The new stopping GLP-1 heart risk study found a higher combined rate of heart attacks, strokes or deaths among people who stopped treatment. Everyone studied had type 2 diabetes, a condition that raises blood sugar. The study cannot show that stopping itself caused the difference.
The finding makes the reason for stopping and the follow-up plan worth discussing with the prescriber. GLP-1 medicines, including semaglutide, mimic a gut signal involved in appetite and blood sugar.
Stopping GLP-1 heart risk: what the new study found
As of November 12, 2026, the day this article was written, a study published November 9, 2026, in The Lancet Regional Health – Europe has brought this question into sharper focus. Antonio Ceriello and colleagues looked back at records from 80,302 people at 301 Italian diabetes clinics.
The records covered 2004–2024. Participants had type 2 diabetes and no previous major heart or blood-vessel event when they first received a GLP-1 prescription. Researchers matched people on recorded health characteristics to make the comparison fairer. Each matched group contained 28,713 people. Average follow-up was roughly four years.
The main measure combined heart attacks and strokes people survived with death from any cause. Death did not have to be heart-related.
| Outcome | After stopping | With continued treatment |
|---|---|---|
| Combined heart attack, stroke or death | 3.94 | 1.96 |
| Death from any cause | 2.23 | 0.48 |
These are events per 100 person-years: 100 people followed for one year each, or another mix totaling 100 years. The combined rate was about twice as high after stopping. It does not mean each person's chance of a heart attack doubled.
Time before stopping counted as time on treatment; time afterward counted as time off. Stopping meant no prescription at two consecutive visits. This was not a study of one late dose.
Why the reason for stopping matters
This was an observational study: researchers watched what happened in ordinary care. They did not randomly assign people to stop or continue.
The health problem that leads someone to stop can also raise their risk of a later event. Researchers call this confounding by indication. A person becoming weaker because of another illness may be more likely to stop and more likely to die.
Matching recorded health measures helps, but it cannot account for every unrecorded difference. The study did not record why each person stopped or their financial circumstances. Its authors specifically warn that the large death-rate difference is unlikely to come from stopping treatment alone.
That leaves a serious finding with an uncertain explanation. It neither proves that stopping causes heart attacks nor establishes that stopping is harmless. It also cannot compare starting and stopping with never having taken a GLP-1 medicine.
Cost, side effects and access problems deserve different conversations. The research on treatment costs and refill gaps explains why financial barriers can interrupt care. Those barriers were not established as the explanation for these Italian results.
Does this apply to Ozempic, Mounjaro or weight loss?
Semaglutide, the medicine in Ozempic and Wegovy, accounted for only a small part of this study group. The paper cannot supply a reliable semaglutide-specific stopping-risk estimate.
Tirzepatide, the medicine in Mounjaro and Zepbound, was not included. The authors explicitly say their findings cannot be extended to it. A headline about stopping Mounjaro therefore goes beyond this study.
The same limit applies to people using these medicines for weight management without diabetes. They were not the population studied.
The SELECT trial addressed a different question: adding semaglutide to care in people with established heart disease and no diabetes. Evidence about taking a medicine cannot simply be read backward as proof of what stopping will cause.
The guide to stopping tirzepatide covers its separate evidence about what happens after treatment ends. That evidence belongs in a prescribing discussion about tirzepatide; this Italian study cannot replace it.
What a stopping conversation can cover
A useful discussion starts with why treatment might end: side effects, another illness, pregnancy planning, access or a change in treatment goals. The prescriber can review blood sugar, blood pressure and cholesterol, alongside other medicines and the person's history.
Stable readings are useful information, but they cannot guarantee that heart risk is unchanged. In the Italian study, changes in these measures did not fully explain who later had an event.
The semaglutide stopping guide covers appetite, weight changes and tapering, meaning gradually reducing the dose. This news adds a reason to keep follow-up broader than the scale. The prescriber sets any dose changes and the plan for diabetes care after treatment ends.
Safety instructions still matter. As of November 12, 2026, the day this article was written, the FDA's November 9 safety communication requires warnings about a potential risk of a rare eye condition. That condition involves reduced blood flow to the nerve carrying sight signals to the brain.
FDA tells patients with sudden painless sight loss, blurred or missing vision, or changes in color or sharpness to contact a health professional right away. It tells clinicians to consider stopping the medicine if that condition develops, unless treatment benefits clearly outweigh the risk. A heart-risk headline does not override that instruction.
What this means for a compounded prescription
Semaglutide and tirzepatide prescribed through Promise are dispensed by a licensed U.S. compounding pharmacy as compounded medications. Compounding means preparing a medication for an individual prescription.
These preparations are different from an FDA-approved product: the formulations offered here are not FDA-approved. FDA explains that it does not review compounded drugs for safety, effectiveness or quality before marketing. The Italian study does not establish the same outcomes for these preparations.
A licensed provider may still prescribe — that decision is between you and your doctor.
What to watch next
The next studies need to record why people stop, separate individual medicines and include people without diabetes.
As of November 12, 2026, the day this article was written, an earlier BMJ Medicine study, published March 18, 2026, also linked longer treatment gaps with more major events. It examined U.S. veterans with type 2 diabetes. It used health records rather than random assignment, so it also leaves uncertainty about cause.
For now, the useful question is what ending treatment would mean for this person's overall care. At Promise, a licensed U.S. provider reviews every request, and not everyone qualifies.





