Millions get GLP-1 drugs through channels insurance records may miss

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Oct 26 2026

Behind the rising use of GLP-1 medications, an 18-month US study examines who takes these drugs, why they start treatment, and how they obtain them.

GLP-1/GIP Uptake, Indication, and Access Pathways Among US Adults in the Understanding America Study. Image Credit: Celso Pupo / Shutterstock

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

In a recent study posted to the medRxiv preprint* server, researchers characterized the prevalence, indication, and access pathways of glucagon-like peptide-1 (GLP-1)-based medications, including dual GLP-1/GIP receptor agonists, among adults in the United States (US).  

Randomized trials have shown that GLP-1/GIP treatment improves weight, glycemic control, and cardiovascular outcomes, but the generalizability of this evidence to real-world settings remains limited. Trials often underrepresent healthier, younger adults who may have attenuated benefits and older adults at risk of adverse events. Their supervised dosing may not reflect variation in dose escalation and adherence outside clinical trials.

About the study

The researchers included investigators from the University of Southern California's Center for Economic and Social Research, Viterbi School of Engineering, and Schaeffer Center for Health Policy and Economics in Los Angeles, and the RAND Corporation in Santa Monica, California.

The team examined the uptake, indications, access pathways, and clinical profile of GLP-1/GIP treatment during 18 months of surveillance in the US. They used data from the nationally representative GLP-1/GIP Influence on Metabolism and Mind Examined in the Realworld, Realtime (GLIMMER) study.

The study included adults who reported using these medications in any of the three surveillance waves between 2024 and 2025. The study’s main comparison was GLP-1/GIP use, defined as either any use (including current or former use) or never use. The study estimated the prevalence of any GLP-1/GIP use. Secondary measures included how medications were obtained and the primary reason for use (e.g., weight loss, diabetes).

Access pathway was non-traditional if respondents reported prescriptions from online providers or medical spas, fills from a foreign, compounding, or online pharmacy, or no prescription. All other access pathways were categorized as traditional. The researchers assessed participants' cardiometabolic health, medication type, treatment duration, dose changes, and discontinuation.

Physical activity and heart rate measures were derived from smartwatch data for a subsample of participants. The team reported prevalence estimates adjusted to reflect the US adult population and examined group differences by indication and access pathway. The characteristics of GLP-1/GIP users in this study were also compared with those of participants in published trials.

Findings

The study included 9,150 adults; of these, 945 had never used the medications but intended to do so, and 1,274 reported any GLP-1/GIP use. Smartwatch data were available for 1,550 participants, including 156 users. Weighted prevalence of any use (current or former) increased from 8.2% to 12% between March 2024 and October 2025, corresponding to a 46% increase. Before 2024, initiations were mainly for diabetes, while weight loss dominated among more recent initiators.

By October 2025, weight-loss indications accounted for nearly half of all use, and the share of US adults reporting weight-loss-related use rose from 4.1% to 5.6% since December 2024, while diabetes indications remained stable. Semaglutide had the largest share of use (50%), followed by tirzepatide (30%). New initiations of tirzepatide exceeded those of semaglutide.

Participants who reported any use were more often female, tended to be older, and had a greater burden of cardiometabolic disease than never users. Compared with diabetes-indicated users, those taking the medications for weight loss tended to be younger, more often female, more educated, and healthier.

Weight-loss- and diabetes-indicated users differed from the trial populations. Participants in selected diabetes trials were generally older, predominantly male, and had higher cardiovascular risk. Weight-loss trials included participants with higher body mass index and typically excluded individuals with cardiovascular comorbidities and diabetes. Among weight-loss-indicated users in this study, about one in nine had cardiovascular disease and nearly one in four had diabetes.

Among adults with any use, 64% were still taking the medications, while 36% had discontinued. Diabetes-indicated users had longer treatment duration, lower discontinuation rates, and slightly higher adherence to prescribed dosing than weight-loss-indicated users, but differences in observed treatment duration may partly reflect longer follow-up in the diabetes group.

Treatment access was traditional for almost all diabetes-indicated users, whereas large proportions of weight-loss-indicated users had non-traditional access. Drug coverage was absent for 43% of weight-loss-indicated users, compared with about 3% of diabetes-indicated users. Among people obtaining treatment through non-traditional pathways, approximately 11% met the researchers' criteria for likely cosmetic use, but their reasons for taking the medications were not independently verified.

In the subsample with smartwatch data, GLP-1/GIP users had higher resting heart rates, fewer daily steps, lower heart rate variability, and lower maximum activity heart rates than never users. These associations cannot establish whether treatment caused the differences, particularly given the small number of users with smartwatch data. The researchers also noted that treatment indications and access routes were self-reported.

Conclusions

The proportion of adults reporting current or previous GLP-1/GIP use increased by 46%, from 8.2% to 12%, in a nationally representative adult cohort over 18 months of GLIMMER surveillance in the US. This growth was predominantly driven by weight-loss indications.

Users differed substantially in clinical profile, indication, and access pathways. Approximately one-third of users obtained these medications through nontraditional channels, suggesting that insurance claims may miss a substantial proportion of treatment use.

This raises questions about medication safety, effectiveness, and coverage, but the study did not establish whether these access routes affected clinical outcomes.

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.