Weight & Metabolic
How Long Do People Actually Stay on GLP-1s?
The short answer
Considerably less long than clinical trials suggest. In a large US commercial insurance claims analysis, only 32.3% of people starting a GLP-1 for obesity were still on it at one year, and persistence varied enormously by drug — 47.1% for semaglutide versus 19.2% for liraglutide in the same dataset. In a German digital weight-loss program, 88% of patients had dropped out of the medication-and-tracking protocol within 12 months, even with structured support. A US academic obesity clinic offering the drugs at no cost to patients still saw a median persistence of well under a year. Every real-world cohort checked lands in roughly the same place: most people don't stay on these drugs as long as the trials that established their benefits.
Quick answers
- What is it?
- How long people actually stay on GLP-1 medications in real-world use.
- Why are people interested in it?
- People starting these drugs want to know if the trial-level weight loss results are realistic for someone who doesn't stay on the drug as long as trial participants did.
- What has actually been studied?
- A US commercial insurance claims analysis of 4,066 patients, a no-cost academic obesity clinic cohort of 2,306 patients, a German digital weight-loss program of 4,535 patients, and real-world adherence data from Israel and a US electronic health record database.
- How strong is the evidence?
- Considerably less long than clinical trials suggest — real-world cohorts consistently show most people stop within a year, even when cost is removed as a barrier.
- What don't we know yet?
- Why people actually stop, since claims data can't say, what happens to people after they stop, and whether persistence improves or worsens as lower-cost formulations become available.
- Where can I read the original research?
- Sources at the bottom of this page: Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes and Real-world titration, persistence & weight loss of semaglutide and tirzepatide in an academic obesity clinic.
The short answer: Not nearly as long as the pivotal trials — with their fixed treatment schedules and built-in follow-up — might imply. In a large US commercial claims analysis, only about a third of people who started a GLP-1 for obesity were still taking it a year later, and that number swung wildly depending on which specific drug they started. A German digital weight-loss program saw 88% of its patients drop out of the full protocol within a year, even with structured tracking support. A US academic clinic that removed cost as a barrier entirely still saw median persistence measured in single-digit months for many patients. Every real-world cohort examined converges on the same basic conclusion, even though the exact numbers differ: staying on these drugs long-term is the exception, not the norm, once people leave a clinical trial setting.
What the research shows
The baseline number, from US commercial insurance
A 2024 analysis used integrated pharmacy and medical claims data covering 16.5 million average monthly members to identify 4,066 obese adults without diabetes who newly started a GLP-1 for obesity treatment in 2021, each with a full year of enrollment before and after. Overall persistence — defined as no gap in therapy of 60 days or more — was 46.3% at 180 days and 32.3% at one year. Average proportion of days covered across the year was 51.0%, with only 27.2% of patients meeting the stricter 80%-or-higher adherence threshold. Just 11.1% switched to a different GLP-1 product rather than stopping altogether. The drug-by-drug breakdown was stark: one-year persistence was highest for semaglutide (Ozempic) at 47.1% and lowest for liraglutide (Saxenda) at 19.2%, within the same dataset and time period.
Removing cost didn't fix it
A 2025 study followed 2,306 patients at an academic medical center's multidisciplinary obesity clinic, all enrolled in a "no cost to patient" prescription bundle for semaglutide or tirzepatide between 2022 and 2024. Persistence was defined as continuous prescription fills without a gap of 84 days or more. Median persistence across the cohort was 10.7 months (interquartile range 5.4 to 16.3 months). Among the subset who stayed on therapy for at least 6 months, median weight loss was 9.4%; among those who made it to 12 months, it was 14.4% — figures that approximate what the pivotal trials reported. But reaching those milestones required staying on the drug, and the median patient in this no-cost program didn't clear a year.
A digital program with structured support
A 2026 retrospective cohort followed 4,535 patients in a semaglutide-supported digital weight-loss service in Germany, tracking both medication adherence and program engagement over 12 months. Only 12% of the total cohort (546 patients) remained adherent to both the medication and data-submission protocols for the full year. Mean 12-month weight loss in that adherent minority was 15.13% — again, close to trial-level results. But modeling the full initiating cohort under a worst-case scenario that assumed zero weight loss for anyone who dropped out brought the population-level average down to 2.40%. The authors described the program's population-level impact as "fundamentally limited by high attrition (88%) in an unsubsidised environment."
