Stopping a GLP-1 Was Linked to Less Heart Protection in a Large Study of Veterans With Type 2 Diabetes
Author
Date Published
- Twitter
- Facebook
- LinkedIn
- Instagram
- Copy Link
In large randomized trials, some GLP-1 medicines lowered the combined risk of heart attack, stroke or heart-related death. But what happens to that benefit when you stop? A large study of US veterans with type 2 diabetes looked at exactly that. It was published in March 2026 in BMJ Medicine (BMJ Medicine) and ScienceDaily republished the university's news release in September (ScienceDaily). The longer people stayed off the drug, the more of the heart benefit seemed to slip away. The study is large, but it's observational, so it shows a link and can't prove that stopping caused the change.
Who This Helps
This is for anyone on a GLP-1 who is thinking about stopping, whether because of cost, side effects or reaching a goal weight. It's also for anyone who has seen headlines saying stopping can "quickly erase" the heart benefits and wants to know what the research really says.
What the study looked at
Researchers at Washington University in St. Louis and the VA used veterans' health records (WashU). The study included people who started treatment between 2017 and 2023, tracked for up to three years: 132,551 veterans who started a GLP-1, compared with 201,136 who started an older type of diabetes pill called a sulfonylurea (BMJ Medicine).
A few details shape what the results can tell us:
- Everyone had type 2 diabetes. The people in it were mostly older men, with an average age of about 66 (BMJ Medicine). It didn't look at people taking these drugs only for weight loss.
- Most were on semaglutide. About two-thirds took semaglutide, the drug in Ozempic and Wegovy. Very few took tirzepatide, well under 1% (BMJ Medicine).
- "Stopping" meant a gap in refills. Someone counted as stopping if they went at least 90 days without a refill after their supply ran out (BMJ Medicine).
- The outcome was broad. The researchers counted heart attack, stroke or death from any cause, over three years (BMJ Medicine).
It's an observational study. The team used methods built to mimic a trial, but nobody was randomly told to stop or keep going. Keep that in mind as you read the results.
What it found
Staying on was linked to lower risk. People who took a GLP-1 for all three years had about an 18% lower relative risk of heart attack, stroke or death than people who started the older pill. In real numbers, the team estimated that's about 3.8 fewer of these events for every 100 people over three years (BMJ Medicine).
Risk rose the longer people were off. Compared with people who stayed on for all three years, people who stopped for good had a higher relative risk. It grew with time off the drug (BMJ Medicine):
- About 4% higher for those off the drug for the last six months.
- About 14% higher for those off for the last year.
- About 22% higher for those off for the last two years.
Those are relative numbers, so a 22% higher risk doesn't mean 22 out of 100 people had a heart attack. On average, people who stopped and never restarted had about the same three-year risk as people who started the older pill. Most of them stopped within the first year (BMJ Medicine). Dr. Melanie Jay, who directs NYU Langone's Comprehensive Program on Obesity, told CNN, "It's not like you're worse off than if you'd never taken it. You're just worse off than if you've stayed on" (CNN). The study didn't test that directly. It compared people who stopped with people on an older pill, not with people who took no diabetes drug.
Shorter use showed less benefit. People who took a GLP-1 for a year and a half or less, then stayed off, showed little or no lower risk than people on the older pill. People who took it for two years before stopping kept a little benefit, about 7% lower relative risk (BMJ Medicine).
Gaps mattered too. People who took a break and then restarted still had a higher relative risk than people who never stopped. On average they kept more benefits than people who never went back. Their breaks were short, about five months for the typical person (BMJ Medicine).
The authors wrote that even brief breaks "might progressively erode and could ultimately reverse" the protection (BMJ Medicine). The study can't prove that stopping caused the change, and the authors use careful words like might and could.
Find a provider in your state
Answer a few questions and see live pricing from the providers that ship where you live.
What the study can't tell us
Because no one was randomly picked to stop, people who stopped may have been different in ways the data didn't catch. The authors say people may stop because the drug doesn't seem to work or because of side effects. They also say they can't fully rule out other hidden differences (BMJ Medicine).
There are a few other limits too:
- Refills were used as a stand-in for taking the drug, which isn't always the same thing (BMJ Medicine).
- The study doesn't tell us why the benefit faded. Whether it's weight coming back or something else still needs to be confirmed (BMJ Medicine).
- The people studied were mostly older men with diabetes, and the authors note the study plan wasn't registered ahead of time. An editorial in the same journal said the results need testing in younger people, women, and people who take these drugs for weight but don't have diabetes (BMJ Medicine editorial).
