5E-01 - Safety and effectiveness of GLP-1 receptor agonists among insulin-treated persons with type 1 diabetes

Background: Despite growing use of GLP-1 receptor agonists (GLP-1RAs) among persons with type 1 diabetes (T1D), their safety and effectiveness remain unclear as pivotal trials excluded this population.

Objectives: To evaluate the risks of severe hypoglycemia and diabetic ketoacidosis (DKA) and the cardiovascular benefits associated with GLP-1RA initiation in persons with T1D.

Methods: We conducted a sequential trial emulation study among individuals with T1D on insulin therapy using de-identified electronic health records from the Optum Labs Data Warehouse. Patients were classified as GLP-1RA initiators or non-initiators for every consecutive month between January 2013 and March 2024, and were followed until the occurrence of outcomes, death, or last encounter (through June 30, 2024). Safety outcomes were defined as hospitalization with a primary diagnosis of hypoglycemia or DKA, and the cardiovascular outcome was a composite of myocardial infarction, stroke, heart failure, and mortality. We estimated an intention-to-treat hazard ratio (HR) comparing GLP-1RA initiation versus no initiation using Cox regression model. The model was adjusted for baseline confounders, including demographics, body mass index (BMI), hemoglobin A1c (HbA1c), systolic blood pressure, kidney function, comorbidities, and medication/technology use and healthcare utilization in the prior year. We also used traffic accident as a negative control outcome.

Results: Among 5,614,691 person-trials (161,841 individuals across 135 sequential trials), there were 14,141 GLP-1RA initiators. Compared to non-initiators, GLP-1RA initiators were older (mean age 50 vs. 43 years), had higher mean HbA1c (9.0% vs. 8.4%) and more obesity (65% vs. 32%). Hospitalization for DKA occurred in 596 (4.1%) GLP-1 RA initiators and 446,312 (7.3%) non-initiators, with an adjusted HR of 0.82 (95% CI: 0.76-0.89) for GLP-1RA initiation versus no initiation. Hospitalization for hypoglycemia occurred in 212 (1.5%) GLP-1RA initiators compared with 123,858 (2.0%) non-initiators (adjusted HR=0.84, 95% CI: 0.74-0.97). GLP-1RA was associated with a lower risk of composite cardiovascular outcome (410 out of 11,686 [3.5%] vs. 176,888 out of 5,240,491 [3.4%] events in GLP-1RA initiators vs. non-initiators; adjusted HR=0.82, 95% CI: 0.74-0.90). There was no association between GLP-1RA initiation and traffic accident.

Conclusions: GLP-1RA initiation in persons with T1D on insulin therapy was associated with a lower risk of cardiovascular outcomes, without an increased risk of hypoglycemia or DKA, compared with non-initiation. While these findings support the safety and effectiveness of GLP-1RAs in this population, randomized trials are needed for confirmation.