GLP-1 Persistence Statistics 2026: The Start-and-Stop Numbers, Honestly

By Matt · Host, Gila Podcast · Published / Last reviewed June 17, 2026 · Editor-in-chief: Sezen Soykut, ICF ACC

TL;DR GLP-1 persistence statistics 2026 tell a start-and-stop story, not a failure story. Roughly four in ten people pause within the first year, but most restart within a year or two, and one-year persistence has nearly doubled since 2021. The number that actually predicts kept loss isn't the prescription — it's the behavior built around it.

If you only read one statistic about GLP-1 medications this year, you'd probably read a scary one. The headlines have been loud: people stop, people regain, people churn. And the raw numbers, pulled out of context, do look alarming.

But I host a podcast where we read the whole study, not the press-release sentence. And when you line the 2026 persistence research up side by side, a different picture comes into focus. People aren't quitting and disappearing. They're starting and stopping and starting again — and the gap between those who regain and those who hold the line has very little to do with the molecule and a lot to do with what's built around it.

Here are the numbers, read honestly.

Key Takeaways

1. About 4 in 10 people on a GLP-1 stopped within the first year, and ~6 in 10 by year two, in a 60,000+ patient U.S. claims analysis (ENDO 2026, Sontha et al.).
2. Stopping is rarely the end: 41.5% of people who stopped restarted within a year, and 58% within two (ENDO 2026, Sontha et al.).
3. One large cohort reported 46.5% of people with type 2 diabetes and 64.8% without discontinued within a year — but that's a single 2024-era cohort, and persistence has been rising since (JAMA Network Open / PMC).
4. One-year persistence on Wegovy climbed across recent years: 33.2% → 34.1% → 39.8% → 58.6% from 2021 into the first half of 2024 (JMCP 2026).
5. In a real-world study of ~40,262 adults, 52% stopped semaglutide for weight loss within a year — 18% by three months, 31% by six (Medscape).
6. Two real-world studies put 180-day discontinuation at 30.8% and 58.2% — a wide spread that tells you "discontinuation" is not one number (Diabetes, Obesity & Metabolism).
7. Roughly 60% of the weight lost during treatment is modeled to return a year after stopping — when nothing replaces the support (ScienceDirect).
8. The leading reason people stop is financial: 47.6% cited cost or insurance, vs 14.6% side effects and 11.8% shortages (Cleveland Clinic, Obesity, 2026).
9. 37% of patients reported nausea or GI side effects, and an endocrinologist prescriber was linked to a 10% lower chance of stopping (ENDO 2026, Sontha et al.).
10. Daily steps fell from 5,047 to 4,487 and moderate-to-vigorous activity from 28 to 22 minutes/day on GLP-1s — with no sign weight loss alone increased movement (ENDO 2026, Maharjan et al.).
11. By two years, most people are no longer filling these prescriptions — roughly one in four remain on Wegovy and about one in five on Ozempic (Prime Therapeutics analysis, 2026).
12. In a real-world analysis of nearly 8,000 patients who stopped, average regain was small — the obesity group regained only about 0.5% a year later, and 45% kept losing or held steady (Cleveland Clinic, 2026).

How many people stop: year one vs. year two

Start with the most-cited number of 2026. In a U.S. claims study of more than 60,000 adults with type 2 diabetes on liraglutide, semaglutide, or tirzepatide, about four in ten stopped within the first year and roughly six in ten by year two (ENDO 2026, Sontha et al.). Here, "stopping" meant a refill gap longer than 60 days — a reasonable proxy, but a proxy.

A separate cohort of 125,474 patients found that 46.5% of those with type 2 diabetes and 64.8% of those without discontinued within a year; after accounting for censoring, that worked out to about 53.6% by one year and 72.2% by two (JAMA Network Open / PMC). I'm giving you that figure with its context attached on purpose: it's one 2024-era cohort, discontinuation is not the same as permanently quitting, and — as the next section shows — one-year persistence has roughly doubled since 2021 (JMCP 2026). Read alone, it would mislead you.

The honest summary: a meaningful share of people pause in the first year. That's real. It's also the beginning of the story, not the end.

Restart is the norm, not the exception

This is the part the scary headlines leave out. In the same 60,000-patient analysis, 41.5% of people who stopped restarted within a year, and 58% restarted within two (ENDO 2026, Sontha et al.).

So "discontinuation" is often a pause, not a goodbye. People stop for a season — a cost gap, a supply gap, a rough stretch of side effects — and then they come back. That reframes nearly every alarming stat above: a 52% one-year discontinuation rate (Medscape) does not mean 52% of people walked away forever.

One caveat worth holding: restarting after a gap may not pick up exactly where you left off. Researchers note that effectiveness may depend heavily on consistency, and inconsistent use can blunt the response on resuming (JCI Insight, via NewsNation). If a pause or restart is on your mind, that's a conversation for your prescriber.

