Stopping GLP-1 drugs erases heart benefits within months, veteran study shows

The prescriptions are easy to start. Semaglutide, tirzepatide - the drugs sold as Ozempic, Wegovy, Mounjaro, and Zepbound. Roughly one in eight American adults now takes a GLP-1 receptor agonist for diabetes or weight loss. The drugs lower blood sugar. They reduce body weight. And increasingly, clinical trials have shown they protect the heart.

But about half of users stop within the first year. The reasons are familiar: cost (often exceeding $1,000 per month without insurance coverage), persistent nausea and gastrointestinal side effects, drug shortages, or simply the difficulty of maintaining a weekly injection indefinitely. When they stop, the weight comes back. That much is well known.

What happens to their hearts? A study of more than 333,000 U.S. veterans, published March 18 in BMJ Medicine, provides an answer that should concern both patients and clinicians: the cardiovascular benefits of GLP-1 drugs erode quickly after discontinuation - far faster than they took to build up.

333,687 veterans, three years of follow-up

The study, led by Ziyad Al-Aly, MD, a clinical epidemiologist at Washington University School of Medicine in St. Louis and chief of the Research and Development Service at the VA Saint Louis Health Care System, compared two groups of veterans with type 2 diabetes. One group (132,551 participants) was prescribed GLP-1 receptor agonists. The other (201,136 participants) was prescribed sulfonylureas - an older class of diabetes medication that includes glipizide, glimepiride, and glyburide.

The researchers tracked outcomes for up to three years, evaluating GLP-1 users' treatment status every six months. Over the course of the study, 26% of GLP-1 users stopped taking the medication entirely. Another 23% experienced at least one interruption of six months or more before resuming treatment. That means nearly half of GLP-1 users had some significant gap in their treatment.

Benefits that build slowly and collapse fast

The pattern that emerged was stark. Veterans who took GLP-1 medications continuously for the full three-year study period saw the most protection: an 18% reduction in the risk of major adverse cardiovascular events - heart attack, stroke, and death - compared to the sulfonylurea group. That translates to roughly 4 fewer major cardiovascular events per 100 people over three years.

But these benefits accumulated gradually. Veterans who used GLP-1 drugs for two years before stopping still gained a 7% risk reduction. Those who continued for two and a half years before discontinuing saw 15% risk reduction. The protective effect built up over time, requiring sustained treatment to reach its full potential.

The collapse, by contrast, was rapid. Veterans who stopped GLP-1 treatment for one year saw enough erosion of benefits to largely cancel out the gains from prior treatment. Those who discontinued for two years experienced a 22% increase in cardiovascular events compared to those who stayed on the drugs - effectively wiping out the protection that had taken years to accumulate. Fewer than 18 months of GLP-1 use before discontinuation produced no statistically significant cardiovascular benefit at all compared to the sulfonylurea group.

The metabolic reversal you cannot see

Al-Aly describes the phenomenon as metabolic whiplash. When patients stop GLP-1 drugs, the visible change is weight regain. But beneath the surface, a cascade of metabolic reversals occurs: inflammation resurges, blood pressure climbs, cholesterol levels deteriorate. These changes are not visible to the patient and may not be immediately apparent in routine clinical visits, but they translate directly into cardiovascular risk.

The asymmetry between how slowly benefits build and how quickly they erode is particularly concerning. Cardiovascular protection from GLP-1 drugs appears to require years of continuous treatment to reach meaningful levels. But interrupting that treatment, even briefly, can undo a substantial portion of the accumulated benefit. The biological processes that protect blood vessels and reduce inflammation while on the drug do not simply pause during a gap in treatment - they reverse.

Restarting helps, but not enough

The study also examined what happens when patients restart GLP-1 treatment after an interruption. The news was mixed. Resuming treatment did restore some cardiovascular protection: veterans who experienced an interruption and then restarted achieved a 12% risk reduction on average, compared to the 18% reduction seen with continuous use.

But the gap between continuous users and those who interrupted and restarted was consistent and meaningful. Even a six-month interruption before resuming treatment reduced the cardiovascular benefit by 4 to 8 percentage points compared to uninterrupted use. Longer gaps corresponded to larger reductions in restored benefit.

Al-Aly frames this as evidence that discontinuation leaves a lasting scar - a period of unprotected metabolic deterioration whose cardiovascular consequences are not fully reversed by resuming treatment. The heart, it seems, keeps a record of the interruption.

Cost, side effects, and the adherence problem

The clinical implications are straightforward but logistically difficult. If GLP-1 drugs need to be taken continuously and indefinitely to maintain cardiovascular protection, then the barriers to adherence - cost, side effects, supply shortages - become cardiovascular risk factors in their own right.

At current pricing, many patients face monthly costs that exceed $1,000 without employer insurance or Medicare coverage. Insurance formulary changes, prior authorization requirements, and periodic drug shortages create involuntary gaps in treatment even for patients who want to continue. Side effects, particularly nausea and gastrointestinal distress, cause a substantial fraction of patients to abandon treatment voluntarily.

Al-Aly argues that health systems need to treat adherence as a clinical outcome, not an afterthought. That means proactive management of side effects, transparent conversations about the long-term nature of treatment, infrastructure to identify patients at risk of stopping, and policy efforts to address cost barriers. Given the study's findings, a patient who stops GLP-1 treatment due to a coverage gap or a formulary change is not just losing weight management - they may be losing cardiovascular protection that took years to establish and that may never be fully restored.

Observational data with real-world relevance

Several limitations apply. This was an observational study using VA electronic health records, not a randomized controlled trial. Patients who stopped GLP-1 treatment may have done so for reasons correlated with higher cardiovascular risk - worsening health, medication intolerance, loss of healthcare access - creating potential confounding that the study's statistical methods can reduce but not eliminate.

The study population was predominantly male U.S. veterans with type 2 diabetes, a demographic that may not be representative of the broader population of GLP-1 users, which increasingly includes younger patients and those using the drugs primarily for weight management rather than diabetes control. Whether the same patterns of benefit accumulation and erosion apply to non-diabetic users taking GLP-1 drugs for weight loss alone is an important unanswered question.

The comparison group used sulfonylureas, which have their own cardiovascular risk profile. The study design - a target trial emulation framework - attempted to mimic the structure of a randomized trial using observational data, but inherent limitations of this approach remain.

Despite these caveats, the study's central finding carries practical weight: stopping GLP-1 drugs is not a neutral act. The cardiovascular benefits do not persist in the drug's absence. They fade, and they fade fast. For the millions of Americans now taking these medications, that reality needs to be part of the conversation from the very first prescription.