What Off-Label GLP-1 Use Reveals About Access and Risk

More than 1.1 million U.S. adults received a GLP-1 prescription between 2021 and 2025 without an FDA-approved indication recorded in their medical files, according to an NYU-led analysis published in Obesity.

That number sounds like a clean measure of off-label use. It is actually a pile of different clinical stories wearing one billing code.

Some patients may have had obesity or a related condition that was missing from the database. Some may have qualified before entering the health system or continued medication after losing weight. Others may have been taking GLP-1s for conditions still under investigation. A smaller but concerning group may have been using the drugs despite having a normal recorded body mass index or a history of disordered eating.

Meanwhile, another study found that 9.3% of U.S. children ages 8 to 11 with obesity but without diabetes were receiving GLP-1 drugs, up from 0.03% in 2019. These prescriptions were described as off-label because the children were younger than the approved population discussed in the reporting.

Put the adult and pediatric findings together and a more useful question emerges. Once prescribing travels beyond a drug’s established boundaries, what systems determine whether that experiment comes with careful monitoring, informed uncertainty and equitable access?

“No Recorded Indication” Leaves a Lot Unresolved

The adult medical-records analysis examined roughly 90 million adults without a current recorded diagnosis of type 2 diabetes, obesity, or overweight accompanied by a related complication. Just over 1.1 million received liraglutide, semaglutide or tirzepatide from 2021 through 2025.

The annual prescribing rate within this group rose from 0.11% in 2021 to 1.5% in 2025, a 13.6-fold increase. The researchers tried to reduce coding noise by requiring an earlier BMI measurement. They also examined people who had no previous recorded history of obesity and found the same upward trend, although the total fell to roughly 350,000.

Those checks strengthen the finding that prescribing expanded. They cannot establish why each prescription was written or whether it was clinically appropriate. Electronic records are excellent at counting what clinicians entered. They are less talented at reconstructing a patient’s entire medical life.

Several signals still deserve scrutiny. Roughly one-third of recipients categorized as lacking an approved indication had a normal recorded BMI. People in this group were six times as likely as non-users to have an eating-disorder history, although the researchers could not reliably separate different eating disorders.

That uncertainty cuts in more than one direction. Emerging research may eventually support GLP-1 treatment for some conditions, including binge-eating disorder. The same appetite-suppressing effects could also be misused to sustain restrictive eating. A diagnosis code cannot settle which situation applies, and body size alone cannot diagnose an eating disorder.

Clinicians and patients therefore need more context than the word “off-label” supplies: the treatment goal, relevant diagnoses, eating-disorder screening, anticipated benefits, known risks and a plan for follow-up. That is not a verdict against prescribing. It is the information required to judge a prescription rather than its label status.

Children Expose the Monitoring Gap

The pediatric pattern has a different clinical shape. In the study of children ages 8 to 11, prescriptions were more common among those with severe obesity or complications such as prediabetes, sleep apnea, high blood pressure or high cholesterol. This suggests that at least some clinicians were using the drugs in response to substantial current risk.

Childhood obesity is a complex condition influenced by genetic, environmental and social factors. Treating it as a failure of willpower loads stigma onto children while explaining very little. It also obscures the clinical dilemma: severe obesity can carry serious health consequences, while evidence about years of GLP-1 exposure during growth and puberty remains limited.

The recent history shows how rapidly practice has moved. Pediatric prescribing in the study climbed from 0.03% in 2019 to 9.3%. Reporting on a separate nutritional study says the American Academy of Pediatrics began recommending GLP-1 therapy for childhood obesity in 2023 while emphasizing nutritional support alongside medication. The exact populations and years in the adult and pediatric studies differ, so their growth rates should not be directly compared. Both document practice changing faster than long-term outcome data can accumulate.

A claims-based study of roughly 2,000 patients ages 10 to 17 sharpens the concern. Among children who started a GLP-1 without a preexisting deficiency in their records, about 17% received a diagnosis of a nutritional deficiency or related complication within a year. Vitamin D deficiency was most common, at 12%.

The study cannot prove that GLP-1 treatment caused those diagnoses. It was observational, relied on claims, and covered 2017 through 2022, before the drugs became as widely used in children. Intentional weight loss itself may affect nutritional status, while increased medical contact could uncover deficiencies that were already present but undocumented.

The care surrounding those prescriptions was thin in the records: about 5% received dietary counseling within 30 days, and one-quarter received it within six months. Counseling codes may miss conversations that occurred during ordinary visits, but the gap is large enough to make monitoring part of the risk discussion.

For families evaluating treatment with a qualified clinician, useful questions include how nutrition, growth and muscle health will be followed; what symptoms or test results would prompt a change; who will provide dietary support; and what the plan is if medication becomes unaffordable or is stopped. Medication can be one component of care without carrying the entire burden of a child’s food environment, family resources, mental health and access to safe physical activity.

Access is Selecting the Patients Too

The adult analysis found prescribing without a recorded indication more often among women, white patients, privately insured patients and people living in less socially vulnerable areas. In the pediatric study, children receiving GLP-1s also came from wealthier families more often, and study leader Babak Orandi noted that treatment can cost hundreds of dollars a month.

These studies do not prove that income or insurance caused each prescription. Together, they support a narrower inference: when evidence is unsettled and treatment is expensive, people with fewer access barriers appear better positioned to obtain it. Clinical need may determine who could benefit, while coverage, geography, specialist availability and family time determine who reaches the prescription pad.

That creates a peculiar two-sided access problem. Some comparatively advantaged adults may be receiving GLP-1s without a clearly documented indication, while children and adults with established obesity-related risks may still struggle to get treatment. Broad warnings about “overuse” can miss under-access. Broad enthusiasm can miss weak screening and follow-up.

What Evidence Catching Up Can Look Like

Tirzepatide’s use for obstructive sleep apnea offers a limited comparison. A 2026 review of GLP-1-related treatments reported that Phase III trials reduced breathing disturbances by roughly 50% to 63%, or about 20 to 24 events per hour. Under the trial definitions, approximately 42% to 50% of participants taking tirzepatide reached remission after one year, compared with 14% to 16% receiving placebo. Those results preceded FDA approval for moderate to severe obstructive sleep apnea in adults with obesity in December 2024, according to the review.

This is one path by which yesterday’s experimental use can become today’s labeled indication: a defined population, randomized trials, measurable outcomes and regulatory review. The comparison has limits. Sleep apnea in adults is not a proxy for obesity treatment in a growing child, and the review itself was not a new randomized trial. It also found that CPAP reduced breathing interruptions more immediately on average than tirzepatide, suggesting that approval does not make a drug the sole or best intervention for every patient.

Labels will continue to change as evidence develops. Until then, “off-label” can describe an evidence gap, an administrative gap, an emerging use or a warning sign. The quality of care depends on finding out which one applies before the prescription becomes just another line in a database.