A growing body of real-world evidence is exposing sharp limits to the GLP-1 receptor agonist boom, with large studies showing that most patients prescribed drugs like semaglutide and tirzepatide for weight loss quit within a year, while regulators on both sides of the Atlantic flag serious but uncommon safety signals ranging from pancreatitis to eye damage. The findings arrive as prescriptions for non-diabetic patients have surged from roughly 21,000 new starts in 2019 to more than 174,000 in 2023, a roughly 700% increase over four years that has outpaced the clinical evidence base tracking long-term outcomes.
Most Non-Diabetic Users Quit Within a Year
The clearest sign that real-world GLP-1 use diverges from clinical-trial conditions comes from a study of 125,474 U.S. adults who started liraglutide, semaglutide, or tirzepatide between 2018 and 2023. Among patients without type 2 diabetes, 64.8% discontinued treatment within one year, compared with 46.5% of those who had the disease. The gap suggests that side effects, cost, or access barriers hit weight-management users harder than the population for which these drugs were originally designed. Gastrointestinal problems, including moderate or severe adverse events documented in the same cohort, were a consistent driver of early exits, and the authors emphasized that persistence on therapy was substantially lower in routine care than in the pivotal obesity trials.
A separate analysis of 15,811 commercially insured adults starting semaglutide specifically for weight management between June 2021 and December 2023 reinforced those numbers: roughly 46% had stopped the drug by month five. Most participants also deviated from the recommended dose-escalation schedule, and discontinuation rates held steady even after supply shortages eased in late 2023, indicating that logistics alone do not explain the dropout pattern. In both datasets, people with higher baseline body mass index tended to stay on treatment longer, but even in these subgroups, adherence eroded over time, challenging the assumption that once patients start a GLP-1 drug they will stay on it long enough to sustain weight loss.
Prescriptions Surged Faster Than Safety Data
The scale of uptake among people without diabetes has been extraordinary. A study using the TriNetX research network, reported by The BMJ, estimated that new GLP-1 receptor agonist prescribing among people without diabetes was around 21,000 in 2019 and exceeded 174,000 by 2023. That 700% jump over four years happened largely on the strength of randomized trials demonstrating double-digit percentage weight loss, but the pharmacovigilance infrastructure has struggled to keep pace with such rapid expansion into a comparatively healthy population. Many of these new users are younger and have fewer comorbidities than typical diabetes patients, raising questions about how tolerable long-term side effects must be for a drug that is, in many cases, addressing risk rather than acute disease.
Evidence from toxicology services suggests that real-world use has been messy. U.S. poison centers recorded a rising number of GLP-1 exposure cases from 2017 through 2022, driven primarily by therapeutic errors and accidental overdoses that caused symptoms such as vomiting and dizziness, according to a study in the Journal of Medical Toxicology. Most cases were not life-threatening, but the trend underscores how quickly complex injectable regimens have moved into primary care and even informal weight-loss channels. The authors noted that confusion over dose titration and pen devices contributed to misdosing, highlighting an educational gap that has not received the same attention as supply constraints or insurance coverage debates.
Side Effects, Expectations, and the Adherence Gap
About half of GLP-1 agonist users experience nausea, and roughly one-third report diarrhea, according to a RAND Corporation report, which characterized most of these side effects as mild and not serious. That framing, however, sits uneasily alongside the high discontinuation rates documented in large insurance and health-system cohorts. If side effects are generally tolerable, other forces must be pushing patients off treatment at rates that would be alarming for any chronic therapy. Out-of-pocket costs for branded injectables can run into hundreds of dollars per month in the United States, and coverage policies often require repeated prior authorizations or proof of ongoing weight loss, creating friction that may compound even modest physical discomfort.
Unmet expectations may also be playing a role. Clinical trials are tightly structured: participants receive lifestyle counseling, regular follow-up, and clear messaging that weight returns when treatment stops. In everyday practice, patients may start GLP-1 drugs expecting a short-term “reset” rather than an open-ended commitment. When weight loss plateaus or side effects persist, the perceived benefit can shrink quickly, especially for those using the drugs for cosmetic or social reasons rather than to manage diabetes or severe obesity. The combination of gastrointestinal symptoms, administrative hurdles, and ambiguous long-term goals helps explain why real-world adherence looks far more fragile than promotional narratives suggest.
Pancreatitis, Eye Damage, and Mental Health Flags
Beyond gut symptoms, regulators have begun cataloguing rarer but more severe risks. The UK’s Medicines and Healthcare products Regulatory Agency has warned of a small risk of severe acute pancreatitis with semaglutide and tirzepatide, citing Yellow Card reports that included deaths. In response, the agency and Genomics England have launched the Yellow Card Biobank to study adverse drug reactions at the genomic level, a signal that regulators view the current reporting system as insufficient for drugs prescribed at this scale. A recent review hosted on the National Institutes of Health platform similarly noted that while gastrointestinal adverse events are widely recognized as a class effect of GLP-1 receptor agonists, potential pancreatic and thyroid harms remain less well defined, with conflicting signals across trials, observational studies, and animal data.
Ocular safety has also emerged as a concern. A large Canadian observational study in JAMA Ophthalmology found that GLP-1 receptor agonist use was linked to a higher risk of neovascular age-related macular degeneration in diabetic patients, with the association growing stronger at six months and again at 30 months of use. The authors cautioned that residual confounding could not be ruled out, but the pattern has prompted calls for more frequent eye screening in high-risk users. At the same time, U.S. regulators have been evaluating reports of suicidal thoughts and self-harm in people taking GLP-1 drugs, with the Food and Drug Administration stating in a January 2024 communication that available data did not establish a causal link but warranted continued monitoring. Together, these signals illustrate how quickly the safety profile of a drug class can evolve once millions of relatively healthy people begin long-term use.
Rebalancing Hype With Long-Term Evidence
The emerging evidence does not negate the substantial benefits GLP-1 receptor agonists can offer, particularly for people with type 2 diabetes and severe obesity, where weight loss and improved glycemic control can dramatically reduce cardiovascular and renal complications. Instead, it highlights a mismatch between the blockbuster narrative and the realities of chronic treatment in everyday life. High discontinuation rates among non-diabetic users suggest that many will not experience the sustained weight loss and cardiometabolic improvements seen in trials, while growing but still uncertain safety signals complicate risk-benefit calculations for younger and lower-risk patients. For clinicians, this means emphasizing that these drugs are tools rather than cures, and that they work best as part of a broader strategy that includes diet, activity, and attention to mental health.
For policymakers and regulators, the challenge is to build surveillance systems that can keep up with rapid adoption. Initiatives like the Yellow Card Biobank and large-scale electronic health record analyses are steps toward a more nuanced understanding of who benefits most and who is most vulnerable to harm. As more long-term data accumulate, guidelines may need to differentiate more sharply between indications, reserving GLP-1 drugs for those with clear medical need while discouraging casual or purely aesthetic use. In the meantime, the combination of widespread discontinuation, rising reports of complex side effects, and incomplete knowledge about organ-level risks argues for a more cautious framing of these medicines: powerful but imperfect interventions whose true value will only become clear with time, careful monitoring, and honest communication about their limits.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
