Comparative Effectiveness of Glucagon-like Peptide-1 Receptor Agonists Versus Oral Agents for Insulin Discontinuation in Type 2 Diabetes : A Target Trial Emulation
AI Summary
Addition of GLP-1RA to basal insulin did not increase insulin discontinuation versus SGLT-2i or DPP-4i (RR 0.93 and 0.98; non-significant).
Target trial emulation in U.S. Veterans Health Administration EHR with 8869 matched sets; cohort predominantly older White males with high baseline HbA1c.
Findings robust in modified per protocol and subgroup analyses; limitations include residual confounding and exposure or outcome misclassification in EHR data.
Ann Intern Med. 2026 Jul 14. doi: 10.7326/ANNALS-25-05216. Online ahead of print.
ABSTRACT
BACKGROUND: Addition of glucagon-like peptide-1 receptor agonists (GLP-1RAs) to basal insulin can decrease insulin requirements, but whether it permits insulin discontinuation is unclear.
OBJECTIVE: To compare rates of insulin discontinuation among patients with type 2 diabetes (T2D) receiving basal insulin who initiated treatment with a GLP-1RA, sodium-glucose cotransporter-2 inhibitor (SGLT-2i), or dipeptidyl peptidase-4 inhibitor (DPP-4i) between 2020 and 2022.
DESIGN: Target trial emulation.
SETTING: U.S. Veterans Health Administration electronic health record (EHR) data.
PARTICIPANTS: Veterans with T2D receiving basal insulin.
MEASUREMENTS: Insulin discontinuation, defined as the first gap in insulin prescription fills of 12 months or more over 3 years of follow-up.
RESULTS: Among 8869 matched sets of GLP-1RA (76.6% semaglutide, 15.2% dulaglutide, 7.9% liraglutide, and 0.3% exenatide), SGLT-2i (99.7% empagliflozin), and DPP-4i (95.9% alogliptin) initiators, 63% were 65 years or older, 93% were male, 70% were White, and 48% had a hemoglobin A1c (HbA1c) level of 9% or more. Over 3 years of follow-up, 1480 (16.7%) GLP-1RA initiators compared with 1585 (17.9%) SGLT-2i initiators and 1517 (17.1%) DPP-4i initiators discontinued insulin therapy in the intention-to-treat analysis (risk ratio, 0.93 [95% CI, 0.86 to 1.01] and 0.98 [CI, 0.87 to 1.09] for the GLP-1RA arm compared with the SGLT-2i and DPP-4i arms, respectively). Results were not substantively different in a modified per protocol analysis. None of the subgroups showed a comparative advantage of GLP-1RAs with respect to insulin discontinuation over SGLT-2is or DPP-4is.
LIMITATION: Possible residual confounding; misclassification of exposure and outcome using EHRs may bias associations toward the null.
CONCLUSION: Among veterans with T2D receiving basal insulin therapy, addition of GLP-1RA did not increase the chances of stopping insulin therapy compared with SGLT-2i or DPP-4i therapy.
PRIMARY FUNDING SOURCE: U.S. Department of Veterans Affairs.
Front Public Health. 2026 Jun 29;14:1887736. doi: 10.3389/fpubh.2026.1887736. eCollection 2026.ABSTRACTBACKGROUND: Bioterrorism involves the intentional release of biological agents capable of causing mass casualties, public health...
Acta Paediatr. 2026 Jul 14. doi: 10.1111/apa.70678. Online ahead of print.ABSTRACTAIM: Physical activity in childhood and adolescence is a cornerstone of holistic health, supporting physical...
JMIR Res Protoc. 2026 Jul 14;15:e94797. doi: 10.2196/94797.ABSTRACTBACKGROUND: Attention-deficit/hyperactivity disorder (ADHD) is a prevalent neurodevelopmental condition affecting approximately 7% to 8% of children and adolescents,...
JAMA Netw Open. 2026 Jul 1;9(7):e2623051. doi: 10.1001/jamanetworkopen.2026.23051.ABSTRACTIMPORTANCE: Child maltreatment is a major global public health concern with well-documented psychological consequences, but its associations with...
