More than 1.1 million US adults were prescribed a GLP-1 drug with no approved reason recorded in their chart, and 35.1% of those patients were already at a normal body weight, according to a study of national medical records published September 22, 2026 in the journal Obesity.
Researchers at New York University searched Cosmos electronic health records covering 92,415,648 adults with no apparent FDA-approved indication for these drugs. Of that group, 1,133,953 (1.2%) were prescribed liraglutide, semaglutide or tirzepatide anyway, the ingredients sold as Saxenda, Ozempic, Wegovy, Mounjaro and Zepbound.
If you or someone in your family takes one of these, the useful question is short: which qualifying diagnosis is actually written in your chart? That one line decides whether your insurance keeps paying, and whether anyone is monitoring you for the muscle and bone loss these drugs cause along with the fat loss. Bring it to your prescriber. It is not a reason to stop a drug on your own.
Whether a daily pill can replace the weekly shot is a separate question, and one trial tested a pill against semaglutide.
Key takeaways
- Among US adults with no apparent FDA-approved indication, GLP-1 prescribing rose 15-fold, from 0.1% in 2021 to 1.5% in 2025.
- Of the 1,133,953 patients who got a prescription, 35.1% were at a normal BMI when it was written and 85.5% were women.
- Eating disorders were documented six times as often in the patients prescribed a GLP-1 drug without an approved reason, 1.8% versus 0.3%, about 15 extra cases per 1,000 people.
What the study found
The team reviewed Cosmos records from January 2021 through December 2025, counting adults prescribed these drugs “without a documented indication (type 2 diabetes or obesity with qualifying comorbidities).” Prescribing in that group climbed 15-fold, from 0.1% in 2021 to 1.5% in 2025.
The patients who got a prescription were heavier than the adults who got nothing, but not by much: a median BMI of 25.9 versus 24.5. More than a third of the patients who got a prescription, 35.1%, were at a normal BMI. They were overwhelmingly female at 85.5% and mostly White at 60.0%. High blood pressure and abnormal cholesterol were more common among them, coronary artery disease and heart failure less common. They were also more likely to have private insurance and to live in less socially vulnerable areas. Eating disorders were documented in 1.8% of them versus 0.3% of the comparison group.
Dr. Kumar’s take
The “off-label prescribing” headline misses the real signal. Off-label prescribing is ordinary and often good medicine. What deserves attention is the six-fold excess of documented eating disorders sitting next to the fact that one in three of the patients given these drugs was already at a normal weight. Those two numbers point the same direction. Some share of this prescribing is not treating obesity, it is feeding disordered eating.
That matters because of what the drugs do. The weight that comes off is not all fat. Muscle and bone come off with it. In someone with obesity and diabetes, that trade is worth making, because the disease is doing real damage. In someone at a normal weight who wants to be smaller, there is no disease on the other side of the ledger, and the muscle and bone loss is the whole effect.
The headline number needs a brake on it. “No apparent indication” reflects what a chart contains, not a verdict on the prescriber. Someone can carry a diabetes diagnosis recorded at another health system, an obesity history treated years earlier, or a qualifying condition a clinician never coded. So the 1.2% counts adults with no apparent FDA-approved indication recorded, which is what the authors set out to count. The authors themselves conclude: “The risk-benefit profile in these populations remains uncertain.”
What sloppy charting does not explain is who is getting these drugs. The patients getting these prescriptions were mostly women, mostly privately insured, mostly in less socially vulnerable neighborhoods, and a third of them were at a normal weight. Missing documentation does not sort itself that way.
What it means for you
Ask which diagnosis is documented in your chart. If the answer is vague, better to learn that from your prescriber than from your insurer.
Ask what is being monitored. Muscle mass and bone health belong on that list for anyone losing weight on these drugs, and they matter more, not less, in someone who started at a normal weight.
If you or the person taking it has any history of disordered eating, say so out loud. This study found that history six times as often in the patients who were prescribed these drugs, and it changes the calculation.
If the appeal is the results rather than the needle, an oral drug reached nearly 12% weight loss in a trial.
A normal BMI does not rule out a legitimate reason to take one of these drugs. That call belongs to you and a doctor who can see your whole history. The study supports having the conversation, not settling it from a headline.