Research Lab Review

Real-world data: most stop semaglutide or tirzepatide within a year

News · Tirzepatide

Two large real-world cohorts find most adults who start semaglutide or tirzepatide for obesity stop within a year, often before weight is recorded.

Two large records-based studies published this year find that most adults who start semaglutide or tirzepatide for obesity are no longer on it a year later [1, 2]. In the larger study, about half of people who started treatment had no weight recorded at 12 months [1]. Together, the findings help explain why weight loss in routine care can look smaller than in randomized trials, which run under structured follow-up [1].

Key takeaways

  • In a US records study of 743,153 adults with obesity and without diabetes, 29.9% of semaglutide starters and 35.9% of tirzepatide starters were still being prescribed the drug at 12 months. This study is a preprint and has not yet been peer reviewed [1].
  • In an Israeli cohort of 34,309 adults without diabetes who started semaglutide, 67.6% had stopped within a year [2].
  • About half of the US starters had no weight on record at 12 months. Documented weight loss was most common among people who stayed on therapy, reached a maintenance dose and were weighed [1].
  • Both studies are observational. They describe patterns and associations, not causes, and neither records why people stopped [1, 2].

Microscope view of adipose tissue showing clusters of fat cells (adipocytes)

The US study: a care cascade across 350+ health systems

The first study is a retrospective cohort study. It drew on an electronic health record network covering more than 350 US health systems [1]. The researchers identified adults with obesity and without diabetes who started injectable semaglutide or tirzepatide in 2024 and followed them for 12 months [1]. Of 743,153 people, 60.1% started semaglutide. Their mean age was 48.5 years, and 72.5% were women [1]. The paper was posted on a preprint server and has not been peer reviewed, so its figures could change [1].

The main measure was “prescribing persistence”, meaning whether a person was still being prescribed the drug. At 12 months, 29.9% of semaglutide starters and 35.9% of tirzepatide starters were persistent [1]. The researchers also tracked whether people reached a maintenance dose as defined in the US Food and Drug Administration label. Reported rates were 28.3% for semaglutide and 55.0% for tirzepatide [1]. Only 8.8% and 16.4%, respectively, were still persistent at a maintenance dose with a weight recorded at 12 months [1].

The authors arranged the results as a “care cascade”: each step narrows the group to people who completed the step before it [1]. Across all starters, at least 5% total body weight loss was documented in 25.6% of the semaglutide group and 30.4% of the tirzepatide group [1]. These shares rose at each step:

  • 51.9% and 61.9% among people with a recorded weight [1].
  • 66.1% and 79.5% among those who were also still being prescribed the drug [1].
  • 82.9% and 87.3% among those who were also at a maintenance dose [1].

The authors suggest that continuation, dose escalation and monitoring are targets for improving care [1].

Limits. The word “documented” matters here. A missing weight in the record does not mean a person lost no weight; it may only mean no one entered a measurement [1]. Persistence was measured from prescriptions, not from whether people filled or took them [1]. Each step of the cascade is also more selected than the last. People who respond well may be more likely to keep going, so the higher rates further down do not by themselves show that continuing causes more weight loss [1]. Finally, people were not randomly assigned to semaglutide or tirzepatide, so differences between the two groups may reflect who received each drug [1].

The Israeli study: who stops, and when

The second study is also a retrospective cohort. It used records from a large Israeli health fund and included adults without diabetes who started semaglutide for obesity between January and May 2025 [2]. Stopping was defined as a gap of more than 60 days after medication coverage ended [2]. The cohort numbered 34,309 people: 67.0% were women, and the mean age was 44.6 years [2].

Cumulative discontinuation reached 36.5% at 180 days and 67.6% at 365 days [2]. Several factors were linked to a higher chance of stopping:

  • Belonging to the Arab population sector (hazard ratio 1.65) [2].
  • Low socioeconomic status (hazard ratio 1.13) [2].
  • Current smoking (hazard ratio 1.13) [2].

The researchers also ran a separate analysis of 28,218 people who were still on treatment at day 90 [2]. In this group, still being on the lowest starting dose at day 90, rather than a higher dose, was linked to later stopping (hazard ratio 1.88) [2]. That link weakened substantially once the analysis accounted for how consistently people had refilled early on [2]. Early weight response also mattered: each percentage point less of BMI reduction was associated with a 3% higher hazard of stopping [2].

