GLP-1 Agonists vs Insulin: What Head-to-Head Trials Show

GLP-1 and Blood Sugar

3.71 kg

In a meta-analysis of 19 head-to-head trials, patients on GLP-1 receptor agonists lost 3.71 kg more than those on insulin therapy.

Abd El Aziz et al., Diabetes, Obesity & Metabolism, 2017

Abd El Aziz et al., Diabetes, Obesity & Metabolism, 2017

View as image

For decades, the standard progression for type 2 diabetes was straightforward: start with metformin, add oral medications, then move to insulin injections when those fail. That sequence is now outdated. A meta-analysis of 19 head-to-head clinical trials found that GLP-1 receptor agonists matched or exceeded insulin for blood sugar control while producing 3.71 kg more weight loss and 34% fewer hypoglycemic episodes.^[1] These findings have reshaped international treatment guidelines, positioning GLP-1 agonists as the preferred first injectable therapy over insulin for most people with type 2 diabetes.

How GLP-1 Agonists and Insulin Work Differently

Insulin is a direct hormone replacement. Injected insulin enters the bloodstream and forces glucose into cells regardless of blood sugar levels. This works reliably but carries inherent risks: if the dose is too high relative to food intake, blood sugar drops dangerously low.

GLP-1 receptor agonists take a fundamentally different approach. They mimic the incretin hormone GLP-1, which the gut releases after eating. This triggers insulin secretion from pancreatic beta cells, but only when blood glucose is elevated.^[2] When blood sugar is normal or low, the insulin-stimulating effect diminishes. This glucose-dependent mechanism explains why GLP-1 agonists carry far lower hypoglycemia risk than exogenous insulin.

Beyond insulin secretion, GLP-1 agonists suppress glucagon (the hormone that raises blood sugar), slow gastric emptying to blunt post-meal glucose spikes, and reduce appetite through central nervous system signaling.^[2] Insulin does none of these things. This multi-mechanism profile is why GLP-1 agonists produce weight loss while insulin typically causes weight gain.

The dosing schedules also differ. Modern long-acting GLP-1 agonists like semaglutide and dulaglutide require a single weekly injection. Basal insulin requires daily injections, and patients who add mealtime insulin may inject three to four times daily. For patients earlier in the treatment cascade, the comparison against first-line oral medications like metformin involves different considerations entirely, but the insulin-vs-GLP-1 decision point is where injection burden and monitoring requirements diverge most.

Head-to-Head Trials: Blood Sugar Control

The most comprehensive comparison comes from Abd El Aziz and colleagues, who pooled data from 23 publications describing 19 randomized trials that directly compared GLP-1 receptor agonists with insulin in patients taking oral glucose-lowering medications.^[1]

GLP-1 agonists reduced HbA1c by 0.12 percentage points more than insulin (P < .0001). That difference is modest in absolute terms, but it came alongside dramatically better weight and safety profiles. Long-acting GLP-1 agonists (liraglutide, semaglutide, dulaglutide) outperformed short-acting ones (exenatide twice daily) for both HbA1c and fasting glucose reduction.

One area where insulin maintained an advantage: fasting plasma glucose. Basal insulin reduced fasting glucose by 1.8 mmol/L more than GLP-1 agonists (P < .0001).^[1] This makes physiological sense. Basal insulin provides continuous glucose-lowering overnight and between meals, while GLP-1 agonists primarily act in the fed state when incretin signaling is active.

The SURPASS-2 trial demonstrated the trajectory of this drug class. Tirzepatide, a dual GIP/GLP-1 receptor agonist, reduced HbA1c by 2.01% to 2.30% across three dose levels, compared to 1.86% for semaglutide 1 mg.^[3] All tirzepatide doses were statistically superior to semaglutide, the previous best-in-class GLP-1 agonist. These reductions exceed what most insulin regimens achieve in clinical practice, though direct head-to-head data between tirzepatide and insulin are still accumulating.

Weight: The Decisive Difference

Weight is where the GLP-1 vs insulin comparison becomes most stark. Across 19 head-to-head trials, GLP-1 agonists produced 3.71 kg more weight loss than insulin (P < .0001).^[1] Since insulin typically causes weight gain of 2 to 4 kg, the net difference between the two approaches can exceed 6 kg.

