6, retinopathy complications were more frequent on semaglutide, with a hazard ratio of 1.76 (Marso 2016). The leading explanation is that rapidly lowering a high A1c can transiently worsen pre-existing retinopathy, a phenomenon known independently of this drug class. It is a reason for an eye exam before starting in someone with type 2 diabetes and known retinopathy, not a general warning for everyone.
Aspiration and procedures. Delayed gastric emptying means a stomach that is not reliably empty before sedation. The 2024 multisociety guidance concluded that most patients can continue their GLP-1 agonist, but those in the escalation phase, on higher doses, or with gastrointestinal symptoms should follow a liquid diet for 24 hours beforehand or take other individualised measures. Anyone having a procedure should tell the anaesthesia team, including people using a compounded or grey-market product.
The absolute contraindications are narrower than people assume: a personal or family history of medullary thyroid carcinoma, MEN 2, and prior serious hypersensitivity.
The thyroid question, honestly
GLP-1 receptor agonists activate thyroid C-cells in rodents, causing calcitonin release and C-cell proliferation (Bjerre Knudsen 2010). That finding produced the boxed warning. Rodent C-cells express far more GLP-1 receptor than human C-cells do, so the translation is genuinely uncertain.
Human data conflict. A French nested case-control study reported increased risk of all thyroid cancer and of medullary thyroid cancer, particularly after one to three years of use (Bezin 2023). A large Scandinavian active-comparator cohort found no substantial increase (Pasternak 2024). Detection bias, meaning more neck imaging in treated patients, may account for part of the difference.
The labels describe routine calcitonin testing and thyroid ultrasound as being of uncertain value, and the contraindication stays in place as a precaution while the question is unsettled. That is the accurate state of knowledge as of September 2026: neither reassurance nor alarm.
What changed, and what people get wrong
Suicidality. The FDA requested removal of the suicidal ideation and behaviour warning from the semaglutide, liraglutide and tirzepatide weight-management labels in a Drug Safety Communication dated 13 January 2026, after a class review found no increased risk. The removal was confirmed in label supplements in February 2026. If you read an older article saying otherwise, that is why.
NAION. Non-arteritic anterior ischaemic optic neuropathy, a form of sudden painless vision loss, emerged as a signal from observational data (Hathaway 2024) and was examined in later population studies (Tesfaye 2026). In June 2025 the European Medicines Agency’s safety committee concluded NAION is a very rare side effect of semaglutide, at up to 1 in 10,000, and advised stopping the drug if it is confirmed. The Canadian Ozempic monograph lists NAION among postmarket reports. The US Wegovy label as of June 2026 does not list it. Absolute risk appears very low, and sudden painless vision loss warrants urgent ophthalmology assessment regardless of cause.
Lean mass. DXA substudies suggest roughly a quarter to two fifths of the weight lost on these drugs is lean mass. In SURMOUNT-1 about 25% of weight lost was lean mass, and that figure was the same for tirzepatide and placebo (Look 2025). Reported proportions range from about 15% to 40-60% depending on population and method, and “lean mass” includes water, organs and bone, not only muscle. MRI work suggests muscle quality, meaning less fat infiltration, may actually improve. It matters most in older adults and anyone with low baseline muscle, and resistance training plus adequate protein is the standard response.
Fractures. In SELECT, more hip and pelvic fractures occurred on semaglutide than placebo among women (1.0% versus 0.2%) and among people aged 75 and older, per the Wegovy label. Another reason resistance training belongs in the plan.
Pancreatic enzymes. Asymptomatic lipase rises are expected: the Wegovy label reports mean lipase up 39% and amylase up 15-16% in the weight trials. Screening enzymes in someone with no symptoms generates worry, not information.
Where the safety data stops applying
Everything above describes reviewed products at known doses. It does not describe a vial bought online.
An analysis of semaglutide purchased without a prescription from illegal online sellers found all three delivered vials were probably substandard or falsified, with semaglutide content 29% to 39% above label, endotoxin detected in every sample, and purity far below the claimed 99%; ordered pens never arrived at all (Ashraf 2024). Compounded semaglutide has separately generated 10-fold overdoses from confusion between milligrams, millilitres and syringe units, which the FDA flagged in an alert dated 26 July 2024. Mg to ml syringe conversion exists because that arithmetic causes real harm.
As of September 2026, Health Canada’s advisory of 9 April 2026 is explicit that a “research use only” label confers no legal status of its own. The FDA had received 990 adverse-event reports for compounded semaglutide by 31 May 2026.
The bottom line
The GLP-1 side effect profile is unusually well characterised: mostly gastrointestinal, mostly dose-related, mostly early, with a short and specific list of serious warnings. The thyroid question remains genuinely unresolved, the suicidality warning has been withdrawn, and lean-mass loss is best treated as a training-and-protein problem rather than a reason to avoid treatment. None of this data describes an unregulated vial, where content, purity and sterility are all unknown.
Education only. This page is educational. Most compounds referenced are not approved for human use, and a “research use only” label carries no legal status of its own. Nothing here constitutes medical advice, diagnosis, or treatment.
Sources
- Wharton S, et al. Gastrointestinal tolerability of once-weekly semaglutide 2.4 mg in adults with overweight or obesity, and the relationship between gastrointestinal adverse events and weight loss. Diabetes Obes Metab, 2022.
- Marso SP, et al. Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes. N Engl J Med, 2016.
- Lincoff AM, et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med, 2023.
- Bjerre Knudsen L, et al. Glucagon-like Peptide-1 receptor agonists activate rodent thyroid C-cells causing calcitonin release and C-cell proliferation. Endocrinology, 2010.
- Bezin J, et al. GLP-1 Receptor Agonists and the Risk of Thyroid Cancer. Diabetes Care, 2023.
- Pasternak B, et al. Glucagon-like peptide 1 receptor agonist use and risk of thyroid cancer: Scandinavian cohort study. BMJ, 2024.
- Ashraf AR, et al. Multifactor Quality and Safety Analysis of Semaglutide Products Sold by Online Sellers Without a Prescription. J Med Internet Res, 2024.
- FDA Drug Safety Communication, removal of the suicidal behaviour and ideation warning from GLP-1 receptor agonist labels, 13 January 2026; Wegovy US prescribing information, revised June 2026; EMA PRAC conclusion on NAION and semaglutide, June 2025.
- Health Canada, “Think twice before injecting peptides bought online”, 9 April 2026.