Tirzepatide became covered by the health fund in December 2025. Over a median of 3.1 months after that, 2,908 of 33,936 eligible participants (8.6%) started it [2]. People who started tirzepatide tended to have higher socioeconomic status and higher baseline BMI [2]. The authors say the socioeconomic differences suggest inequalities in access to sustained obesity medication [2].

Limits. The persistence results in this study cover semaglutide only, and the tirzepatide follow-up was short [2]. Like the US study, it is observational, so the factors linked to stopping are associations, not proven causes [2]. The records also do not say why people stopped. Side effects, cost, supply or personal choice cannot be separated [2]. The findings come from one country’s health system and may not apply elsewhere [2].

Why real-world results can differ from trials

Randomized trials of these drugs have reported clear results under protocol-driven care. A recent review summarizes several of them [3]:

  • SELECT reported a 20% reduction in major adverse cardiovascular events with semaglutide in people with obesity and established cardiovascular disease but without diabetes.
  • SUMMIT reported a 38% reduction in cardiovascular death or worsening heart failure with tirzepatide.
  • SURPASS-CVOT found tirzepatide noninferior, but not superior, to dulaglutide for major cardiovascular events.

An exploratory analysis of SURPASS-CVOT looked at people with type 2 diabetes and atherosclerotic cardiovascular disease, followed for an average of 4 years [4]. Among those with heart failure, it reported a hazard ratio of 0.83 for all-cause death or heart failure events with tirzepatide versus dulaglutide [4]. Because this was an exploratory subgroup analysis, it generates hypotheses rather than confirming an effect [4].

These trials enrolled specific groups, such as people with diabetes or established heart disease, and kept them under structured follow-up [3, 4]. The two cohorts describe a different setting: adults without diabetes in everyday care, most of whom stopped within a year [1, 2]. The US authors point to this gap between structured trial care and routine practice as the motivation for their study [1]. None of these four papers tests whether trial outcomes carry over to people who stop early. The review also notes that gastrointestinal side effects remain a concern with this drug class [3], but neither cohort recorded side effects as a reason for stopping [1, 2].

What we still don’t know

  • Why people stop. Neither cohort captured the reasons. Side effects, cost, supply problems, insurance changes and personal choice all remain possible explanations [1, 2].
  • What happens after stopping. Neither study reports weight or health outcomes in people after they discontinued [1, 2].
  • What missing weights hide. It is unclear how much of the gap in documented weight loss reflects less weight loss and how much reflects missing records [1].
  • Whether the US findings hold up. The US study has not yet been peer reviewed, and its numbers could change [1].
  • Long-term tirzepatide persistence. The Israeli data on tirzepatide cover only early uptake [2].
  • Whether persistence can be improved. Neither study tested ways to help people stay in care, such as more follow-up visits or support with dose escalation [1, 2].
  • How far the results apply. Both studies come from specific health systems, and patterns may differ in other countries or insurance settings [1, 2].

References

  1. Post-Initiation Care Patterns and Weight Loss With Semaglutide and Tirzepatide for Obesity in Clinical Practice. medRxiv : the preprint server for health sciences (2026). doi:10.64898/2026.10.04.26364715
  2. Early Treatment Trajectory and Long‐Term Persistence to Semaglutide for Obesity: A Population‐Based Cohort Study. Diabetes, Obesity and Metabolism (2026). doi:10.1111/dom.71488
  3. GLP-1-based therapies in cardiovascular disease: outcomes, mechanisms, and emerging clinical applications. European Journal of Clinical Pharmacology (2026). doi:10.1007/s00228-026-04199-0
  4. Outcomes in patients with diabetes, atherosclerotic cardiovascular disease, and heart failure treated with tirzepatide or dulaglutide: SURPASS-CVOT analysis. European Journal of Preventive Cardiology (2026). doi:10.1093/eurjpc/zwag452

Image credits

Free to use (CC0 / public domain); credited as a courtesy.

  • Connective Tissues Adipose (26915768267).jpg, by Berkshire Community College Bioscience Image Library (CC0, via wikimedia-commons): source
  • Adipose tissue.png, by Yuki T. Okamura (CC0, via wikimedia-commons): source
Open the hub