Stopping GLP-1 drugs raises heart risk within two years, study finds

Researchers at Washington University School of Medicine in St. Louis tracked more than 333,000 U.S. veterans with type 2 diabetes for up to three years to see what happens to heart health after people stop taking GLP-1 drugs such as semaglutide and tirzepatide. These drugs, sold under names including Ozempic, Wegovy, Mounjaro and Zepbound, are now used by roughly one in eight U.S. adults for diabetes and weight loss, and they also lower cardiovascular risk while people stay on them. The study compared 132,551 participants prescribed GLP-1 drugs with 201,136 participants prescribed a different diabetes drug class, sulfonylureas, which include glipizide, glimepiride and glyburide. The results appeared March 18 in BMJ Medicine.

People who stayed on GLP-1 drugs for the full three years had the largest drop in risk of heart attack, stroke or death — 18% lower than the sulfonylurea group, or roughly 4 fewer major cardiovascular events per 100 people over three years. But stopping erased that protection quickly. Six months off the drug was enough to produce a meaningfully higher risk compared with staying on it, and the risk kept climbing the longer someone stayed off: a 14% increase in risk after one year without the drug, and a 22% increase after two years, compared with continuous use. People who took GLP-1 drugs for less than 18 months before quitting saw no significant risk reduction at all compared with the sulfonylurea group.

The team checked each GLP-1 user's treatment status every six months over the three-year follow-up, sorting participants into 16 different patterns of continued use, stopping, or stopping and later resuming. This let them measure not just whether someone was on the drug at the end of the study, but exactly how long they had used it, how long they had gone without it, and whether they went back on it. Over the course of the study, 26% of GLP-1 users stopped taking the medication altogether, and about 23% had a gap of six months or more before resuming.

The comparison group taking sulfonylureas served as the baseline against which every GLP-1 pattern was measured. Researchers also adjusted for people's underlying health conditions, since the two drug groups were not identical to begin with. Senior author Ziyad Al-Aly, a WashU Medicine clinical epidemiologist, said the study grew out of noticing that about half of users stop taking GLP-1 drugs shortly after starting.

The researchers describe the pattern as benefits that build slowly but erode quickly. Continuing GLP-1 treatment for two or two-and-a-half years before stopping still produced a significant risk reduction — 7% and 15%, respectively — showing that longer exposure buys some lasting protection. But even people who resumed the drug after an interruption did not fully recover the benefit: those who interrupted and later resumed gained a 12% average risk reduction, compared with the 18% reduction seen in people who never stopped. A gap of just six months before resuming still cut into the benefit, producing a 4% to 8% increase in risk compared with continuous use.

Al-Aly said the visible sign of stopping is weight regain, but the cardiovascular damage is not something patients can see. "When they stop, it's not just weight that comes back; they experience a resurgence in inflammation, blood pressure, and cholesterol," he said. "Weight regain is visible; the metabolic reversal is not." He added that restarting the medication helped restore some protection, but only partially, calling discontinuation a "lasting scar." The study's authors note that discontinuation is already known to be common, with prior work cited in the paper putting first-year discontinuation rates as high as 36% to 81%, and often driven by cost, side effects, or drug shortages.

Al-Aly argued that adherence to GLP-1 treatment should be treated as an outcome that matters in its own right, not an afterthought for clinicians. He called for health systems to build in support for indefinite treatment, including managing side effects proactively, having direct conversations with patients about the long-term nature of the treatment, identifying patients at risk of stopping, and addressing the cost barriers that make sustained GLP-1 therapy unaffordable for many patients. The study's authors write that strategies to reduce treatment gaps should be developed and evaluated to make the most of the drugs' heart-protective effects.

The study followed veterans with type 2 diabetes specifically, using Veterans Affairs healthcare records, and compared GLP-1 users against people taking sulfonylureas rather than against a group taking no diabetes medication at all. The researchers describe their approach as a target trial emulation, a method for estimating what a randomized trial would show using existing patient records rather than randomly assigning people to stop or continue the drug. The findings point to a duration-dependent relationship — longer use built more protection, and longer gaps eroded it — but the authors frame their conclusions in terms of associations found in this veteran population, drawn from health records rather than a randomized clinical trial.