Long-Term Ozempic Use: What We Know and What We Don't

The question patients ask most often about Ozempic is not about side effects or cost. It is: "How long do I have to take this?" The answer, based on current evidence, is probably indefinitely, and the reasons for that have more to do with the biology of obesity and diabetes than with any limitation of the drug itself. But the long-term data is still maturing, and there are legitimate questions that deserve honest answers.

What happens when you stop

The most important piece of long-term data on semaglutide comes from the STEP 1 extension study, published in Diabetes, Obesity and Metabolism in 2022 by Wilding and colleagues. In the original STEP 1 trial, participants on semaglutide 2.4 mg lost an average of 14.9% of their body weight over 68 weeks. The extension study followed a subset of these participants for an additional year after the drug was withdrawn.

The results were sobering. Within one year of stopping semaglutide, participants regained approximately two-thirds of their lost weight. Their metabolic improvements, including reductions in waist circumference, blood pressure, C-reactive protein, and HbA1c, also reversed proportionally. By the end of the follow-up period, the group that had stopped semaglutide looked metabolically similar to the group that had received placebo all along.

This is not a failure of the drug. It is a reflection of what obesity researchers have understood for decades: the biological systems that regulate body weight are persistent and powerful. When caloric intake drops and body fat decreases, the body responds with increased hunger hormones (particularly ghrelin), reduced energy expenditure, and changes in reward signaling that collectively push weight back toward the previous set point. Semaglutide suppresses these compensatory mechanisms while it is active, but removing the drug removes that suppression.

The parallel to other chronic diseases is instructive. No one expects blood pressure to remain controlled after stopping antihypertensive medication, or cholesterol to stay low after stopping a statin. The emerging consensus in obesity medicine is that pharmacotherapy for obesity, when effective, should be considered long-term treatment rather than a temporary intervention.

The long-term safety profile

Semaglutide has been available in some form since 2017 (Ozempic's FDA approval), which gives us roughly eight years of post-marketing surveillance data. GLP-1 receptor agonists as a class have been in clinical use since exenatide (Byetta) was approved in 2005, providing nearly two decades of accumulated safety data on the mechanism.

The major long-term safety considerations include:

• Thyroid C-cell tumors: The boxed warning on all GLP-1 receptor agonists is based on findings in rodent studies, where semaglutide caused thyroid C-cell tumors at clinically relevant exposures. However, rodents have a much higher density of GLP-1 receptors on thyroid C-cells than humans. To date, epidemiological studies and post-marketing data have not shown an increased risk of medullary thyroid carcinoma in humans taking GLP-1 agonists. A large pharmacovigilance study published in 2024 covering more than 2 million patient-years of GLP-1 agonist use found no signal. That said, the drug remains contraindicated in patients with a personal or family history of medullary thyroid carcinoma or MEN 2.
• Pancreatitis: Acute pancreatitis has been reported in patients taking semaglutide, though at rates that are difficult to distinguish from the baseline risk in the diabetic and obese population, which already has an elevated pancreatitis risk. In the SUSTAIN and STEP trials, pancreatitis occurred in less than 0.5% of participants. Patients with a history of pancreatitis should discuss the risk-benefit calculation with their physician.
• Gallbladder disease: Rapid weight loss from any cause increases the risk of gallstones, and semaglutide is no exception. In the STEP trials, cholelithiasis (gallstones) occurred in approximately 2.6% of semaglutide-treated patients versus 1.2% on placebo. The risk appears to be proportional to the rate and magnitude of weight loss.
• Diabetic retinopathy: In SUSTAIN 6, there was an unexpected increase in diabetic retinopathy complications in the semaglutide group (3.0% vs. 1.8% for placebo). This was primarily seen in patients with pre-existing retinopathy and is thought to be related to the rapid improvement in glycemic control, which can paradoxically worsen retinopathy in the short term. Patients with existing retinopathy should be monitored closely.

The cardiovascular case for long-term use

The SELECT trial, with its median follow-up of 39.8 months, provides the most direct evidence for the long-term cardiovascular benefit of semaglutide. The 20% reduction in MACE was observed over more than three years of continuous treatment, and the Kaplan-Meier curves for semaglutide and placebo continued to diverge throughout the study period. This suggests that the cardiovascular benefit is not a one-time effect but accumulates with ongoing treatment.

For patients with established cardiovascular disease and obesity, this creates a compelling argument for indefinite treatment. Stopping the medication does not just reverse the weight loss; it likely also reverses the cardiovascular protection. The analogy to statins is again relevant: just as statin therapy is considered lifelong for patients with atherosclerotic disease, semaglutide may occupy a similar role for patients with obesity-related cardiovascular risk.

What we still do not know

Despite the extensive clinical trial program, there are important gaps in the evidence:

• Safety beyond 5 years: The longest randomized data we have is approximately 4 years (from SELECT). There are no 10- or 20-year safety datasets. While the GLP-1 class has a reassuring track record over nearly two decades, semaglutide-specific ultra-long-term data does not exist yet.
• Optimal long-term dosing: It is unclear whether patients who reach their target weight can step down to a lower maintenance dose while preserving most of the benefit. Some clinicians are experimenting with dose reduction after stabilization, but there is no randomized evidence to guide this practice.
• Body composition over many years: The concern about lean mass loss is well-documented in the first 1 to 2 years of treatment. Whether continued use produces ongoing muscle loss or whether body composition stabilizes is not well-characterized in data beyond 68 weeks.
• Effects on bone health: Weight loss is associated with decreased bone mineral density, and some bariatric surgery patients experience increased fracture risk years later. Whether semaglutide-induced weight loss carries similar skeletal risks over the long term is still being studied.
• Demographic gaps: The major trials enrolled predominantly White populations. Long-term safety and efficacy data in Black, Hispanic, and Asian populations are limited, which is particularly relevant given the higher prevalence of type 2 diabetes in some of these groups.

A practical framework for patients

For patients currently on semaglutide or considering starting, the long-term picture looks like this: the drug works well for as long as you take it, and most of its metabolic benefits reverse when you stop. This does not make it a bad medication. It makes it a chronic therapy, like blood pressure medication or thyroid hormone replacement.

The practical considerations for long-term use include cost (a monthly expense measured in hundreds or thousands of dollars, depending on insurance), injection adherence over years or decades, and ongoing monitoring for side effects. Patients should have realistic conversations with their healthcare providers about these factors before starting treatment.

The good news is that the safety signal over nearly a decade of use is reassuring, the cardiovascular benefit is substantial and well-documented, and the field is actively studying how to optimize long-term outcomes including maintenance dosing, exercise protocols to preserve lean mass, and combination therapies that may allow lower doses of any single agent.

The honest answer to "how long do I have to take this?" is: for as long as you want the benefits to continue. Whether that framing feels discouraging or simply pragmatic depends on how you think about chronic disease management in general.