Question explored with the scientific record
Conversation so far
- 1 Is wegovy safe
- 2 What long-term safety data exists beyond 2–3 years, and what’s unknown?
Answered on this page ↓
What long-term safety data exists beyond 2–3 years, and what’s unknown?
Beyond 2–3 years, nobody has run the trial that would answer this. The longest follow-up in front of me is about three years, and it was not designed to catch slow harms.
The longest outcome data here is a VA comparison of 21,790 veterans on GLP-1 drugs, with a median follow-up of about 2.74 to 3.01 years [3]. That is the ceiling in this stack. It found broadly similar heart and kidney outcomes across drugs, with a gallbladder signal for semaglutide [3]. It is observational, not randomized, and it stops at three years [3]. The cardiovascular trials are shorter still: SUSTAIN-6 ran a median of 2.1 years [6]. A 2025 systematic review of 25 studies concluded that "some long-term safety conclusions remain uncertain," and rated all 25 at high risk of bias [1].
| What is known | Follow-up | Design |
|---|---|---|
| VA head-to-head, 21,790 veterans | ~2.7–3.0 years | Observational [3] |
| SUSTAIN-6 cardiovascular trial | 2.1 years median | Randomized [6] |
| Systematic review, 25 studies | Mixed, short | High risk of bias [1] |
What is genuinely unknown past three years: whether muscle loss during rapid weight loss recovers or becomes permanent frailty, especially in older people [5]; whether rare harms like pancreatitis and the pancreatic tissue growth seen in one case report become more common with years of exposure [2][4]; and whether the drug must be taken indefinitely to hold the weight off, and what stopping does to the body over a decade. None of that is in these records. The retrieval also has no post-marketing surveillance data, and passive reporting systems catch almost nothing by design.
My call: there is no long-term safety data beyond roughly three years, and the three-year data is observational and underpowered for slow harms. Confidence: moderate that the gap is real.
Sources used 6
-
Investigation of clinical outcomes on the oral or injectable use of semaglutide and cardiovascular risks: a systematic review
A PRISMA-guided systematic review evaluating clinical outcomes and cardiovascular safety of both oral and injectable semaglutide in adults with type 2 diabetes, concluding that subcutaneous semaglutide tends to cause more endocrine-related adverse events while oral semaglutide m…
DOI: 10.54448/ijn25207 -
Semaglutide induced Pancreatitis
This case report discusses a 48-year-old male patient who developed acute pancreatitis after starting semaglutide, a glucagon-like peptide-1 receptor agonist, highlighting the potential risks associated with its use.
DOI: 10.36948/ijfmr.2024.v06i04.26238 -
Liraglutide vs Semaglutide vs Dulaglutide in Veterans With Type 2 Diabetes
A large, real-world, VA-wide head-to-head comparison of three commonly used GLP-1 receptor agonists (liraglutide, semaglutide, and dulaglutide) in veterans with type 2 diabetes, finding broadly similar kidney and cardiovascular outcomes across drugs, with potential mortality dif…
DOI: 10.1001/jamanetworkopen.2025.37297 -
MON-636 Growth of Pancreatic Heterotopic Tissue: A Rare Adverse Effect of Semaglutide
A 47-year-old woman with type 2 diabetes developed small bowel obstruction from growth of undiagnosed pancreatic heterotopia after one year of semaglutide treatment, suggesting a rare adverse effect of GLP-1 receptor agonists.
DOI: 10.1210/jendso/bvaf149.996 -
The impact of semaglutide and tirzepatide on skeletal muscle: Significant benefit or substantial risk?
This narrative review concludes that semaglutide and tirzepatide reduce skeletal muscle mass proportionally to weight loss, with adaptive changes and improved muscle function.
DOI: 10.21518/ms2025-467 -
Series: Cardiovascular outcome trials for diabetes drugs Semaglutide and SUSTAIN-6
This article in a cardiovascular outcome trials series summarizes SUSTAIN-6, a pre-licensing trial of once-weekly semaglutide in type 2 diabetes showing non-inferiority (with nominal superiority) for MACE and a reduction in non-fatal stroke, but an increased rate of retinopathy …
DOI: 10.15277/bjd.2021.301