GLP-1 Agonists in Type 1 Diabetes: The Off-Label Insulin Adjunct Case

Why Type 1 Patients and Clinicians Are Looking Beyond Insulin

Type 1 diabetes (T1D) has long been managed through a single non-negotiable cornerstone: exogenous insulin. Unlike type 2 diabetes, where a growing pharmacological arsenal — metformin, SGLT2 inhibitors, GLP-1 receptor agonists, DPP-4 inhibitors — offers layered glycemic control, people with T1D have historically had little beyond insulin dose titration and lifestyle modification to optimize their outcomes.

Yet two stubborn clinical realities make this single-agent paradigm increasingly inadequate. First, even with best-in-class insulin regimens and continuous glucose monitoring (CGM), a substantial proportion of T1D patients fail to achieve guideline-recommended A1c targets below 7.0%. Second, the modern T1D population is increasingly burdened by overweight and obesity — a combination that drives insulin resistance, worsens glycemic variability, and compounds cardiovascular risk. In the landmark DCCT/EDIC cohort, participants gained an average of 4.5 kg during intensive insulin therapy; in real-world practice, weight gain of 10–15 kg over a lifetime of insulin therapy is common.

This is the clinical gap that has driven serious interest in GLP-1 receptor agonists (GLP-1 RAs) as adjunct therapy in T1D. These agents — which include liraglutide (Victoza, Saxenda), semaglutide (Ozempic, Wegovy), dulaglutide (Trulicity), and exenatide (Byetta, Bydureon) — lower blood glucose through multiple insulin-independent mechanisms, suppress appetite, slow gastric emptying, and produce meaningful weight loss. The question is not whether they work in T1D, but whether they can be used safely.

The Mechanistic Case: How GLP-1 RAs Can Help Without Beta Cells

GLP-1 receptor agonists were originally developed to exploit the incretin effect — the phenomenon by which oral glucose triggers a far greater insulin response than intravenous glucose. This effect is blunted in type 2 diabetes. But in T1D, where insulin-secreting beta cells have been destroyed by autoimmune attack, there are no beta cells to stimulate. So why would a GLP-1 RA help at all?

The answer lies in the multiple non-beta-cell mechanisms through which GLP-1 RAs influence glucose metabolism:

• Glucagon suppression: GLP-1 receptors are present on pancreatic alpha cells, and GLP-1 RA stimulation suppresses postprandial glucagon secretion. Unregulated glucagon in T1D is a major driver of postprandial hyperglycemia and is increasingly recognized as a therapeutic target in its own right.
• Gastric emptying delay: By slowing the rate at which nutrients enter the small intestine, GLP-1 RAs blunt postprandial glucose spikes — reducing peak glucose excursions and decreasing the required mealtime insulin bolus.
• Central appetite suppression: GLP-1 receptors in the hypothalamus and brainstem mediate satiety signaling, reducing caloric intake and producing weight loss independent of insulin changes.
• Insulin sensitivity improvement: Weight loss itself reduces adipose-driven insulin resistance, which in T1D means lower total daily insulin requirements, fewer dose miscalculations, and reduced glycemic variability.

The ADJUNCT Trial Program: What the Evidence Shows

The most robust clinical data on GLP-1 RAs in T1D comes from the ADJUNCT trial program, sponsored by Novo Nordisk and conducted between 2012 and 2016. These were the largest phase 3 randomized controlled trials ever conducted in T1D with a non-insulin agent, and their results shaped everything we know about the risk-benefit profile of this approach.

ADJUNCT ONE: 52-Week Efficacy and Safety

ADJUNCT ONE enrolled 1,398 adults with T1D (mean duration 20 years, mean A1c 8.2%, mean BMI 27.9 kg/m²) across 17 countries. Participants were randomized to liraglutide 1.8 mg/day, 1.2 mg/day, 0.6 mg/day, or placebo, all on top of basal-bolus insulin therapy. Duration was 52 weeks.

The headline results were promising but nuanced. At 52 weeks, liraglutide 1.8 mg produced a statistically significant A1c reduction of 0.47% versus placebo — modest compared to the 1.0–1.5% reductions seen in type 2 diabetes, but clinically meaningful in a population already near-target. Weight loss was more striking: the 1.8 mg dose produced a mean 4.5 kg reduction versus placebo. Bolus insulin doses fell by roughly 10%, reflecting the glucagon-suppressive and gastric-slowing effects.

However, the DKA signal was the trial's critical finding. Diabetic ketoacidosis occurred in 2.4% of patients on liraglutide 1.8 mg versus 0.6% on placebo — approximately a 4-fold increase. Investigation of these events revealed a consistent pattern: patients had reduced their insulin doses inappropriately (encouraged by falling glucose readings), some had adopted low-carbohydrate diets, and several had not adequately monitored ketone levels. This was not an intrinsic pharmacological property of liraglutide causing DKA — it was a behavioral risk amplified by inadequate patient education.

