GLP-1 Adherence Statistics: What the Real-World Data Actually Shows (2026)

Best GLP-1 adherence statistics to use (by what you're trying to answer)

If you only read one table on this page, read this one. It maps the most-cited statistics to the questions they actually answer.

What are the latest GLP-1 adherence statistics?

The two most current peer-reviewed real-world studies are Xie et al. (JAMA Network Open, March 2026), which found 24.5% one-year persistence and 29.6% adherence in 126,984 adults with overweight or obesity without diabetes (MarketScan 2019–2024), and Marshall et al. (JMCP, March 2026), which found 60.9% one-year persistence in 33,607 commercially insured Wegovy and Zepbound starters in 1H 2024. Same drug class, same year of publication, very different headlines — because they measure different populations and different drug eras.

• For a broad U.S. claims benchmark in obesity/no-diabetes: Use Xie 2026 (24.5% persistent, 29.6% adherent). Note that the cohort captured no Zepbound index starts because Zepbound was approved in late 2023 and the study window ended in early 2024.
• For the current high-potency Wegovy/Zepbound era: Use Marshall 2026 (60.9% in 1H 2024 starts, with Wegovy at 58.6% and Zepbound at 64.8%).

Persistence, adherence, discontinuation: read this before quoting any number

This is the #1 reason you've seen “GLP-1 adherence statistics” that seem to contradict each other. They almost never do. They just measure different things.

Persistence

“Did the patient keep taking the drug?” Most studies count someone as persistent if they don't go more than 60 or 90 days without a refill. Stricter cutoff (60 days) = lower persistence numbers. Looser cutoff (135 days) = higher numbers.

Adherence

“Did the patient take the drug consistently?” The most common measure is proportion of days covered (PDC) — the fraction of the study window the patient had medication on hand. PDC ≥0.80 (had supply at least 80% of the time) is the standard cutoff for “adherent.” A patient can be persistent (no big gap) and still non-adherent (lots of small gaps).

Discontinuation

“Did the patient stop?” This is the binary opposite of persistence. Discontinuation usually requires a longer gap — 90 days is common, 135 days is the cutoff one major JAMA Network Open paper used. Longer cutoff = lower discontinuation rate.

Switching

“Did the patient change drugs?” Switching from Wegovy to Zepbound is not the same as quitting. In the 2026 Xie study, switchers actually had higher persistence (36.4%) than non-switchers (21.4%). Treating switchers as quitters understates real persistence by a meaningful amount.

Reinitiation

“Did they come back?” Many people who stop later restart, sometimes on a different drug. Treating discontinuation as permanent gives you a misleading picture.

How many people stop taking GLP-1s within a year?

It depends on the cohort. The most-cited single number for U.S. adults with obesity and no diabetes is around 64.8% one-year discontinuation. For adults with type 2 diabetes, it's around 46.5%. That comes from Rodriguez et al. in JAMA Network Open (January 2025), a large EHR-based study that followed 125,474 patients on dual-labeled GLP-1 receptor agonists.

A different JAMA Network Open study — Do et al. 2024 — used a 135-day discontinuation cutoff (instead of the more common 90-day cutoff) and found: 26.2% discontinued at 3 months, 30.8% at 6 months, and 36.5% at 12 months. Overlapping time periods. Same drug class. Different definitions. Different numbers.

The Master Comparison Table: every major GLP-1 adherence study, side by side

Assembled from the major peer-reviewed studies, payer reports, and institutional analyses into a single dataset with sample size, population, drug class, time horizon, definitions, and key findings.

Blue rows = most current peer-reviewed broad U.S. studies. Last verified May 7, 2026.

Why the numbers vary so widely

Five things drive nearly all of the variation. Hold these in your head and you'll be able to read any GLP-1 study without getting lost.

