GLP-1 Receptor Agonists in Type 1 Diabetes: What the New Cardiorenal Evidence Means

For fifteen years, every landmark cardiovascular trial of GLP-1 receptor agonists has excluded patients with type 1 diabetes. LEADER, SUSTAIN-6, REWIND, AMPLITUDE-O, FLOW — every one of them set its enrollment criteria around T2D. The cardiorenal benefit signal that has reshaped diabetes care in T2D and obesity has, until now, sat over a hole in the evidence base for the roughly two million Americans living with T1D.

That hole started to close in March. A team led by Johns Hopkins published a target trial emulation in Nature Medicine on March 19, drawing on electronic health records from 174,678 T1D patients, and reported a 15% lower five-year risk of major adverse cardiovascular events and a 19% lower risk of end-stage kidney disease in patients who initiated a GLP-1 RA. A separate two-year retrospective from Cleveland Clinic, published in Diabetes Technology & Therapeutics, found lower all-cause mortality with similar safety. And the 2026 ADA Standards of Care now formally recommend GLP-1 RA-based therapy for adults with T1D and obesity — the first ADA recommendation of its kind.

This article walks through what the new T1D evidence actually shows, why these data are not the same as a randomized trial, the persistent DKA question, and how the 2026 ADA Standards reflect what is and is not yet established.

Why every landmark GLP-1 trial excluded T1D

The exclusion was deliberate, and it was not unreasonable. T1D is autoimmune destruction of pancreatic beta cells. Patients depend entirely on exogenous insulin. Adding a drug that suppresses glucagon, slows gastric emptying, and reduces appetite — on top of insulin — creates a real risk of severe hypoglycemia and, more concerning, diabetic ketoacidosis. SGLT2 inhibitors, when tested in T1D, produced enough excess DKA cases that the FDA's 2019 advisory committee declined to expand their T1D indication.

The trial designers excluded T1D for sample-size reasons too. T1D patients account for roughly 5–10% of total diabetes prevalence in the U.S. Cardiovascular and renal events accumulate slowly. Powering a hard-outcomes trial in T1D would require either a very large enrollment or a very long follow-up, and neither has been commercially attractive.

What that left was a 15-year gap. Roughly 31% of adult T1D patients develop major adverse cardiovascular events by middle age, and 7% develop end-stage kidney disease. The mechanisms — endothelial dysfunction, chronic inflammation, accelerated atherosclerosis — overlap heavily with T2D. The question was never whether T1D patients might benefit from cardiorenal protection. It was whether anyone could measure it without putting them in a randomized trial.

The Hopkins target trial emulation

Yunwen Xu and colleagues at the Johns Hopkins Bloomberg School of Public Health used Optum Labs Data Warehouse — de-identified EHR data covering more than 300 million U.S. patients across 60+ health systems — to construct what is called a sequential target trial emulation. The technique applies randomized-trial design rules to observational data: define eligibility, define the intervention, define a clear time-zero, and run the analysis as if the comparison were prospective.

They emulated 135 trials, one starting every consecutive month from January 2013 through March 2024. The cohort included 174,678 adults with T1D, contributing about 6 million person-trials. Mean age was 43. Within the cohort, 14,488 person-trials initiated a GLP-1 RA. Median follow-up was 38 months. Propensity score weighting handled measured confounders.

The primary findings:

• Major adverse cardiovascular events (myocardial infarction, stroke, all-cause mortality): 5-year risk 4.3% with GLP-1 RA versus 5.0% without. Hazard ratio 0.85, 95% CI 0.77–0.95. Kaplan-Meier curves separated at 1.5 years and stayed separated.
• End-stage kidney disease: 5-year risk 1.6% versus 1.9%. HR 0.81, 95% CI 0.69–0.95.
• Hospitalization for heart failure: HR 0.82, 95% CI 0.71–0.94.
• Major adverse liver events: HR 0.72, 95% CI 0.60–0.85.
• Weight loss ≥5%, ≥10%, ≥15%: all significantly more likely with GLP-1 RA initiation.

The safety findings are what move this paper from interesting to important. There was no significant increase in hospitalization for diabetic ketoacidosis. There was no significant increase in severe hypoglycemia. The DKA signal that ended SGLT2 inhibitor expansion in T1D did not appear here.

The effect sizes are also worth holding next to the T2D record. The MACE reduction in T2D landmark trials runs 13–14%. ESKD reduction runs about 16%. The T1D point estimates are 15% and 19% — comparable, possibly slightly larger. That mechanistic equivalence is part of why the finding is plausible: the cardiorenal pathways GLP-1 RAs act on (endothelial repair, inflammation, glucose-independent metabolic improvements) are not specifically T2D pathways. They are general pathways shared across both diseases.

The full PubMed entry is at PMID 41857198 with DOI 10.1038/s41591-026-04274-0.

The Cleveland Clinic two-year analysis

Six days after the Hopkins paper appeared, Cleveland Clinic researchers led by Samita Garg published a retrospective analysis in Diabetes Technology & Therapeutics of T1D patients receiving GLP-1 RAs (with or without GIP) over a two-year window. Endpoints were all-cause mortality, hospitalization rate, ED visits, and gastrointestinal medication utilization, against propensity-matched T1D controls on insulin alone.