Why people stop: cost, side effects, shortages

If you assume people quit because the medication "stopped working" or because they couldn't tolerate it, the data will surprise you. The leading reason is money. In a Cleveland Clinic study published in the journal Obesity, 47.6% of people who stopped cited financial reasons — insurance denial, an expired discount coupon, or unaffordable out-of-pocket cost — compared with 14.6% for side effects and 11.8% for shortages (Cleveland Clinic, Obesity, 2026).

Side effects are real, of course — 37% of patients reported nausea or GI symptoms in the large ENDO 2026 claims study (ENDO 2026, Sontha et al.). But the dominant driver isn't your body rejecting the drug. It's the system around the prescription: affordability, coverage, and access. That same study found patterns worth naming plainly — Medicaid and Medicare enrollees and Black patients were more likely to stop, and people prescribed by an endocrinologist were about 10% less likely to. The medication itself mattered too: tirzepatide users were 41% less likely to discontinue than liraglutide users, and semaglutide users 28% less (ENDO 2026, Sontha et al.).

If cost is the wall you keep hitting, it helps to see the real annual number for your situation before you decide anything — our cost calculator is built for exactly that.

Persistence is improving

Here's the counterweight the doom narrative never mentions. One-year persistence on Wegovy rose steadily — 33.2% in 2021, then 34.1%, then 39.8%, reaching 58.6% in the first half of 2024 (JMCP 2026). That's nearly a doubling in roughly three years.

More people are staying the course than were a few years ago. Better access, more familiarity, clearer expectations about the early side-effect window — the direction of travel is toward more people holding on, not fewer. That doesn't erase the discontinuation numbers, but it does date them. A 2021 cohort and a 2024 cohort are telling you about two different worlds.

The harder mile is still year two. By that point most people are no longer filling these prescriptions — roughly one in four remain on Wegovy and about one in five on Ozempic (Prime Therapeutics analysis, 2026). The long game remains the open question — which is exactly where behavior comes in.

What regain actually looks like

The fear underneath all of this is regain. And the modeling is sobering: about 60% of the weight lost during treatment is regained a year after stopping (ScienceDirect).

Read that carefully, though. It's a modeled trajectory for what tends to happen when the medication comes off and nothing takes its place. It is a description of an unsupported stop, not a law of physics. The medication quiets appetite signals; remove it with no other scaffolding and those signals return. The regain isn't a personal failure — it's what happens when the only lever in the system gets pulled out.

Which raises the obvious question: what happens when there's more than one lever?

Activity and muscle: the quiet variable

Movement deserves its own line, because it's easy to miss. In an analysis using Fitbit data from people who started GLP-1s, daily steps fell from 5,047 to 4,487 and moderate-to-vigorous activity dropped from 28 to 22 minutes a day. The largest declines showed up in men and in people with joint or muscle pain — and crucially, there was no evidence that weight loss on its own increased activity (ENDO 2026, Maharjan et al.).

That's a gentle nudge worth hearing: the scale moving doesn't automatically make you move more. If anything, activity can quietly drift down. Protecting movement and muscle is something you have to choose on purpose, not something the medication does for you. (If muscle is on your mind, we go deeper in our research roundup on preventing muscle loss.)

What separates regain from kept loss: behavior

This is the finding that reframes everything above. In a real-world analysis of nearly 8,000 patients who stopped semaglutide or tirzepatide, average regain a year later was small: the obesity group had lost about 8.4% of body weight and regained only around 0.5%, and 45% kept losing or stayed the same (Cleveland Clinic, 2026). What did the people who held their loss have in common? They did something next — 27% switched to another treatment, 20% restarted their original medication, and 14% continued with lifestyle support from a dietitian or exercise specialist.

That's the whole story in one sentence. The variable that separates regain from kept loss isn't whether you ever paused. It isn't your starting weight or your willpower on a hard week. It's whether there was a behavior layer underneath the prescription — habits, food awareness, movement, structure — doing work the medication can't do on its own.

This is the layer Gila is built around. The medication can quiet the noise; the habits decide what fills the quiet. When the appetite signal turns down, you get a rare window to build the patterns that hold whether or not the medication is there next year. More people are holding the line — and the ones who do tend to be the ones who treated the medication as the start of the work, not the whole of it.

What this means for the GLP-1 community

If you take one thing from these numbers, let it be this: starting and stopping is normal, regain is not a verdict, and the thing you can actually steer is the behavior around the medication.

You don't have to be perfect to be persistent. The data shows people pause and return all the time. What changes the year-two and year-three picture is whether you've built something underneath — a few habits that outlast any single prescription, a relationship with food noise that doesn't depend on the dose, movement you protect on purpose.