1. The gap definition. A 60-day gap classifies more people as “discontinued” than a 135-day gap. Same patients, same prescriptions — different numbers, just because the cutoff moved.
2. The indication. Adults using a GLP-1 for type 2 diabetes have substantially better persistence than adults using one for obesity. Rodriguez 2025 reported 46.5% one-year discontinuation with T2D vs. 64.8% without. Insurance coverage is one of the clearest likely drivers — diabetes is covered by almost every plan; obesity is not.
3. The drug era. A 2021 cohort started on Saxenda or early Wegovy during massive shortages with limited coverage. A 2024 cohort started on Zepbound or fully-supplied Wegovy with broader employer coverage. Treating those as the same population gives you a misleading average.
4. The data source. Claims data miss cash-pay fills, samples, and compounded products. EHR data miss prescriptions filled outside the health system. Surveys reflect self-report. Each source captures a different slice of reality.
5. The follow-up window. A 4-week dropout statistic and a 3-year persistence statistic answer different questions. Earlier Prime research had already shown poor short-term persistence: in a one-year real-world study from 2023, just 32% of members remained on therapy. A follow-up study found only 15% persisted for two years.

Year-over-year improvement: the story most articles miss

GLP-1 one-year persistence has improved dramatically across recent cohorts. Marshall et al. (JMCP 2026) — the most current peer-reviewed analysis — found one-year persistence on Wegovy and Zepbound nearly doubled from 33.2% in 2021 starts to 60.9% in 1H 2024 starts.

Source: Marshall et al., JMCP 2026 (n=33,607). Last verified May 7, 2026.

Most articles still quote the 2021 numbers — 32%, 33%, 8.1% at three years — because those are the headline-grabbing low figures and they reflect the longest available data. But citing only those numbers in 2026 is like citing 2020 mortgage rates today.

GLP-1 adherence by drug

Drug-by-drug persistence is heavily confounded by indication, year, coverage, and patient selection — but some patterns are consistent.

Wegovy (semaglutide 2.4 mg)

• 2021 cohort: 33.2% one-year persistence
• 1H 2024 cohort: 58.6% one-year persistence (Marshall JMCP 2026)
• 3-year persistence (older 2021 cohort): 14%
• Reaching the recommended 2.4 mg dose: Only 13% of Wegovy users in a 110,748-patient Danish study reached 2.4 mg by their fifth prescription. Only 5.7% stopped after the first prescription (Ladebo et al., Diabetes Care 2024).

Zepbound (tirzepatide)

• 2023 cohort: 64.0% one-year persistence
• 1H 2024 cohort: 64.8% one-year persistence (Marshall JMCP 2026)
• Adverse-event discontinuation in SURMOUNT-1: 4.3% at 5 mg, 7.1% at 10 mg, 6.2% at 15 mg (vs. 2.6% on placebo)

Ozempic (semaglutide, T2D label)

• 2021 cohort: 47.1% one-year persistence — the highest of the older drugs
• Ozempic shows up in obesity studies because clinicians prescribed it off-label during the Wegovy shortage.

Mounjaro (tirzepatide, T2D label)

Limited dedicated obesity-adherence data. Don't treat Mounjaro and Zepbound adherence as interchangeable unless the study explicitly groups tirzepatide by brand or indication.

Saxenda (liraglutide, daily injection)

• 2021 cohort: 19.2% one-year persistence — the lowest of the major obesity GLP-1s. The daily injection burden is the leading hypothesis.

Oral GLP-1s (Rybelsus, Wegovy tablets, Foundayo)

• Foundayo (orforglipron) is an oral once-daily GLP-1 pill for weight management FDA-approved April 1, 2026; unlike other oral GLP-1s, it does not need to be taken on an empty stomach.
• Real-world adherence data for oral weight-management GLP-1s is still emerging; comparisons to injectables won't be reliable until 2027.

Compounded GLP-1s

Persistence data is fragmentary because most compounded use is cash-pay and outside insurance claims. The FDA's 2024–2025 actions ended the shortage-based basis for routine compounding of semaglutide and tirzepatide after enforcement-discretion periods.

Why people stop taking GLP-1s

The most consistently documented reason for stopping a GLP-1 in U.S. clinical practice is cost or insurance — about 47.6% of patients in the largest chart-review study cited it as the primary reason. Side effects are second at 14.6%, and drug shortages account for 11.8%.

Different methods produce different breakdowns. Here are the major ones side by side.