The directional findings matched Hopkins. GLP-1 RA recipients had significantly lower all-cause mortality, fewer hospitalizations, fewer ED visits, less endoscopy utilization, and lower use of laxatives and prokinetics. DKA rates and severe hypoglycemia rates were similar between groups.

Two retrospective analyses do not equal a randomized trial. They do represent independent investigative teams, different EHR sources, different statistical approaches, and different endpoints, all converging on the same general signal. That kind of cross-replication is the strongest version of observational evidence available short of an RCT.

Cleveland Clinic's summary of the analysis includes Garg's own framing: "using these medications adjunctively to insulin in patients with type 1 diabetes was associated with very favorable outcomes." She is careful to add that prospective trials remain needed.

What the 2026 ADA Standards of Care actually say

The ADA's Standards of Care in Diabetes — 2026, released in late 2025 and updated through 2026, contains the change clinicians have been waiting on. Recommendation 8.29 was added for the first time, including "GLP-1 RA–based therapy and/or metabolic surgery as treatment options for obesity in people with type 1 diabetes."

That language is narrower than it might first appear. The recommendation is for obesity, not for cardiorenal risk reduction. The ADA did not, in this revision, recommend GLP-1 RA-based therapy for the prevention of major cardiovascular events or kidney disease in T1D. The Standards reflect a hierarchy: an obesity indication that the existing trial data and clinical experience can support, with the cardiorenal claim still requiring confirmation.

The companion 2026 ADA-EASD Management of Type 1 Diabetes Consensus Report goes further in scope. It treats GLP-1 RAs as part of the broader cardiometabolic toolkit available to T1D patients and includes them in the cardiovascular risk-management discussion. The consensus document is a guidance and synthesis document; it does not carry the same prescriptive force as the Standards of Care.

The combined picture: T1D + obesity is now a formal ADA-recommended indication. T1D + cardiorenal risk reduction is a clinically defensible use that the data supports but the ADA has not yet codified.

The DKA question is still the question

Whether GLP-1 RAs are safe in T1D rises and falls on whether they raise the DKA risk that limits SGLT2 inhibitors. The published evidence so far says they do not. The Hopkins paper found no significant increase in DKA hospitalization. The Cleveland Clinic analysis found similar DKA rates between groups. Earlier randomized trials in T1D — ADJUNCT ONE and TWO with liraglutide — did report some DKA signal, but at doses and clinical contexts that have been refined since.

The mechanistic argument is intuitive. SGLT2 inhibitors increase ketone production through urinary glucose loss; GLP-1 RAs do not. SGLT2 inhibitors enable euglycemic DKA by clearing the glucose that would normally signal a metabolic problem; GLP-1 RAs do not. The pharmacology of the two drug classes diverges precisely at the place where the safety signal differs.

What this means for clinicians: insulin dose reduction in patients starting a GLP-1 RA needs to be done carefully. Starting a GLP-1 RA reduces appetite and improves insulin sensitivity, both of which lower the basal insulin requirement. Cutting insulin too aggressively risks hyperglycemia and ketosis. Cutting it too slowly risks hypoglycemia. The CGM data is essential. Patients without continuous glucose monitoring should not be dose-titrated against memory and finger sticks.

Which GLP-1 RA, and at what dose

The Hopkins target trial emulation pooled GLP-1 RAs as a class. It does not separate the effect of semaglutide from liraglutide, dulaglutide, exenatide, or tirzepatide (the GIP/GLP-1 dual agonist). The class effect is the finding. Drug-specific effects in T1D are still an open question.

Some patterns from adjacent literature are worth holding alongside this. A 2025 trial published in NEJM Evidence (Shah et al.) studied semaglutide in adults with T1D and obesity using automated insulin delivery (AID) systems, and found roughly one-third of treated patients met a combined endpoint of improved time in range, low hypoglycemia, and at least 5% weight loss. A small CGM-focused trial reported time in range improving from 69.4% to 74.2% with daily insulin use down 11.3 units. Tirzepatide observational data in T1D is more limited, but early reports describe weight loss of 10–20 kg and insulin dose reductions exceeding 30%. The phase-3 SURPASS T1D trial of tirzepatide is ongoing.

Doses in T1D have generally tracked the lower end of the T2D and obesity ranges, with slower titration. This is partly defensive — to give insulin titration time to keep up — and partly because the T1D population starts thinner on average than the obesity-trial population. There is no T1D-specific dosing label for any GLP-1 RA in 2026.

For context on the underlying compounds, see profiles for semaglutide, tirzepatide, liraglutide, dulaglutide, and exenatide, plus our metabolic peptides category and cardiovascular category.