ADJUNCT TWO: 26-Week Replication

ADJUNCT TWO enrolled 835 T1D patients in a 26-week trial using the same liraglutide doses. Results were broadly consistent: A1c reduction of 0.37% with 1.8 mg versus placebo, weight loss of approximately 3.6 kg, and a reduction in total daily insulin dose. The DKA rate was again elevated in the liraglutide arms, though the absolute event rate was lower given the shorter duration.

Importantly, ADJUNCT TWO included a subgroup analysis showing that patients who maintained careful insulin dose management and used CGM had substantially lower DKA rates — suggesting that the elevated DKA risk is modifiable with the right infrastructure and patient education.

Why the FDA Did Not Approve Liraglutide for T1D

In January 2019, the FDA's Endocrinologic and Metabolic Drugs Advisory Committee met to review Novo Nordisk's supplemental NDA for liraglutide in T1D. The committee voted 8 to 7 against recommending approval — one of the closest margins in recent FDA advisory history. The committee acknowledged the efficacy signal was real but expressed concern that the DKA risk could not be adequately managed in real-world clinical practice without the intensive monitoring infrastructure used in the trials. They also noted that the modest A1c benefit might not justify the risk for all T1D patients.

The EMA (European Medicines Agency) similarly declined to approve any GLP-1 RA for T1D, citing comparable concerns. As of mid-2026, no regulatory agency has approved a GLP-1 RA for type 1 diabetes, making all use in this population off-label.

Understanding DKA Risk in Type 1 Diabetes with GLP-1 RAs

Diabetic ketoacidosis is the most serious safety signal in GLP-1 RA use in T1D, and it deserves careful mechanistic understanding rather than reflexive avoidance. DKA in this context is not caused directly by the GLP-1 RA itself — these drugs have no direct ketogenic effect. Rather, DKA risk in T1D patients on GLP-1 RAs arises from a cascade of downstream events:

The Insulin Under-Dosing Cascade

When a T1D patient starts a GLP-1 RA, several things happen simultaneously: gastric emptying slows, postprandial glucose peaks blunt, and the CGM or fingerstick readings trend lower. This is the desired therapeutic effect. However, if the patient or clinician responds to these improved readings by reducing insulin doses too aggressively — without accounting for ongoing basal insulin requirements — absolute insulin deficiency can develop. In T1D, even brief periods of absolute insulin deficiency can trigger ketogenesis. Unlike type 2 diabetes patients, T1D patients have no residual beta cell function to produce compensatory endogenous insulin.

Low-Carbohydrate Diets and DKA Amplification

Many T1D patients who pursue GLP-1 RA therapy simultaneously adopt low-carbohydrate or ketogenic diets, attracted by the promise of reduced glucose variability. This combination is particularly problematic. A very low-carbohydrate diet in T1D already drives the body toward fat metabolism and mild ketonemia. When combined with GLP-1 RA-driven insulin dose reduction, the threshold for progressing from euglycemic ketosis to frank DKA becomes dangerously low. Crucially, euglycemic DKA — where blood glucose is near-normal but ketone levels are life-threateningly elevated — is possible in this scenario, meaning patients may not recognize the danger until severely ill.

Risk Mitigation Strategies

Clinical experience and sub-analyses from the ADJUNCT trials suggest that DKA risk in T1D patients on GLP-1 RAs is substantially modifiable with appropriate precautions:

• Never reduce basal insulin when initiating a GLP-1 RA — only titrate bolus doses and only in response to documented hypoglycemia
• Mandate CGM use before and throughout GLP-1 RA therapy
• Provide blood or urine ketone monitoring equipment and clear protocols for when to check (illness, vomiting, extended fasting)
• Educate patients specifically about euglycemic DKA — normal blood sugars do not exclude DKA
• Avoid simultaneous very low-carbohydrate diets without expert supervision and more frequent ketone monitoring
• Start at the lowest available dose and titrate slowly over 8–12 weeks

Hypoglycemia Risk: More Complex Than It Appears

Counterintuitively, GLP-1 RAs do not significantly increase the rate of severe hypoglycemia in T1D — and in well-managed patients, they may actually reduce it. This is because the mechanisms of hypoglycemia risk with GLP-1 RAs differ fundamentally from those seen with insulin dose escalation.