Side effects: trial rates vs. real-world rates

In randomized clinical trials, only about 4–10% of GLP-1 patients discontinued because of side effects. In real-world practice, side effects account for roughly 14–28% of stoppages.

Specific clinical-trial discontinuation rates for adverse events:

• Wegovy adult weight-reduction trials: 4.3% of Wegovy injection-treated adults permanently discontinued because of a gastrointestinal adverse reaction, vs. 0.7% on placebo (Wegovy FDA prescribing information)
• SURMOUNT-1 (tirzepatide): 4.3% at 5 mg, 7.1% at 10 mg, 6.2% at 15 mg vs. 2.6% on placebo. Pooled SURMOUNT-1 to -4: 1.0% to 10.5% discontinued due to GI adverse events.

What patients actually experience in trials:

• Wegovy adult trials: 73% reported gastrointestinal adverse reactions
• SURMOUNT-1 (tirzepatide): nausea 24.6–33.3%; diarrhea 18.7–23.0%; vomiting 8.3–12.2% (rates rose with higher doses)

GI side effects are common, especially during dose escalation. Most patients ride them out. The clinical-trial environment — structured titration support, no cost barrier, no prior auth — makes it much easier to do so.

Cost and insurance: the adherence killer most articles underplay

Cost is the single largest documented reason U.S. adults stop taking GLP-1 drugs for obesity, and it hits Medicaid and Medicare patients harder than commercial-insurance patients.

• KFF Health Tracking Poll, November 2025: 56% of GLP-1 users said the drugs were difficult to afford; 27% reported paying the full cost out-of-pocket.
• Cleveland Clinic by payer (primary reason for stopping): Medicaid 76.5%, Medicare 61.0%, Private 43.0%, Self-pay 44.4% cited cost.

What's actually happening with Medicare and Medicaid in 2026

A coupon cliff is a real adherence event. Manufacturer savings programs have eligibility rules, expiration dates, and monthly limits that change. Treat coupon expiration as an adherence risk, but verify current program terms before quoting a specific timeline.

The dose-escalation trap

A surprising amount of GLP-1 adherence trouble comes from how the drugs are titrated — not from the drugs themselves. In a Danish study of 110,748 Wegovy users (Ladebo et al., Diabetes Care 2024):

• Only 13% reached the maximum dose of 2.4 mg by their fifth prescription
• Only 5.7% stopped after the first prescription
• Only 10% followed recommended dose increases every 4 weeks
• Overall, 25% filled at least one prescription of 2.4 mg, while 33–48% continued with the 1.0-mg dosage from the fourth prescription onward

What happens to weight when you stop

The strongest trial evidence says weight tends to come back substantially after stopping. In the STEP 1 trial extension (semaglutide 2.4 mg), participants regained on average about two-thirds of their lost weight in the 12 months after stopping. In SURMOUNT-4 (tirzepatide), only 17% of withdrawn participants kept ≥80% of their weight loss after 52 weeks off-treatment.

From the clinical trial extensions

• STEP 1 extension (Wilding et al., 2022): Mean weight loss was 17.3% with semaglutide through week 68. After treatment withdrawal — including withdrawal of lifestyle support — semaglutide participants regained 11.6 percentage points of weight by week 120, resulting in a net loss of 5.6% from baseline. Patients had regained about two-thirds of what they lost.
• SURMOUNT-4 (Aronne et al., 2024): Patients on tirzepatide for 36 weeks lost a mean of 20.9%. Then half were randomized to placebo. Withdrawal was followed by a mean weight regain of 14% over the next 52 weeks, while continuation resulted in additional weight loss of 6.7%. Only 17% of withdrawn patients maintained at least 80% of their weight loss.
• SURMOUNT-4 cardiometabolic post-hoc (Horn et al., 2025): Patients regaining ≥25% of weight saw greater reversal of cardiometabolic gains compared with patients who maintained weight loss.

Real-world post-stopping outcomes

The Cleveland Clinic followed 7,938 real-world discontinuers and found something more nuanced: 55% of obesity patients gained weight in the year after discontinuation, while 45% maintained or kept losing. In the diabetes group, 44% gained weight and 56% maintained or kept losing. Real-world stoppers do better than trial extensions because many restart, switch to another drug, or use other interventions.

