Researchers at Washington University School of Medicine in St. Louis followed more than 333,000 U.S. veterans with type 2 diabetes for up to three years. They found that interrupting GLP-1 treatment for six months or stopping it completely was associated with a significantly higher risk of major cardiovascular events compared with continuing treatment.

The longer people remained off treatment, the greater the risk appeared to be. Participants who stopped taking GLP-1 drugs for two years had up to a 22% higher risk of heart attack, stroke, and death than those who continued treatment. Much of the cardiovascular protection associated with ongoing GLP-1 therapy was also lost.

The study, published in BMJ Medicine, suggests that stopping GLP-1 drugs may have effects beyond weight regain. The findings also highlight the potential importance of uninterrupted treatment for maintaining cardiovascular protection.

“There’s so much excitement about starting GLP-1 drugs, but not enough attention has been paid to what happens when people stop,” said lead author Ziyad Al-Aly, MD, a clinical epidemiologist at WashU Medicine and chief of research and development services at the VA St. Louis Healthcare System.

“Many people stop taking it after a few months because of cost, side effects, and lack of supplies. When they stop, not only do they regain the weight, but they also see a return of inflammation, blood pressure, and cholesterol. You see weight gain back, but you don’t see a metabolic reversal.”

“Our data suggest that this metabolic whiplash is detrimental to heart health,” Al-Aly added. “When I restarted the medication, I regained some protection, but it was only partial, and I found that discontinuing the medication left permanent scars.”

Continuous GLP-1 treatment provided the greatest heart protection

GLP-1 drugs include the semaglutide medications Ozempic and Wegovy, as well as the tirzepatide medications Mounjaro and Zepbound. After observing that about half of users stop taking GLP-1 drugs shortly after starting treatment, Al-Aly and his colleagues investigated what happens to cardiovascular health after treatment is discontinued.

The researchers analyzed data from 333,687 veterans with type 2 diabetes. Of those participants, 132,551 were prescribed GLP-1 drugs and 201,136 were prescribed sulfonylureas, another type of diabetes medication.

Sulfonylureas include glipizide (Glucotrol), glimepiride (Amaryl), and glyburide (such as Diabeta). Participants were followed for up to three years, and researchers reassessed their GLP-1 treatment status every six months.

During the study period, 26% of GLP-1 users stopped taking the medication completely. Approximately 23% experienced a treatment gap lasting at least six months before eventually restarting treatment.

The clearest cardiovascular benefits were seen among people who continued GLP-1 therapy throughout the three-year study.

Compared with participants taking sulfonylureas, those who consistently continued GLP-1 treatment had an 18% lower risk of major cardiovascular events. That represents approximately four fewer serious cardiovascular events per 100 people over three years.

Participants who continued GLP-1 treatment for two years or 2.5 years before stopping for the remainder of the study also experienced significant risk reductions of 7% and 15%, respectively.

By contrast, participants who discontinued GLP-1 therapy before reaching 18 months of treatment did not show a significant reduction in cardiovascular risk compared with those taking sulfonylureas by the end of the study.

GLP-1 treatment gaps may weaken cardiovascular protection

Interrupting and then restarting GLP-1 treatment appeared to reduce its cardiovascular benefits.

The risk of major cardiovascular events was 18% lower among participants who continued taking GLP-1 drugs for three years. Among those who temporarily stopped and later restarted treatment, the average risk reduction was 12%.

Even a six-month interruption before treatment resumed was associated with waning protection. Compared with continuous use, treatment gaps were associated with a 4% to 8% increase in cardiovascular risk.

Longer periods without treatment were linked to an even greater loss of protection. Participants who stopped GLP-1 therapy for one year without restarting had a 14% higher risk of cardiovascular events than continued users. After two years off treatment, the increase reached 22%.

This pattern suggests that cardiovascular benefits associated with GLP-1 treatment may be lost relatively quickly after the medication is discontinued.

Restarting GLP-1 drugs may not fully restore lost benefits

The study findings support the importance of continued treatment when patients and clinicians aim to maintain the cardiovascular effects of GLP-1 therapy. They also suggest that reducing treatment interruptions may help maximize the medications’ potential heart benefits.

“Clinicians should treat compliance with GLP-1 therapy as an important outcome in its own right, rather than an afterthought,” Al-Aly said. “Health care systems need to recognize that GLP-1 treats chronic diseases and develop plans to help people continue taking the medication indefinitely.”

He said those plans should include proactive management of side effects, open conversations about the long-term nature of treatment, systems to identify and support patients at risk of discontinuation, and solutions to cost barriers that make GLP-1 therapy unsustainable for many people.

Al-Aly noted that this issue is particularly important because cardiovascular protection associated with GLP-1 treatment appears to accumulate over time but may disappear more quickly after treatment stops.

Among the study participants, one year off treatment was enough to lose benefits built up over years of continuous therapy. Restarting the medication restored some protection, but not all of it.

This study was funded by the U.S. Department of Veterans Affairs. The funders had no role in the study design, data collection, analysis, interpretation, report writing, or decision to publish the findings. The content does not represent the views of the U.S. Department of Veterans Affairs or the U.S. Government.