Highlight
- Among veterans with type 2 diabetes on basal insulin, initiation of GLP-1 receptor agonists (GLP-1RAs) did not lead to higher rates of insulin discontinuation compared to sodium-glucose cotransporter-2 inhibitors (SGLT-2is) or dipeptidyl peptidase-4 inhibitors (DPP-4is) over three years.
- The study leveraged target trial emulation methods and matched cohorts from U.S. Veterans Health Administration electronic health records (EHR) covering 2020–2022.
- Risk ratios for insulin discontinuation were 0.93 (95% CI, 0.86 to 1.01) for GLP-1RAs versus SGLT-2is, and 0.98 (95% CI, 0.87 to 1.09) versus DPP-4is, indicating no statistically significant differences.
- Results were consistent across subgroups and in per-protocol analyses despite a high burden of hyperglycemia and comorbidities in the veteran population.
Study Background and Disease Burden
Type 2 diabetes (T2D) remains a prevalent chronic condition globally, requiring tailored pharmacotherapy to optimize glycemic control and mitigate complications. Insulin therapy is often employed when oral and non-insulin injectables fail to maintain target glycemic levels. However, insulin use adds complexity and risk, including hypoglycemia and weight gain. Recent guidelines emphasize early use of agents with cardiovascular and renal benefits such as GLP-1 receptor agonists and SGLT-2 inhibitors. GLP-1RAs reduce appetite and promote weight loss, and can potentially reduce insulin requirements. Whether GLP-1RA initiation facilitates complete insulin discontinuation compared to other oral agents, in real-world clinical practice, remains unclear and is clinically significant for patient-centered diabetes management.
Study Design
This study employed a target trial emulation design utilizing electronic health record (EHR) data from the U.S. Veterans Health Administration to approximate the conditions of a randomized controlled trial in a real-world cohort. Eligible participants were veterans with T2D receiving basal insulin who initiated either a GLP-1RA, an SGLT-2i, or a DPP-4i between 2020 and 2022. Medications analyzed included mainly semaglutide (76.6%) among GLP-1RAs, empagliflozin (99.7%) among SGLT-2is, and alogliptin (95.9%) among DPP-4is. The primary endpoint was insulin discontinuation, operationalized as having a prescription gap of 12 months or more over a follow-up period of up to 3 years. Propensity score matching yielded 8869 matched triplets balanced on key covariates such as age, sex, race, baseline HbA1c, and comorbidities.
Key Findings
Over three years, insulin discontinuation occurred in 16.7% of GLP-1RA initiators compared to 17.9% with SGLT-2i and 17.1% with DPP-4i. Intention-to-treat analysis showed a risk ratio of 0.93 (95% CI, 0.86–1.01) comparing GLP-1RA to SGLT-2i and 0.98 (95% CI, 0.87–1.09) when compared to DPP-4i, indicating no statistically significant superiority of GLP-1RAs in facilitating insulin discontinuation. Per-protocol and subgroup analyses did not reveal meaningful differences. Most participants were older (63% ≥65 years), predominantly male (93%), and had poorly controlled diabetes (48% with HbA1c ≥9%).
These findings challenge previously held assumptions that GLP-1RAs inherently improve insulin independence more than oral antihyperglycemics when added to basal insulin. This equivalency in real-world discontinuation rates might reflect patient selection, clinical inertia, or the complex nature of insulin dependence in advanced T2D.
Expert Commentary
The use of target trial emulation here provides robust comparative effectiveness evidence by reducing biases typical of observational studies. The large veteran cohort with detailed EHR data enhances generalizability to older, high-risk populations commonly seen in clinical practice. However, possible residual confounding and misclassification of drug exposure or discontinuation based on pharmacy fills must be acknowledged, potentially biasing results toward no difference.
Clinicians should interpret the absence of insulin discontinuation benefit with GLP-1RAs in the context of their other established benefits, including cardiovascular protection and weight reduction. Furthermore, treatment individualization remains paramount. Future randomized trials or pragmatic studies focused explicitly on insulin discontinuation might refine these insights.
Conclusion
In this large real-world cohort of veterans with type 2 diabetes managed on basal insulin, adding a GLP-1 receptor agonist did not increase the likelihood of insulin discontinuation compared with initiation of SGLT-2 inhibitors or DPP-4 inhibitors over three years. These findings support clinicians in selecting adjunctive therapies based on comprehensive patient needs rather than expectations of insulin cessation alone. Ongoing research should explore patient-level predictors of insulin independence.
Funding and Trial Registration
The study was supported by the U.S. Department of Veterans Affairs. No clinical trial registration was reported due to the observational design utilizing existing EHR data.
References
1. Lipska KJ, Zawack K, Yan L, Mutalik P, Li J, Gulanski B, Ioannou GN, Aslan M. Comparative Effectiveness of Glucagon-like Peptide-1 Receptor Agonists Versus Oral Agents for Insulin Discontinuation in Type 2 Diabetes: A Target Trial Emulation. Ann Intern Med. 2026 Jul 14; PMID: 42441965.
2. American Diabetes Association. Standards of Medical Care in Diabetes—2024. Diabetes Care. 2024 Jan;47(Suppl 1):S1–S194.
3. Nauck MA, Meier JJ. GLP-1 receptor agonists in the treatment of type 2 diabetes—state-of-the-art. Mol Metab. 2019 Oct;30:170-182.
4. Zinman B, Wanner C, Lachin JM, et al. Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes. N Engl J Med. 2015;373(22):2117-2128.
5. Oak JS, Lipska KJ. Insulin therapy and discontinuation in diabetes care: clinical realities and research directions. Endocrinol Metab Clin North Am. 2023;52(1):1-14.