Switching is not the same as quitting

Roughly 1 in 5 GLP-1 users switches between drugs within the first year — and switchers actually have higher persistence than non-switchers. From the Xie 2026 paper (n=126,984 patients):

• 12-month persistence across entire cohort: 24.5%
• Switchers (n=26,197): 36.4% persistent; mean PDC 63.0%; 33.3% adherent (PDC ≥80%)
• Non-switchers (n=100,787): 21.4% persistent; mean PDC 52.0%; 28.6% adherent

Why would switchers persist better? A switch usually means active care — the patient is engaged, the prescriber is engaged, and someone is making decisions about therapy. Patients who get past the friction of a switch are demonstrating the kind of persistence that predicts staying on therapy.

Reinitiation rates tell the same story. From Rodriguez et al. 2025: 47.3% of patients with type 2 diabetes and 36.3% without restarted a GLP-1 RA within one year of stopping. So when you read “64.8% of obesity patients discontinued at one year,” remember: about 36% of those discontinuers came back within another year.

How adherence affects real-world weight loss

Real-world weight loss tracks adherence almost linearly. From Gasoyan et al. 2025 (Cleveland Clinic, n=7,881):

The gap between the trial weight loss numbers everyone quotes (15–21%) and the typical real-world average is mostly an adherence and dosing gap, not a drug efficacy gap. The drugs work. The system around them frequently doesn't deliver them at full dose for a full year.

Who is most likely to stay on a GLP-1 (and who isn't)

Higher persistence is consistently associated with: type 2 diabetes (vs. obesity-only), higher household income, age 40+, female sex, prescription by an endocrinologist or obesity specialist, weight-related comorbidities, and starting in 2024.

• By indication: T2D patients persist ~18 pp better than obesity-only at one year (46.5% vs. 64.8% discontinuation, Rodriguez 2025).
• By payer (share citing cost as primary reason for stopping): Medicaid 76.5%, Medicare 61.0%, Private 43.0%, Self-pay 44.4%.
• By age (MassHealth): Members younger than 40 were less likely to be persistent (58.9%) or adherent (57.6%) at 6 months vs. 40+ (62.6% persistent, 62.5% adherent).
• By sex (MassHealth): Women more likely than men to be persistent (61.7% vs. 56.8%) and adherent (60.7% vs. 57.3%).
• By GI/psychiatric history (Mailhac Danish 2025): Patients with prior GI medication: +9% discontinuation risk. Prior psychiatric medication: +12%.
• By prescriber type (Blue Health Intelligence): Endocrinologists and obesity specialists had higher 12-week completion rates than primary care.
• By initiation year: A 2024 starter has roughly twice the one-year persistence of a 2021 starter, even before adjusting for patient-level factors.

Trial adherence vs. real-world adherence: why the gap is structural

Clinical trial completion rates above 85% reflect a fundamentally different setup than real-world prescribing. The 50–65% one-year discontinuation in older real-world cohorts isn't a sign the drugs don't work — it's a sign the system around them often doesn't support sustained use.

What trials provide that real life doesn't:

• Medication at no cost to participant
• Pre-screening that excludes patients with severe GI conditions or recent psychiatric medication changes
• Dedicated study coordinators tracking each patient
• Slow, structured dose escalation with clinical support
• No insurance prior authorization process
• No supply shortages
• A defined endpoint, after which the patient knows they're done

What actually improves GLP-1 adherence

The best-supported adherence levers are the ones that show up repeatedly in the data.

• Affordability stability. Cost is the largest documented driver of stopping, and it's the most fixable. Plans that cover obesity drugs without an annual maximum or improvement clause plausibly retain more members long-term — but measure your own population to be sure.
• Specialist prescribing. Patients prescribed by endocrinology or obesity-medicine clinicians have higher 12-week completion rates. Specialists tend to know how to manage GI side effects and how to titrate slowly.
• Frequent early follow-up. Patients who saw their provider more often during the first 12 weeks were more likely to persist (Blue Health Intelligence). Side-effect management is the time-sensitive piece.
• Slow titration. The Wegovy label permits delaying dose escalation by four weeks if the patient isn't tolerating the current dose. The Ladebo Danish data shows only 10% of real-world users actually follow recommended four-week steps.
• Supply continuity. A patient who can't fill their next prescription is at high risk of falling off therapy entirely. Resolved shortages explain a meaningful share of the 33%-to-61% persistence improvement from 2021 to 2024.
• Setting expectations. The STEP 1 extension and SURMOUNT-4 data make clear that GLP-1s require ongoing use to maintain weight loss. Patients who understand that are more likely to plan around it — to budget for it, build it into their lives, and not stop after reaching a goal weight.

