A significant proportion of patients prescribed glucagon-like peptide-1 (GLP-1) receptor agonists for type 2 diabetes follow a "start-and-stop" pattern of usage rather than consistent, long-term adherence, according to a comprehensive study presented at ENDO 2026. The research, which analyzed the treatment trajectories of more than 60,000 Americans, reveals that while discontinuation rates are higher than previously estimated, a substantial number of patients eventually return to the therapy, suggesting that the path to metabolic health via these medications is rarely linear.
The study, led by Sainikhil Sontha, MS, a research associate at the Boston University School of Public Health, sought to address critical gaps in clinical understanding regarding how long patients remain on these highly sought-after drugs. As GLP-1 medications like semaglutide and tirzepatide have surged in popularity for both diabetes management and weight loss, the medical community has grappled with questions about long-term persistence and the factors that drive patients to abandon or resume treatment.
Methodology and Study Design
To investigate these patterns, researchers conducted a retrospective cohort study using extensive insurance claims data from Komodo Health, spanning from January 2019 to June 2025. This six-year window allowed the team to track patients through the height of the GLP-1 "boom" and through various cycles of drug availability and insurance policy shifts.
The study population consisted of 60,000 adults between the ages of 18 and 64. All participants had a recorded body mass index (BMI) of 25 kg/m² or higher and a diagnosis of type 2 diabetes. The cohort was specifically limited to those who had initiated treatment with one of three primary GLP-1 or dual-agonist medications: liraglutide, semaglutide, or tirzepatide. To ensure data quality, the researchers required that participants had been enrolled in their insurance plans for at least one year prior to starting the medication and had at least six months of follow-up data available.
For the purposes of the study, "discontinuation" was strictly defined as a gap of more than 60 days in filling a GLP-1 prescription. Conversely, "reinitiation" was defined as the act of obtaining a new fill at any point following a period of discontinuation. This 60-day threshold is a standard metric in pharmacological research to differentiate between minor logistical delays and a formal cessation of therapy.
The Start-and-Stop Phenomenon: Key Findings
The results of the analysis paint a complex picture of patient behavior. Sontha reported that approximately 40% of patients discontinued their GLP-1 medication within the first year of treatment. This number grew significantly over time, with nearly 60% of patients having stopped their medication by the end of the second year.
However, the data also offered a more optimistic perspective on patient persistence. Among those who discontinued the medication, 41.5% restarted the therapy within one year. Within two years of stopping, nearly 58% of those patients had reinitiated treatment.
"This suggests that for many patients, these medications aren’t being abandoned permanently; use is more start-and-stop than most people assumed," Sontha noted during the presentation. This "revolving door" effect indicates that while patients may struggle with the medication initially—whether due to side effects, cost, or supply issues—the perceived benefits often drive them to attempt the therapy again.
Identifying Predictors of Discontinuation
Using Cox proportional hazards models, the research team analyzed a variety of sociodemographic, clinical, and provider-level factors to determine what influenced a patient’s likelihood of stopping treatment. The findings highlighted significant disparities and clinical hurdles.
Socioeconomic and Demographic Factors
The study found that patients enrolled in Medicaid or Medicare were significantly more likely to discontinue their GLP-1 medications within the first year compared to those with private insurance. This likely reflects the ongoing challenges of coverage and the high out-of-pocket costs associated with these drugs, even for those with some form of government-sponsored insurance. Additionally, Black patients were found to have higher rates of discontinuation, underscoring potential inequities in healthcare support and access.
The Role of the Prescribing Physician
One of the more actionable findings of the study was the impact of the prescribing provider. Patients whose first GLP-1 prescription was written by an endocrinologist were 10% less likely to stop the medication compared to those whose prescriptions were written by general practitioners or other specialists. This suggests that the specialized counseling, titration management, and expectation-setting provided by endocrinologists may play a crucial role in patient retention.
Clinical Side Effects
Not surprisingly, physical tolerance was a major hurdle. Approximately 37% of patients who discontinued their medication cited nausea or other gastrointestinal (GI) side effects as a primary factor. GLP-1 medications work by mimicking hormones that slow gastric emptying and signal fullness to the brain, which inherently carries a risk of nausea, vomiting, and diarrhea, particularly during the dose-escalation phase.