Why GLP-1 RAs Are Relatively Hypoglycemia-Sparing

Unlike sulfonylureas, which stimulate insulin secretion independent of glucose levels, GLP-1 RAs work in a glucose-dependent manner. At low blood glucose levels, GLP-1 RA-stimulated glucagon suppression becomes less pronounced — preserving the counterregulatory glucagon response that protects against severe hypoglycemia. This glucose-dependent mechanism is one of the most clinically attractive features of the GLP-1 RA class.

In ADJUNCT ONE, rates of severe hypoglycemia (defined as an episode requiring third-party assistance) were not significantly different between liraglutide and placebo groups. Non-severe hypoglycemia rates were modestly lower in the liraglutide arm, likely due to reductions in bolus insulin requirements reducing the risk of post-meal hypoglycemia.

The Nocturnal Hypoglycemia Exception

One area of genuine concern is nocturnal hypoglycemia in patients who reduce basal insulin in response to initial GLP-1 RA-driven improvements in fasting glucose. If basal insulin is reduced too aggressively during the day and the GLP-1 RA's gastric-slowing effect wears off overnight, late-night glucose levels can drop unexpectedly. This underscores the importance of maintaining — not reducing — basal insulin when initiating GLP-1 RA therapy in T1D.

Weight Loss Benefit in Type 1 Diabetes: Why It Matters Clinically

The weight loss produced by GLP-1 RAs in T1D is clinically meaningful in ways that go beyond aesthetics. A substantial and growing proportion of people with T1D are now overweight or obese — a phenomenon sometimes called "double diabetes" when insulin resistance co-exists with autoimmune insulin deficiency. The consequences of excess weight in T1D include increased total daily insulin requirements, greater glycemic variability, accelerated cardiovascular risk (which is already elevated 2–4 fold in T1D versus the general population), and reduced quality of life.

Insulin Dose Reduction and Metabolic Benefits

The 3–5 kg of weight loss observed in ADJUNCT trials translated to meaningful reductions in total daily insulin dose — approximately 10–13% in the liraglutide 1.8 mg arm. Reducing exogenous insulin requirements has compounding benefits: each unit of insulin avoided reduces the risk of dose-calculation errors, reduces weight gain from insulin's anabolic effects, and decreases glycemic variability. For patients on closed-loop insulin delivery systems (artificial pancreas), lower total insulin requirements also improve the algorithm's ability to maintain time-in-range.

Cardiovascular Risk Reduction

In T1D, cardiovascular disease (CVD) remains the leading cause of premature death, with risk elevated 2–10 fold compared to age-matched individuals without diabetes depending on duration and control quality. GLP-1 RAs have demonstrated cardiovascular outcome benefits in large T1D-like high-risk T2D populations — most prominently in LEADER (liraglutide) and SUSTAIN-6 (semaglutide). While no dedicated cardiovascular outcomes trial has been completed in a pure T1D population, the mechanisms of benefit (weight loss, blood pressure reduction, anti-inflammatory effects, direct cardiomyocyte protection) are not fundamentally different in T1D.

Who Is a Candidate for Off-Label GLP-1 RA in Type 1 Diabetes?

Not every T1D patient is an appropriate candidate for off-label GLP-1 RA therapy. The clinical case for use is strongest in a specific patient profile, and candidacy should always be determined in partnership with an endocrinologist experienced in T1D management.

Strongest Candidates

The patients most likely to benefit — and least likely to experience harm — from GLP-1 RA adjunct therapy in T1D share several characteristics:

• BMI ≥27 kg/m²: Overweight or obese T1D patients have the most to gain from the weight loss and insulin-sensitizing effects. In lean T1D patients, the benefit-risk calculation shifts unfavorably.
• Already using CGM: Continuous glucose monitoring is not optional in this context — it is a prerequisite for safe therapy. Patients who are unwilling or unable to use CGM should not receive GLP-1 RA adjunct therapy for T1D.
• A1c above target despite optimized insulin therapy: Patients with persistent A1c above 7.5–8.0% despite best efforts at insulin optimization represent the core target population.
• No history of recurrent DKA: Patients with prior episodes of DKA — particularly if related to insulin omission or ketogenic dieting — represent a higher-risk group where additional caution is warranted.
• Access to endocrinology follow-up: Off-label GLP-1 RA use in T1D requires quarterly monitoring at minimum, with proactive management of insulin dose adjustments.