How to use these statistics

Which GLP-1 adherence statistic should I cite?

Pick the statistic that matches your population, metric, time horizon, and drug era. Use this decision tree.

1. Are you talking about obesity/weight-loss use, T2D use, or both?
   • Obesity, no diabetes → Marshall JMCP 2026 or Xie JAMA Net Open 2026
   • Type 2 diabetes → Rodriguez JAMA Net Open 2025 T2D arm
   • Both/mixed → Do JAMA Net Open 2024
2. What metric do you actually need?
   • Staying on therapy → persistence
   • Days covered → adherence (PDC)
   • Stopping for >X days → discontinuation
   • Restarting after stopping → reinitiation
   • Changing drugs → switching
3. What time horizon?
   • 4 weeks → Blue Health Intelligence (BCBSA 2024)
   • 12 months → Marshall 2026, Xie 2026, Rodriguez 2025
   • 3 years → Prime Therapeutics 2025
4. What drug era?
   • 2024 high-potency → Marshall 2026
   • 2021–2023 → Gleason JMCP 2024 or earlier Prime cohorts
5. What source quality do you need?
   • Peer-reviewed → for academic citation
   • Payer/industry → for benefits planning
   • Survey → for patient-perception framing
   • Clinical-trial extension → for biological mechanism

Limitations of the current evidence

Most GLP-1 adherence statistics come from observational claims data, EHR records, or surveys — none of which capture the full picture.

• Cash-pay GLP-1 use is invisible to insurance claims. A meaningful share of compounded semaglutide and tirzepatide moved through cash channels during 2023–2024 shortages.
• Sample medication isn't recorded in pharmacy claims. Some early use is undercounted.
• Out-of-network EHR fills mean a patient might appear to have stopped in their primary health system's record while actually continuing therapy elsewhere.
• Switching to compounded during shortages then back to branded creates “discontinuation” signals that aren't real treatment cessation.
• Patient-reported reasons in surveys reflect what people remember and what they're willing to say.

The patterns that show up across studies — cost dominates, persistence falls steeply year one, weight regain after stopping is real, switching ≠ quitting, 2024 cohorts persisting better than 2021 — are robust to any one source's limitations.

What's coming in 2026 and beyond

Three structural shifts are likely to change GLP-1 adherence rates over the next several years:

• Medicare GLP-1 Bridge and Medicaid BALANCE. This is the single largest near-term policy change affecting adherence. Cost-driven discontinuation in Medicare patients (currently 61% of stoppers) should fall sharply for beneficiaries who qualify under Bridge and pay the $50 copay. Whether Medicaid BALANCE delivers similar improvements depends on which states and manufacturers participate.
• Orforglipron (Foundayo). Oral once-daily eliminates injection burden, and not requiring an empty stomach removes a meaningful adherence friction. Whether the adherence pattern matches injectables, beats them, or underperforms them is the open question. Real-world adherence will tell the rest of the story by 2027.
• 2024 cohort maturation. Did the 60.9% one-year persistence hold to year 2? Year 3? If yes, the long-term picture for the current era is much better than the 8.1%-at-3-years number from older cohorts. We'll know in 2026 and 2027 when those cohorts mature.
• Compounded GLP-1 enforcement. With the FDA's 2024–2025 actions ending the shortage-based basis for routine semaglutide and tirzepatide compounding, patients who relied on compounded versions for affordability are now facing the choice of full-price branded or stopping.

