GLP-1 Prescribing Without an Indication

Between 2021 and 2025, the share of US adults with no documented reason for a GLP-1 prescription who received one anyway rose from 0.1% to 1.5%, a 15-fold increase over four years^[1]. The figure comes from a retrospective cohort study published in the journal Obesity on 22 September 2026 by a research team at New York University. They examined the records of 92,415,648 adults who had neither a documented type 2 diabetes diagnosis nor obesity with a qualifying weight-related comorbidity, and found that 1,133,953 of them (1.2%) had been prescribed liraglutide, semaglutide or tirzepatide. One number inside that group is the reason this paper is being talked about: 35.1% of the people receiving these medicines had a normal BMI.

What the researchers counted

The design is retrospective, so nobody was assigned to anything and no treatment was tested. The authors queried Cosmos, a de-identified database built from Epic electronic health records, covering January 2021 through December 2025^[1]. An adult entered the analysis if a GLP-1 receptor agonist appeared in their medication record without a documented type 2 diabetes diagnosis and without obesity plus a qualifying comorbidity. Anyone with no recorded body weight in the six months before their first prescription was excluded, because without a weight there is no BMI and no way to judge whether an obesity indication was present. The word the authors chose for the headline is "apparent", and it carries the whole caveat. A missing indication in a chart can mean the indication was missing, or it can mean the documentation was.

1.2% sounds small until you multiply it out. Across 92.4 million adults it is 1,133,953 people^[1].

What an approved indication actually requires

The US thresholds for weight management are specific and public. Zepbound, the tirzepatide product the FDA approved on 8 November 2023, is licensed for chronic weight management in adults with obesity (a body mass index of 30 kg/m2 or greater) or overweight (a BMI of 27 kg/m2 or greater) with at least one weight-related condition such as high blood pressure or high cholesterol, used alongside a reduced-calorie diet and increased physical activity^[2]. The Wegovy label sets out the same structure for semaglutide: weight reduction in adults and paediatric patients aged 12 and older with obesity, or in adults with overweight and at least one weight-related comorbid condition, with separate indications covering cardiovascular event risk reduction and noncirrhotic MASH with moderate to advanced liver fibrosis^[3]. An adult with a BMI of 24 and no comorbidity fits none of those descriptions.

Who received these prescriptions

The group prescribed a GLP-1 without an apparent indication did not resemble the population it was drawn from. 85.5% were female and 60.0% were recorded as White^[1]. Median BMI among recipients was 25.9 kg/m2 against 24.5 in the comparison group, so slightly heavier on average, but still inside the overweight band and not the obese one. Hypertension and dyslipidemia were more common among recipients, while coronary artery disease and heart failure were less common, which is roughly the profile of a healthier and younger group. Recipients also had lower social vulnerability scores and were more likely to hold private insurance. That last detail points at who can afford a medicine when the prescription falls outside what an insurer will cover.

The trend line matters as much as the snapshot. In 2021, 0.1% of adults without an apparent indication had a GLP-1 prescription in their record. By 2025 it was 1.5%^[1]. Those four years are the same period in which tirzepatide reached the US market and semaglutide became a household name.

The normal-BMI and eating-disorder findings

Two results deserve to be read slowly. The first is that 35.1% of people prescribed a GLP-1 without an apparent indication had a normal BMI^[1], meaning they were not marginally short of the 27 threshold but comfortably below it. The second is that recorded eating disorders appeared in 1.8% of recipients against 0.3% of non-recipients, roughly a six-fold difference^[1]. That association is real in the data and unexplained by it. An eating disorder diagnosis may have been entered in the chart long before the first prescription, or only after it, and a record review of this design cannot separate the two orderings. The authors claim no direction of causation. Neither do we.

What we don't yet know

Off-label prescribing is legal in the United States and is sometimes the right call. A clinician can hold a defensible reason that never reaches a structured field in a chart, and Cosmos records what was coded, not what was discussed in the room. The database also draws on health systems running Epic, which makes it very large without making it a random sample of the country. Most importantly, the study reports no clinical outcomes at all: no weight change, no adverse events, no discontinuation rates for this population. That absence is where the authors end, concluding that the risk-benefit profile in people without an apparent indication remains uncertain^[1].

There is a related gap worth naming. Prescriptions written outside an approved indication are still prescriptions for approved products, which is a different situation from the compounded and counterfeit GLP-1 market the FDA has warned about repeatedly. The agency had logged 990 adverse-event reports tied to compounded semaglutide and more than 730 tied to compounded tirzepatide as of 31 May 2026, and notes those counts are almost certainly an undercount because state-licensed pharmacies are not required to report^[4].

What this means if you are considering a GLP-1

Nothing in this paper shows that a GLP-1 works, or fails, in people with a normal BMI. It shows that a growing number of them are being prescribed one, and that the trial evidence supporting these drugs was built in patients who looked different. The useful question to put to a prescribing clinician is which licensed indication applies to you, and what the evidence looks like for someone in your situation rather than in the registration trials. Country-level prescribing rules and further safety detail sit on our semaglutide regulation page and our tirzepatide regulation page. This article is for educational and journalistic purposes only and does not constitute medical advice. Speak to a qualified healthcare professional before starting or stopping any prescription medicine.