Contraindications and Cautions

• Personal or family history of medullary thyroid carcinoma or MEN2 (contraindication to all GLP-1 RAs)
• Severe gastroparesis — the gastric-slowing effect of GLP-1 RAs can worsen this condition
• Severe renal impairment (eGFR <15 mL/min) — dose adjustments may be required
• Active eating disorder history — the appetite suppression of GLP-1 RAs can interact unpredictably with restrictive eating patterns
• Pregnancy or planned pregnancy — GLP-1 RAs are not approved for use in pregnancy
• Very lean T1D patients (BMI <23 kg/m²) — weight loss in this group may be counterproductive

Choosing the Right GLP-1 RA for T1D Adjunct Use

While liraglutide has the most clinical trial data in T1D, it is not necessarily the best choice for all patients. The landscape of available GLP-1 RAs has expanded dramatically since the ADJUNCT trials, and clinicians increasingly consider other agents based on efficacy, tolerability, and convenience profiles.

Liraglutide (Victoza)

The only GLP-1 RA with phase 3 RCT data in T1D. Daily injection. The 1.8 mg dose has the strongest efficacy signal. Starting dose of 0.6 mg with weekly titration reduces GI side effects.

Semaglutide (Ozempic)

Weekly injection with superior weight loss efficacy compared to liraglutide in T2D trials. Growing body of off-label case series and small RCTs in T1D suggest A1c reductions of 0.5–0.8% and weight loss of 5–7 kg. The weekly dosing may improve adherence. The 0.5 mg starting dose is recommended, with titration to 1.0 mg at 4 weeks if tolerated.

Dulaglutide (Trulicity)

Weekly injection with a prefilled pen device that some patients find more convenient. Limited T1D-specific data, but pharmacodynamic profile suggests similar mechanisms. The starting dose of 0.75 mg weekly is generally well tolerated.

Oral Semaglutide (Rybelsus)

The only oral GLP-1 RA; taken once daily on an empty stomach. No T1D-specific trial data. Bioavailability is more variable than injectable formulations. Requires strict adherence to administration conditions (empty stomach, water only for 30 minutes before and after). May be appropriate for needle-averse patients with close monitoring.

The Regulatory Landscape: FDA Status and Future Outlook

The FDA's 2019 decision not to approve liraglutide for T1D was not a permanent door closing — it was a request for better data, better patient selection tools, and better risk management strategies. Since then, several developments have reshaped the landscape.

Post-ADJUNCT Research and Regulatory Signals

Following the ADJUNCT trials, several research groups pursued more targeted approaches to GLP-1 RA use in T1D. Investigators at University of California San Diego and several European centers published structured protocols that achieved ADJUNCT-like glycemic benefits with substantially lower DKA rates by mandating CGM, avoiding very low-carb diets, and providing patient education about euglycemic DKA. These real-world data have been cited by advocates for re-engagement with the T1D regulatory pathway.

The FDA's 2023 approval of the first closed-loop insulin delivery system (artificial pancreas) for T1D, and the rapid adoption of CGM as standard of care in T1D management, have also changed the infrastructure calculus. The monitoring capabilities that were considered impractical at scale in 2019 are now approaching routine use in T1D specialist centers.

Semaglutide NDA for T1D

As of 2025–2026, Novo Nordisk and several academic consortia were evaluating semaglutide-specific data in T1D, with potential NDA filing discussions ongoing. The superior weight loss efficacy of semaglutide relative to liraglutide, combined with the improved monitoring infrastructure now available, has reinvigorated the regulatory conversation. However, no approval timeline has been formally announced.

Practical Considerations: Titration, Side Effects, and Monitoring

For T1D patients who proceed with off-label GLP-1 RA therapy after careful evaluation, the practical initiation process requires more active management than is typical in type 2 diabetes.

Starting and Titrating the Dose

Regardless of which agent is chosen, the principle is always to start at the lowest available dose and titrate slowly. For liraglutide: begin at 0.6 mg daily for 4 weeks, then increase to 1.2 mg for 4 weeks, and consider further escalation to 1.8 mg only if the lower dose is tolerated and glycemic improvement has plateaued. For weekly semaglutide: begin at 0.5 mg weekly for 8 weeks before considering escalation to 1.0 mg.

The most common adverse effects during titration are gastrointestinal: nausea, vomiting, and decreased appetite. These are typically most pronounced in the first 2–4 weeks after each dose increase and usually improve with time. Eating smaller, lower-fat meals and avoiding lying down after eating can help manage GI side effects.

Monitoring Parameters

During GLP-1 RA titration in T1D, monitoring should be intensified compared to baseline:

• CGM review weekly for the first 8 weeks, with particular attention to overnight glucose trends and hypoglycemic episodes
• Blood ketone monitoring 2× weekly during weeks 1–4, then as clinically indicated
• Renal function (eGFR, creatinine) at baseline and at 3 months — GLP-1 RAs can affect GFR and some patients with T1D have subclinical nephropathy
• Body weight at every visit — track trajectory and adjust protein intake to preserve muscle mass
• A1c at 3 months and 6 months — response should be assessed at these intervals to determine whether to continue