Frequently asked questions

How we built this page

Sources

1. Xie L, et al. Glucagon-Like Peptide-1 Receptor Agonist Switching and Treatment Persistence in Adults Without Diabetes. JAMA Netw Open. 2026;9(3):e261272.
2. Marshall LZ, et al. Trends in 1-year persistence and adherence among initiators of high-potency, weight loss–indicated glucagon-like peptide 1 receptor agonists. J Manag Care Spec Pharm. 2026;32(3):281.
3. Rodriguez PJ, et al. Discontinuation and Reinitiation of Dual-Labeled GLP-1 Receptor Agonists Among US Adults With Overweight or Obesity. JAMA Netw Open. 2025;8(1):e2457349.
4. Do D, et al. GLP-1 Receptor Agonist Discontinuation Among Patients With Obesity and/or Type 2 Diabetes. JAMA Netw Open. 2024;7(5):e2413172.
5. Gleason PP, et al. Real-world persistence and adherence to glucagon-like peptide-1 receptor agonists among obese commercially insured adults without diabetes. J Manag Care Spec Pharm. 2024;30(8):860-867.
6. Prime Therapeutics. Three-Year Persistence to GLP-1 Therapy for Obesity Among Members Without Diabetes. 2025 industry report.
7. Blue Health Intelligence / Blue Cross Blue Shield Association. Real-world trends in GLP-1 treatment persistence and prescribing for weight management. May 2024 issue brief.
8. Gasoyan H, et al. Reasons for Discontinuation of Obesity Pharmacotherapy With Semaglutide or Tirzepatide. Obesity (Silver Spring). 2025;33(12):2296-2303.
9. Gasoyan H, et al. Changes in Weight and Glycemic Control Following Obesity Treatment by Discontinuation Status. Obesity (Silver Spring). 2025;33:1657-1667.
10. Cleveland Clinic Newsroom. What Happens When Patients Stop Taking GLP-1 Drugs? March 12, 2026.
11. Mailhac A, et al. Discontinuation of semaglutide therapy for weight loss. EASD 2025 Annual Meeting (conference presentation).
12. Ladebo L, et al. Real-World Use of Semaglutide for Weight Management: Dose Titration — A Danish Cohort Study. Diabetes Care. 2024;47(10):1834-1837.
13. Wilding JPH, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: The STEP 1 trial extension. Diabetes Obes Metab. 2022;24(8):1553-1564.
14. Aronne LJ, et al. Continued Treatment With Tirzepatide for Maintenance of Weight Reduction: The SURMOUNT-4 RCT. JAMA. 2024;331(1):38-48.
15. Horn DB, et al. Cardiometabolic Parameter Change by Weight Regain on Tirzepatide Withdrawal: Post Hoc Analysis of SURMOUNT-4. JAMA Intern Med. Published online November 24, 2025.
16. Truveta Research / ISPOR 2025 poster summary. Real-world temporal and indication-specific variation in drivers of GLP-1 RA discontinuation.
17. Weiss T, et al. Real-World Adherence and Discontinuation of GLP-1 RA Therapy in T2DM Patients in the United States. Patient Prefer Adherence. 2020;14:2337-2345.
18. KFF Health Tracking Poll: Prescription Drug Costs, Views on Trump Administration Actions, and GLP-1 Use. Kaiser Family Foundation, November 2025.
19. Wegovy (semaglutide) Prescribing Information. Novo Nordisk Inc. Revised 2026.
20. Zepbound (tirzepatide) Prescribing Information. Eli Lilly and Company. Revised 2025.
21. CMS. Medicare GLP-1 Bridge program, 2026.
22. CMS. BALANCE model. CMS Innovation Center, 2026.
23. FDA. FDA Approves First New Molecular Entity Under National Priority Voucher Program (orforglipron). April 1, 2026.
24. FDA. FDA clarifies policies for compounders as national GLP-1 supply begins to stabilize.

More GLP-1 research from The RX Index

• GLP-1 Alcohol Cravings: What the 2026 Evidence Actually Says
• GLP-1 Hard Contraindications: Who Should Avoid
• Is GLP-1 Safe? 16 Checks Before You Start
• Can You Drink Alcohol on GLP-1 Medications?
• Wegovy Pill vs. Injection: Cost & Results