Individuals prescribed glucagon-like peptide-1 (GLP-1) medications exhibit a greater propensity for initiating and subsequently discontinuing treatment than commonly perceived, according to a study slated for presentation at ENDO 2026, the Endocrine Society’s annual symposium in Chicago, Illinois, scheduled for Sunday.
The central inquiries addressed by our investigation, which have lacked comprehensive answers until now, were: What proportion of individuals with type 2 diabetes utilizing GLP-1 therapies ultimately cease their regimen? And subsequently, how many resume these treatments?
Sainikhil Sontha, M.S., research associate, Boston University School of Public Health, Boston, Mass.
Investigators conducted a retrospective cohort analysis employing Komodo Health U.S. claims data spanning from January 2019 to June 2025. The participant cohort comprised adults aged 18 to 64 years who presented with a body mass index (BMI) of 25 kg/m² or higher and a diagnosis of type 2 diabetes. These individuals had initiated treatment with liraglutide, semaglutide, or tirzepatide and had been enrolled for at least one year prior, with a minimum of six months of follow-up data. Discontinuation was quantitatively defined by a lapse exceeding 60 days in GLP-1 prescription refills. The act of obtaining a new prescription following such a lapse was categorized as reinitiation of therapy.
“Analysis of insurance records pertaining to over 60,000 Americans diagnosed with type 2 diabetes revealed that approximately 40% of patients ceased their GLP-1 medication within the initial year of treatment, and nearly 60% had discontinued by the conclusion of the second year,” Sontha reported.
However, the research also yielded an encouraging observation.
“More than half of those who discontinued their medication subsequently recommenced therapy within a year (specifically, 41.5%), and almost two-thirds did so within the span of two years (58.0%),” Sontha elaborated. “This indicates that for a substantial number of patients, these therapeutic agents are not being permanently abandoned; rather, their utilization pattern appears to be more intermittent—characterized by cycles of initiation and cessation—than widely assumed.”
By employing Cox proportional hazards models, the research team also incorporated sociodemographic, clinical, and provider-level predictive factors into their analysis.
Sontha and his associates determined that individuals covered by Medicaid or Medicare, Black patients, and those experiencing gastrointestinal adverse effects such as nausea (37%) were more predisposed to discontinue GLP-1 medication use within a one-year timeframe.
Conversely, patients whose initial GLP-1 prescription was issued by an endocrinologist demonstrated a 10% reduced likelihood of discontinuing therapy.
Furthermore, individuals prescribed newer pharmacological agents like tirzepatide exhibited a 41% lower probability of discontinuing treatment compared to those administered older medications such as liraglutide. Users of semaglutide showed a 28% decreased likelihood of discontinuing anti-obesity medication use relative to individuals on earlier-generation drugs.
“The significance of this research lies in the fact that the sustained application of these medications is pivotal for realizing their protective benefits,” Sontha emphasized. “Premature cessation of treatment may result in missed opportunities to avert potentially serious health outcomes, including myocardial infarctions, the progression of kidney disease, and other associated complications.”
The researchers aspire for these findings to furnish healthcare providers, insurers, and policymakers with insights to identify patients who may require enhanced support mechanisms to maintain consistent adherence to GLP-1 medications, he concluded.
