As use of injectable GLP-1 medicines expands in the United States, new real‑world data from Cleveland Clinic point to a more nuanced picture of what happens when patients stop therapy. In one of the largest observational analyses to date, adults who discontinued semaglutide or tirzepatide did not, on average, experience major weight rebound over the following year-largely because many restarted treatment or pursued alternative obesity care. The findings, published in the peer‑reviewed journal Diabetes, Obesity and Metabolism, help frame expectations for patients, clinicians, and payers navigating chronic weight management.
What the analysis captured in routine care
The study reflects care delivered inside a large, integrated health system rather than in tightly controlled trial conditions, offering a view of how GLP‑1 use and discontinuation actually unfold in practice.
- Setting: Cleveland Clinic patients in Ohio and Florida receiving care in everyday clinical practice.
- Cohort size: 7,938 adults with obesity or overweight who started semaglutide or tirzepatide for obesity or type 2 diabetes and then stopped within 3-12 months.
- Follow‑up window: Outcomes assessed 12 months after discontinuation.
- Medications examined: semaglutide (Ozempic/Wegovy) and tirzepatide (Mounjaro/Zepbound).
| Patient group | Mean weight change before stopping | Mean weight change 12 months after stopping | Share gaining weight after stopping | Share stable or still losing after stopping |
|---|---|---|---|---|
| Treated for obesity | -8.4% body weight | +0.5% body weight | 55% | 45% |
| Treated for type 2 diabetes | -4.4% body weight | -1.3% body weight | 44% | 56% |
How patients navigated care after stopping GLP‑1 therapy
Rather than a simple on‑off decision, discontinuation often marked a pivot point in longer‑term obesity and diabetes management. Many individuals continued their weight‑management journey through other evidence‑based options in the year after discontinuation.
| Care pathway within 12 months of stopping | Share of patients |
|---|---|
| Switched to a different medication (including older‑generation agents or switching between semaglutide and tirzepatide) | 27% |
| Restarted the original medication | 20% |
| Continued treatment via lifestyle‑modification visits (e.g., dietitians, exercise specialists) | 14% |
| Transitioned to metabolic/bariatric surgery | Less than 1% |
For health‑system leaders and insurers, these patterns underscore that GLP‑1 “stopping” often means treatment cycling, not permanent exit, with knock‑on implications for long‑term budgeting and capacity planning.
Why real‑world outcomes can diverge from trial results
Randomized trials that require complete discontinuation without alternative treatment have documented substantial weight regain within a year. In community practice, however, patients and clinicians rarely accept an abrupt therapeutic void. Step‑through strategies-restarting therapy, switching within or across classes, or intensifying non‑pharmacologic care-appear to blunt average rebound.
As the study lead stated: “Our real-world data show that many patients who stop semaglutide or tirzepatide restart the medication or transition to another obesity treatment, which may explain why they regain less weight than patients in randomized trials,” Dr. Gasoyan said. For policymakers and guideline committees, the message is that discontinuation data from rigid trials may underestimate the stabilizing role of adaptive care pathways in everyday practice.
Cost, coverage, and the restart effect
Economic and policy constraints strongly shaped which patients could restart or switch therapies, making payer design a central part of the clinical story.
- Primary reasons for stopping: cost and insurance coverage limits were leading drivers; side effects were also reported.
- Coverage asymmetry: individuals using these medicines for diabetes were more likely to restart than those using them for obesity, reflecting more consistent benefits coverage for diabetes prescriptions.
- Benefit design pressures: prior authorization, quantity limits, and step‑therapy rules shape whether patients can resume a GLP‑1 or pivot to alternatives.
These dynamics play out within a regulatory environment in which the U.S. Food and Drug Administration authorizes GLP‑1s for specific indications, while public and private payers decide whether and how to cover them. For federal benefit programs, a key constraint remains the way anti‑obesity medications interact with the statutory framework of the Medicare prescription drug benefit, which historically excluded most weight‑loss drugs and still shapes coverage debates in Congress and state legislatures.
Health‑system considerations beyond the prescription
The findings also highlight operational questions for health‑system executives: how to support sustained weight management when GLP‑1 access is intermittent or financially fragile.
- Program capacity: sustainable weight management depends on access to multidisciplinary services (nutrition, physical activity support, behavioral health) that many health systems currently provide at limited scale.
- Care continuity: structured follow‑up after medication pauses-whether to monitor weight trends or to revisit therapeutic options-can reduce fragmentation across primary care, endocrinology, and obesity medicine.
- Data infrastructure: routine electronic health record capture of weights and treatment changes enables systems‑level monitoring of outcomes after discontinuation.
For boards and clinical leadership teams, the study suggests that investing in longitudinal obesity programs and data infrastructure may be as important as decisions about which GLP‑1 agent to place on formulary.
Equity and affordability pressures
Because GLP‑1 drugs now sit at the intersection of chronic disease care and high‑cost specialty pharmacy spending, their use is becoming a test case for how U.S. coverage policy handles long‑term metabolic conditions.
- Patient costs: high list prices and variable cost‑sharing create uneven access, particularly for lower‑income patients and those without robust pharmacy benefits.
- Public coverage variation: state Medicaid programs differ in obesity‑treatment coverage; employer‑sponsored plans vary widely in benefit design.
- Low surgical transition: fewer than 1% moved to metabolic/bariatric surgery within a year, underscoring barriers to procedural care and the centrality of medical and lifestyle pathways.
For policymakers focused on health equity, the data point to a risk that discontinuation and restart patterns will track income and insurance status, not solely clinical need, unless benefit design is recalibrated.
Interpreting the signal responsibly
While the findings offer reassurance against dramatic average rebound, they are not a license for complacency among clinicians or payers.
- Study design: retrospective cohort using real‑world clinical records from two states; not a randomized discontinuation trial.
- Generalizability: outcomes reflect patients engaged with a large integrated health system and may differ in other settings.
- Outcome measurement: weight trajectories in routine care can be influenced by follow‑up intensity and treatment switching.
Decision‑makers should therefore treat the analysis as an important early signal rather than a definitive guide to how long patients can safely pause therapy.
Research agenda and clinical decision‑making
“Many patients do not give up on their obesity treatment journey, even if they need to stop their initial medication,” said Dr. Gasoyan. “In our future work, we will examine the comparative effectiveness of alternative treatment options for obesity in patients who discontinue semaglutide or tirzepatide, to help patients and their clinicians make informed decisions.”
For guideline bodies, health‑plan pharmacy committees, and public payers, the next wave of evidence on comparative effectiveness and restart patterns will help determine whether GLP‑1-based care is financed as a short‑term intervention or recognized-and budgeted for-as a chronic, relapsing condition that requires long‑horizon planning.