A few honest places to go next:

• What weight regain after stopping actually looks like in the research
• Building habits that outlast a GLP-1 prescription
• Preventing muscle loss on GLP-1 medication
• And if cost is your sticking point, run your real numbers with the cost calculator

Methodology notes

The figures here span three kinds of evidence, and it's worth knowing which is which. Claims-data analyses (ENDO 2026, Sontha et al.; JAMA Network Open / PMC; JMCP 2026; Prime Therapeutics, 2026) infer "stopping" from prescription refill gaps, which is a strong but imperfect proxy. Real-world cohort studies (Medscape; Diabetes, Obesity & Metabolism; ENDO 2026, Maharjan et al.; Cleveland Clinic, 2026) track defined patient groups over time. And the regain figure (ScienceDirect) is a modeled trajectory, not an observed outcome for any one person.

The wide spread you see — 30.8% versus 58.2% at 180 days, for instance — is exactly why no single discontinuation number should be treated as "the" number. Definitions, populations, and time windows differ. The reasons-for-stopping breakdown is from a Cleveland Clinic analysis in Obesity; the restart and side-effect figures are from the ENDO 2026 Sontha analysis. Last reviewed June 17, 2026. Gila's own pilot data on the behavior layer is not yet published, so no Gila persistence percentage is cited here.

Sources

1. ENDO 2026, Sontha et al. (Boston University / Komodo Health claims): https://www.endocrine.org/news-and-advocacy/news-room/2026/sontha-press-release-endo-2026
2. ENDO 2026, Maharjan et al. (NIH All of Us + Fitbit): https://www.endocrine.org/news-and-advocacy/news-room/2026/maharjan-press-release-endo-2026
3. JAMA Network Open / PMC — discontinuation & reinitiation of GLP-1 RAs: https://pmc.ncbi.nlm.nih.gov/articles/PMC11786232/
4. JMCP 2026 — Wegovy one-year persistence trend: https://www.jmcp.org/doi/10.18553/jmcp.2026.32.3.281
5. Medscape — real-world semaglutide discontinuation: https://www.medscape.com/viewarticle/real-world-study-finds-over-50-stop-glp-1s-within-1-year-2025a1000obm
6. Diabetes, Obesity & Metabolism — 180-day discontinuation rates: https://dom-pubs.onlinelibrary.wiley.com/doi/10.1111/dom.70579
7. ScienceDirect — modeled weight-regain trajectory after cessation: https://www.sciencedirect.com/science/article/pii/S258953702600043X
8. JCI Insight (via NewsNation) — consistency and effectiveness on restart: https://www.newsnationnow.com/health/stop-restarting-glp-1-effects/
9. Cleveland Clinic — real-world outcomes after stopping GLP-1s (n≈7,938; Diabetes, Obesity & Metabolism, 2026): https://newsroom.clevelandclinic.org/2026/03/12/what-happens-when-patients-stop-taking-glp-1-drugs-new-cleveland-clinic-study-reveals-real-world-insights
10. Cleveland Clinic — reasons for discontinuation (journal Obesity, 2026): https://consultqd.clevelandclinic.org/cost-side-effects-top-reasons-for-quitting-glp-1s-for-obesity
11. Prime Therapeutics — two-year GLP-1 persistence analysis (2026): https://www.primetherapeutics.com/w/most-people-stop-using-wegovy-and-ozempic-after-2-years-analysis-finds

Frequently asked questions

How many people stop taking GLP-1? In a U.S. claims analysis of more than 60,000 patients, about four in ten stopped within the first year and roughly six in ten by year two, where stopping meant a refill gap over 60 days (ENDO 2026, Sontha et al.). Other cohorts report higher one-year figures, but the numbers vary widely by population and definition, and one-year persistence has been rising since 2021 (JMCP 2026).

Do people gain the weight back after stopping? Modeling suggests roughly 60% of the weight lost during treatment returns within a year of stopping when nothing replaces the support (ScienceDirect). But that's a trajectory for an unsupported stop, not a guarantee — a real-world analysis of nearly 8,000 patients found average regain was small (about 0.5% in the obesity group) when people restarted, switched, or added lifestyle support (Cleveland Clinic, 2026).

How many people restart after stopping? Most do, eventually. Among people who stopped in the large claims study, 41.5% restarted within a year and 58% within two (ENDO 2026, Sontha et al.). Stopping is more often a pause than a permanent exit.

Why do most people stop? Cost and insurance lead by a wide margin — 47.6% of people who stopped cited financial reasons, compared with 14.6% for side effects and 11.8% for shortages (Cleveland Clinic, Obesity, 2026).

Is GLP-1 persistence getting better or worse? Better, at the one-year mark. Wegovy one-year persistence rose from 33.2% in 2021 to 58.6% in the first half of 2024 (JMCP 2026). Year-two persistence remains lower — roughly one in four on Wegovy and one in five on Ozempic (Prime Therapeutics, 2026) — which is where building durable habits matters most.

---

This article covers research on GLP-1 persistence. It is not medical advice. For questions about your own medication, talk to your prescriber.