Technological Evolution: Newer Drugs Show Better Retention
A notable highlight of the study was the comparison between older and newer generations of GLP-1 medications. The data revealed a clear trend: the more advanced the drug, the higher the likelihood of patient persistence.
Patients taking tirzepatide—a dual GIP (glucose-dependent insulinotropic polypeptide) and GLP-1 receptor agonist—were 41% less likely to discontinue treatment compared to those taking liraglutide, an older, once-daily injectable. Semaglutide users also showed better retention, being 28% less likely to stop their medication than those on older drugs.
This disparity may be attributed to several factors. Newer medications like semaglutide and tirzepatide are typically administered once weekly, whereas older options like liraglutide require daily injections. Furthermore, newer formulations have shown superior efficacy in both glycemic control and weight loss, which may provide the psychological motivation necessary for patients to push through initial side effects.
The Clinical Stakes of Non-Adherence
The primary concern for healthcare providers regarding these discontinuation rates is the loss of the "protective effects" that GLP-1 medications provide. These drugs have been clinically proven to do more than just lower blood sugar; they offer significant cardiovascular and renal benefits.
"Consistent use of these medications is what produces their protective effects," Sontha emphasized. "Stopping early may mean missed opportunities to prevent heart attacks, kidney disease progression, and other complications."
Type 2 diabetes is a progressive disease that significantly increases the risk of macrovascular complications (such as myocardial infarction and stroke) and microvascular complications (such as retinopathy and nephropathy). GLP-1 agonists have become a cornerstone of modern diabetes care because they address these risks directly. When a patient enters a cycle of stopping and starting, they may experience "rebound" hyperglycemia and weight regain, which can exacerbate the very conditions the medication is intended to treat.
Chronology and Context: The GLP-1 Landscape (2019–2025)
To understand the study’s findings, one must look at the timeline of GLP-1 development and market availability during the study period:
- 2019–2021: This period saw the steady rise of injectable semaglutide (Ozempic) for type 2 diabetes. During this time, liraglutide was still widely used, but the shift toward once-weekly injections was beginning to take hold.
- 2021–2023: The "GLP-1 explosion" occurred as semaglutide gained FDA approval for obesity (under the brand name Wegovy). This led to unprecedented demand, subsequent global shortages, and a surge in off-label use. The 60-day gaps observed in the study data during this period may partially reflect these supply chain disruptions rather than elective discontinuation.
- 2024–2025: Tirzepatide (Mounjaro/Zepbound) entered the market with even higher efficacy profiles. The study’s finding that tirzepatide has the highest retention rate aligns with its clinical reputation for being the most potent agent currently available in the class.
Analysis of Implications for Policy and Practice
The high rates of reinitiation identified in the study suggest that patients recognize the value of these treatments but face barriers that make continuous use difficult. For insurers and policymakers, this data points to a need for more flexible coverage models. If patients are frequently stopping and starting, the current rigid "prior authorization" processes may actually hinder health outcomes by making it harder for patients to resume therapy after a brief hiatus.
For healthcare providers, the "10% endocrinologist advantage" suggests that primary care physicians may need more resources or training to mirror the support structures found in specialized clinics. Effective titration schedules and proactive management of GI side effects are essential to helping patients navigate the first 90 days of treatment, which is often the most volatile period.
Furthermore, the higher discontinuation rates among Black patients and those on public insurance highlight a persistent health equity gap. If the most effective life-saving medications are only accessible or sustainable for those with private insurance and specialized care, the long-term result will be a widening of the diabetes-related mortality gap between different socioeconomic groups.
Conclusion
The study presented at ENDO 2026 serves as a vital reality check for the "miracle drug" narrative surrounding GLP-1 medications. While these drugs offer transformative potential for the treatment of type 2 diabetes and its complications, their effectiveness is heavily dependent on long-term adherence—a goal that nearly 60% of patients struggle to meet within two years.
However, the discovery that more than half of these patients eventually return to the medication offers a unique opportunity for intervention. By identifying the specific populations at risk of stopping—namely those on public insurance, those experiencing significant side effects, and those without specialist support—the healthcare system can begin to build better "safety nets" to ensure that the start-and-stop cycle eventually stabilizes into long-term health success.
As Sainikhil Sontha and his colleagues concluded, the goal for the next phase of GLP-1 integration into public health must be focused on sustainability. Providers and insurers must work in tandem to ensure that once a patient starts this journey, they have the clinical and financial support required to stay the course.