That same editorial called the findings plausible and in line with earlier randomized trial findings. It also warned that key differences may remain between people who kept going and people who stopped (BMJ Medicine editorial). VA drug coverage also made it less likely that cost was the main reason people stopped (BMJ Medicine). That may mean the results fit less well for people who stop because of cost.
How this fits with other research
The best evidence that these drugs lower heart risk comes from randomized trials. The SELECT trial tested semaglutide in adults with overweight or obesity who had already had a heart attack, a stroke or artery disease in the legs, and who didn't have diabetes. Over a little more than three years, heart attack, stroke or heart-related death happened in 6.5% of people on the drug and 8.0% on a dummy shot (ACC). That's about a 20% lower relative risk. By our math, it works out to about 1.5 fewer events for every 100 people over the trial. That trial counted deaths from heart causes, not deaths from any cause, so its numbers don't line up exactly with the VA study.
Other research suggests heart risk markers drift back after stopping. One small follow-up of a semaglutide trial (STEP 1) looked at adults with overweight or obesity and no diabetes. Their average top-number blood pressure fell from 129 to 121 on the drug. A year after people stopped, it was back up to 131 (Diabetes, Obesity and Metabolism). Blood pressure rose in the placebo group too, and the analysis was exploratory. Blood sugar and blood fats also moved back toward where they started, though a year after stopping, blood sugar and some blood fats were still slightly better than in the placebo group. Those are markers, not heart attacks, but they point the same way.
What we don't have is a trial that randomly assigns people to stop and then tracks heart attacks. The journal's editorial said such a trial in people with type 2 diabetes would raise ethics concerns (BMJ Medicine editorial).
What this means for you
The study can't say whether a planned stop works out differently from a missed refill, since it counted any 90-day gap as stopping. If you're thinking about stopping, especially if you have diabetes or heart disease, your prescriber can go over your other heart risks with you first. If cost is the problem, you can compare cash-pay prices and check your plan's coverage before you stop (GLP Winner).
If you do stop, the basics of heart health still count for a lot. The American Heart Association suggests 2½ hours of moderate exercise or 75 minutes of hard exercise a week. It calls blood pressure under 120/80 ideal (American Heart Association). It also lists not smoking, healthy cholesterol and blood sugar, good sleep, and a healthy diet and weight. We covered ways to slow weight regain after stopping in this recent article, and walking in our walking guide.
Final Takeaway
In a large study of veterans with type 2 diabetes, staying on a GLP-1 was linked to a lower risk of heart attack, stroke or death than an older diabetes pill. That lower risk seemed to fade the longer people were off the drug. It's an observational study, so it can't prove stopping caused the change. It may not apply to people who take these drugs for weight alone. The study couldn't tell planned stops from missed refills. If you're thinking about stopping, talk it through with your prescriber first.
If you enjoy posts like these, you can subscribe to receive newsletter updates.
Frequently Asked Questions
Do the heart benefits of Ozempic go away when you stop?
In a large study of veterans with type 2 diabetes, where about two-thirds took semaglutide (the drug in Ozempic), the lower risk of heart attack, stroke or death linked to GLP-1s seemed to shrink the longer people were off the drug. People who stopped for good and stayed off for two of the three years had about 22% higher relative risk than people who stayed on. The study shows a link, not proof of cause.
Am I worse off than if I never took a GLP-1?
The study didn't test that. Compared with people who started an older diabetes pill, people who stopped for good had about the same heart risk, neither clearly higher nor clearly lower. It didn't compare them with people who took no diabetes medicine.
Does this apply if I take a GLP-1 for weight loss only?
We don't know yet. Everyone in the study had type 2 diabetes, and most were older men. The journal's editorial said the findings need to be tested in people without diabetes.
Find a provider in your state
Answer a few questions and see live pricing from the providers that ship where you live.
Keep Reading
A new survey linked GLP-1s to nail lifting. See the real numbers, why experts say it's too early to worry, and what to do if your nails change.
Spirulina may lower blood pressure and cholesterol a little. See what human studies show, what's hype, and why it adds little to your protein goal.
The protein grams on a label are a count, not a grade. Learn why collagen barely counts, how to read the %DV, and what GLP-1 users need.
Categories
Written by
Lauren PescarusLauren Pescarus is a team member with GLP Winner where she works on marketing, content creation, and operations. She has over 10 years experience in the content creation space, including in the GLP-1 space where she works to stay on top of access news, research updates, and lifestyle tips guided by science.