Persistence differs by dosing schedule too
A 2024 nationwide real-world analysis from Israel's Clalit Health Services examined 70,654 patients with type 2 diabetes who purchased any GLP-1 receptor agonist between 2009 and 2021. Once-weekly formulations showed significantly better adherence than daily ones — an adjusted odds ratio of 1.25 for good adherence (proportion of days covered ≥80%) — and were associated with larger HbA1c reductions. Good adherence in this cohort was also more common among older patients, women, and those of higher socioeconomic status, pointing to persistence being shaped by more than just the drug itself.
A different database, similar pattern for tirzepatide specifically
A 2025 analysis of the Veradigm electronic health record database identified individuals without type 2 diabetes who received at least one tirzepatide prescription, then followed an "anti-obesity medication-eligible" subgroup with 6 months of continuous enrollment data. At 6 months, adherence was 55.5% and persistence was 54.2% — better than the 1-year commercial claims figures above, but measured over half the time window, and starting from a group that had already filled at least one prescription and stayed enrolled.
What we don't know yet
Why people actually stop. Claims and EHR data can measure that a prescription wasn't refilled, but not whether the reason was cost, side effects, feeling the goal was reached, a supply shortage, a life event, or something else. Surveys suggest cost and side effects both play large roles, but the studies with hard persistence numbers generally can't attribute individual reasons.
What happens to people after they stop. Persistence studies end at the point of discontinuation. Whether people who drop off later restart, switch to a different weight-management approach, or regain weight and stay off medication entirely isn't tracked within these same datasets.
Whether newer prescribing patterns or lower-cost access channels change the numbers going forward. Most of this data comes from 2021–2024, before wider availability of compounded and lower-cost formulations; whether persistence improves or worsens as the drug landscape shifts isn't yet measured.
A standardized definition of "persistence" across studies. Every study in this space picks its own allowable gap length (60 days, 84 days, others) and its own adherence threshold, which makes the headline percentages hard to compare directly even when they're pointing in the same general direction.
Long-term health outcomes for the majority who don't persist. Given that most people in these cohorts don't make it to a year, and weight regain after discontinuation is well documented separately, there's limited published data connecting real-world persistence patterns directly to real-world long-term health outcomes.
Sources
- Gleason PP, Urick BY, Marshall LZ, et al. Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes. Journal of Managed Care & Specialty Pharmacy, 2024;30(8):860–867. DOI
- Samuels JM, Ye F, Irlmeier R, et al. Real-world titration, persistence & weight loss of semaglutide and tirzepatide in an academic obesity clinic. Diabetes, Obesity & Metabolism, 2025;27(11):6200–6209. DOI
- Talay L, Hom J, Tan M, Ahuja N. Real-World Effectiveness and 12-Month Persistence of a Semaglutide-Supported Digital Weight-Loss Service: A Retrospective Cohort Study in Germany. Diabetes, Obesity & Metabolism, 2026;28(8):6985–6996. DOI
- Kassem S, Khalaila B, Stein N, Saliba W, Zaina A. Efficacy, adherence and persistence of various glucagon-like peptide-1 agonists: nationwide real-life data. Diabetes, Obesity & Metabolism, 2024;26(10):4646–4652. DOI
- Hunter Gibble T, Chinthammit C, Ward JM, et al. Real-world use of tirzepatide among individuals without evidence of type 2 diabetes: Results from the Veradigm® database. Diabetes, Obesity & Metabolism, 2025;27(6):3185–3194. DOI
Article records maintained via PubMed. Peptide Proof reports what studies found. It does not provide medical advice, dosing guidance, or treatment recommendations.
Topic: Weight & Metabolism | Evidence Index: semaglutide · tirzepatide · liraglutide
Sources
- Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes
- Journal of Managed Care & Specialty Pharmacy, 2024 · n=4066 · DOI
- Real-world titration, persistence & weight loss of semaglutide and tirzepatide in an academic obesity clinic
- Diabetes, Obesity & Metabolism, 2025 · n=2306 · DOI
- Real-World Effectiveness and 12-Month Persistence of a Semaglutide-Supported Digital Weight-Loss Service: A Retrospective Cohort Study in Germany
- Diabetes, Obesity & Metabolism, 2026 · n=4535 · DOI
- Efficacy, adherence and persistence of various glucagon-like peptide-1 agonists: nationwide real-life data
- Diabetes, Obesity & Metabolism, 2024 · n=70654 · DOI
- Real-world use of tirzepatide among individuals without evidence of type 2 diabetes: Results from the Veradigm database
- Diabetes, Obesity & Metabolism, 2025 · DOI
Get the evidence, not the hype.
New research briefs, plainly written and fully sourced. No dosing advice, ever.