Semaglutide and Tirzepatide in Type 1 Diabetes: Real-World Insulin Deintensification, Cardiovascular Outcomes and Safety Assessment
This retrospective, propensity-score matched observational study used de-identified federated U.S. electronic health record (EHR) data to examine real-world associations of semaglutide (n=1,424) and tirzepatide (n=578) use in adults with type 1 diabetes (T1D) compared to 1:1 matched T1D controls (n=2,002) who did not receive these agents, over a study period from 2018–2025. Neither drug is approved for T1D. The study found that both agents were associated with statistically significant reductions in total daily insulin dose, HbA1c, and body weight at 12 and 24 months compared to controls. Greater insulin reductions were observed in semaglutide users who achieved higher weight loss or dose escalation. The pre-vs-post safety analysis identified predominantly gastrointestinal adverse events; DKA, severe hypoglycemia, pancreatitis, and retinopathy did not increase significantly overall, though patients with >30% insulin dose reduction had higher DKA rates. Semaglutide and tirzepatide exposure was associated with lower all-cause mortality and major adverse cardiovascular events versus matched controls. Key limitations include the observational design, EHR data quality constraints, off-label prescribing confounding, and preprint status, meaning findings have not yet undergone peer review.
Why this grade: While the study involves a relatively large human cohort with propensity-score matching, it is a retrospective observational analysis using EHR data with inherent confounding risks, covers an off-label indication, and has not yet been peer-reviewed as a preprint.
Semaglutide and tirzepatide are not approved for type 1 diabetes, and their real-world associations with insulin requirements, metabolic outcomes, and safety remain uncertain. Here we analyzed de-identified federated U.S. EHR data, comparing adults with T1D initiating semaglutide (n=1,424) or tirzepatide (n=578) between Jan 1, 2018, and Dec 31, 2025, with 1:1 propensity-score matched T1D controls without GLP-1RA/tirzepatide exposure (n=2,002). At 12 months mean insulin total daily dose (TDD) percent change was −22.0% (median −9.3%) in the semaglutide subcohort and −19.7% (median −17.2%) in the tirzepatide subcohort versus +10.3% (median +0.4%) in matched controls, widening to −27.4% (median −19.4%) and −25.9% (median −31.5%) versus +10.4% (median −1.1%) in controls at 24 months (all P<0.0001). The rate of patients achieving a ≥10% TDD reduction at 6 months was 51.0% (semaglutide) and 50.3% (tirzepatide) versus 18.4% in controls (both P<0.0001). At 12 months median semaglutide dose was 0.50 mg (Q1, Q3: 0.25, 1.00), with 67.8% of patients above the 0.25 mg starter dose. Median tirzepatide dose was 5.00 mg (Q1, Q3: 2.50, 7.50), with 68.2% of patients above the 2.5 mg starter dose. Stratified analysis of 12-month TDD percent change showed significant heterogeneity by weight-loss category in the semaglutide subcohort with mean TDD reduction of −34.5% in patients losing ≥10% of body weight versus −19.6% in patients with <5% weight loss (P<0.001); semaglutide dose escalation was also associated with mean TDD reduction (1.7–2.4 mg, −31.8%; 0.25 mg, −18.2%; P=0.007) while tirzepatide dose escalation was not significantly associated with TDD reduction (ANOVA P=0.800). At 12 months mean HbA1c percent change from baseline was −4.2% (semaglutide, P<0.0001) and −3.0% (tirzepatide, P=0.007) versus −0.3% in controls and mean body weight percent change was −4.0% and −6.8% versus −0.5% (both P<0.0001). The pre-versus-post 365-day safety profile across 13 prespecified events was gastrointestinal-predominant, with nausea or vomiting increasing by 8.5 percentage points (Benjamini-Hochberg adjusted P [BH P] <0.0001) and smaller increases in hypoglycemia (+3.1 pp, BH P=0.022), constipation (+3.0 pp, BH P=0.003), diarrhea (+2.6 pp, BH P=0.016), decreased appetite (+1.1 pp, BH P=0.016), and acute kidney injury (+1.3 pp, BH P=0.022); diabetic ketoacidosis, severe hypoglycemia, pancreatitis, gallbladder disease, gastroparesis, and retinopathy progression did not change significantly. Among clinical and pharmacologic subgroups, patients with a 6-month TDD reduction of >30% had significantly higher incidence of diabetic ketoacidosis (5.42 events per 100 person-years; BH P=0.034). Across 24 prespecified cardiovascular and renal time-to-event endpoints, exposure to semaglutide or tirzepatide was associated with lower 2-year all-cause mortality (1.4% versus 5.7%; BH P<0.001) and major adverse cardiovascular events (4.2% versus 7.5%; BH P=0.002), with mortality reductions for both drugs individually and a reduction in major adverse cardiovascular events for tirzepatide; no renal endpoint remained significant after multiple-testing correction. In summary, this study demonstrates real-world reductions in daily insulin requirements, HbA1c, and weight in T1D patients after initiation of semaglutide or tirzepatide, motivating prospective evaluation of these therapies as adjunctive therapy in selected T1D adults.
Educational summary of published research — not medical advice. Full text is shown only where licensing permits.