What this changes clinically

For an endocrinologist treating a 38-year-old with T1D, BMI 32, and an HbA1c around 7.9%, the 2026 evidence base supports adding a GLP-1 RA to the regimen with reasonable confidence. The obesity indication is now ADA-recommended. The cardiovascular and kidney signal is observational but consistent. The DKA risk does not appear elevated relative to baseline T1D risk. Insulin titration is the operational challenge, and CGM and structured visit cadence are the tools that make it manageable.

What does not change: GLP-1 RAs are an adjunct to insulin in T1D, not a replacement. The condition is still autoimmune insulin deficiency. No GLP-1 RA, no GIP/GLP-1 dual agonist, no triple agonist in development changes that. The framing in patient conversations should make this distinction explicit.

Compounded GLP-1 RAs are a separate question. As covered in our analysis of FDA's April 1 GLP-1 compounding clarification, branded semaglutide and tirzepatide are no longer in formal FDA shortage status, which narrows the legal lane for 503A compounding. Patients seeking an off-label use case in T1D will, in most situations, be receiving a branded prescription rather than a compounded preparation.

What still needs to happen

Two pieces of evidence would close most of the remaining gap.

First, a randomized cardiorenal outcomes trial in T1D. The Hopkins paper makes the prior plausibility strong enough that a sponsor could justify the investment. Whether one of the major manufacturers commits to it in 2026 or 2027 is a business decision, not a scientific one.

Second, drug-by-drug rather than class-level effect estimates in T1D. SURPASS T1D for tirzepatide is the closest active study. A semaglutide T1D outcome trial would matter substantially given semaglutide's market dominance. Smaller drugs in the class — dulaglutide, liraglutide — may end up characterized through observational EHR data alone.

The ADA Standards revision cycle is annual. The 2027 edition will incorporate whatever new evidence appears across 2026. The shape of the recommendation language in the 2027 Standards will tell us whether the cardiorenal indication gets added or whether the field is still waiting for prospective confirmation.

Frequently Asked Questions

Does the Hopkins paper prove GLP-1 RAs prevent heart attacks in type 1 diabetes?

No. It is a target trial emulation — a careful observational study, not a randomized trial. The hazard ratios are consistent with a real protective effect, and the methodology is designed to mimic an RCT as closely as the data allow, but residual confounding remains possible. The signal is strong enough to warrant a randomized trial; it does not, by itself, replace one.

Are GLP-1 RAs FDA-approved for type 1 diabetes?

No. None of semaglutide, tirzepatide, liraglutide, dulaglutide, or exenatide has a T1D indication in 2026. Use in T1D is off-label. The ADA's 2026 obesity-in-T1D recommendation does not change FDA labeling.

What about diabetic ketoacidosis risk?

The two 2026 retrospective studies did not find an increased DKA risk with GLP-1 RA use in T1D. Earlier liraglutide trials had some signal at higher doses. Mechanistically, GLP-1 RAs do not act through pathways that promote ketone production the way SGLT2 inhibitors do. Insulin titration discipline is still essential.

Can I use a compounded GLP-1 RA for type 1 diabetes off-label?

Branded semaglutide and tirzepatide are out of FDA shortage status as of February 2025, which narrows 503A compounding eligibility. Patients receiving GLP-1 RAs adjunctively for T1D are most often receiving branded prescriptions. The compounded supply is a separate regulatory and clinical question.

How does this affect tirzepatide specifically?

Tirzepatide observational data in T1D shows larger weight loss and insulin reductions than GLP-1-only agents, consistent with its T2D and obesity record. Hard outcomes data in T1D will come from SURPASS T1D. See the tirzepatide profile for the broader trial record.

What is the 2026 ADA recommendation for T1D and obesity?

Recommendation 8.29 in the 2026 Standards of Care includes GLP-1 RA-based therapy and/or metabolic surgery as treatment options for obesity in people with T1D. It is the first ADA recommendation specifically supporting GLP-1 RA use in T1D. The recommendation is for obesity, not cardiorenal risk reduction.

Sources

• Xu Y, Daya Malek N, Chang AR, et al. "Glucagon-like peptide-1 receptor agonists for major cardiovascular and kidney outcomes in type 1 diabetes." Nature Medicine, 2026 March 19. DOI: 10.1038/s41591-026-04274-0. PubMed: PMID 41857198
• Johns Hopkins Bloomberg School of Public Health. "Improved Heart and Kidney Outcomes for Type 1 Diabetes Patients Taking GLP-1 Weight-Loss Drugs." March 24, 2026. publichealth.jhu.edu
• Garg S, et al. "Long-term outcomes of GLP-1 therapies in type 1 diabetes." Diabetes Technology & Therapeutics, 2026. Cleveland Clinic summary at consultqd.clevelandclinic.org
• American Diabetes Association. "Standards of Care in Diabetes — 2026: Summary of Revisions." Diabetes Care, December 2025. PMC12690167
• ADA-EASD. "Management of Type 1 Diabetes in Adults: 2026 Consensus Report (draft for consultation)." September 2025. professional.diabetes.org
• diaTribe. "Your Guide to the 2026 ADA Standards of Care." December 15, 2025. diatribe.org

Published May 1, 2026 · PeptideKnow Editorial