This matters beyond aesthetics. Excess weight drives insulin resistance, the core metabolic defect in type 2 diabetes. Weight gain from insulin therapy can worsen insulin resistance, requiring progressively higher doses in a cycle that is difficult to break. GLP-1 agonists interrupt this cycle by reducing appetite and caloric intake through hypothalamic signaling and slowed gastric emptying.^[2]

In SURPASS-2, tirzepatide 15 mg produced 5.5 kg more weight loss than semaglutide 1 mg over 40 weeks.^[3] For patients with type 2 diabetes and obesity, which represents the majority of this population, this combined glucose-lowering and weight-reducing effect addresses both conditions simultaneously.

The GLP-1 agonists also improved lipid profiles. Triglycerides and LDL cholesterol were both significantly lower with GLP-1 agonists compared to insulin, along with lower systolic blood pressure.^[1] These cardiometabolic benefits compound over time and help explain the cardiovascular advantages described below.

Hypoglycemia: A 34% Advantage

Low blood sugar episodes are the most feared acute complication of diabetes treatment. In the Abd El Aziz meta-analysis, the proportion of patients experiencing hypoglycemic episodes was 34% lower with GLP-1 agonists than with insulin (P < .0001), with a similar trend for severe hypoglycemia.^[1]

The mechanism behind this difference is the glucose-dependent nature of GLP-1 receptor activation. When blood glucose falls below approximately 65 mg/dL, GLP-1 receptor-mediated insulin secretion essentially stops.^[2] Injected insulin has no such off switch. Once administered, it works regardless of glucose levels.

This safety advantage has practical consequences. Patients on insulin must carefully match doses to meals, monitor blood sugar frequently, and carry glucose tablets for emergencies. GLP-1 agonists require none of this. The reduced monitoring burden and lower anxiety around hypoglycemia consistently improve quality of life measures in comparative trials.

However, the advantage narrows when GLP-1 agonists are combined with sulfonylureas, which independently stimulate insulin secretion without glucose dependence. In SURPASS-2, hypoglycemia (blood glucose below 54 mg/dL) occurred in 0.2% to 1.7% of tirzepatide-treated patients and 0.4% of semaglutide-treated patients, all on background metformin without sulfonylureas.^[3]

Cardiovascular Outcomes: Where GLP-1 Agonists Pull Ahead

Insulin has never been proven to reduce cardiovascular events in type 2 diabetes. GLP-1 agonists have, repeatedly, in large randomized trials.

The LEADER trial randomized 9,340 patients with type 2 diabetes and high cardiovascular risk to liraglutide or placebo over a median 3.8 years.^[4] Liraglutide reduced the primary composite outcome of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke by 13% (hazard ratio 0.87, P = 0.01 for superiority). Cardiovascular death specifically fell by 22% (hazard ratio 0.78, P = 0.007), and all-cause mortality dropped by 15%.

SUSTAIN-6 showed even larger effects with semaglutide. Among 3,297 high-risk patients followed for 104 weeks, semaglutide reduced the composite cardiovascular endpoint by 26% (hazard ratio 0.74).^[5] Nonfatal stroke was reduced by 39% (hazard ratio 0.61, P = 0.04). One complication emerged: retinopathy events increased with semaglutide (hazard ratio 1.76), which subsequent analysis linked to the speed of HbA1c reduction rather than a direct drug effect.^[6]

The REWIND trial extended these findings to a broader population. Among 9,901 patients followed for a median 5.4 years, dulaglutide reduced major cardiovascular events by 12% (hazard ratio 0.88, P = 0.026).^[7] Critically, only 31% of REWIND participants had established cardiovascular disease at baseline, suggesting that GLP-1 agonists may provide cardiovascular protection even in lower-risk populations.

Taken together, three major trials enrolling over 22,500 patients demonstrated consistent cardiovascular protection from GLP-1 agonists. No insulin trial in type 2 diabetes has shown a comparable benefit. The large ORIGIN trial (2012) randomized 12,537 patients to insulin glargine versus standard care and found no cardiovascular benefit from insulin therapy over a median 6.2 years, reinforcing the distinction between glucose lowering and cardiovascular protection.

These cardiovascular benefits represent the strongest argument for choosing GLP-1 agonists over insulin. Guidelines now specifically recommend GLP-1 agonists for patients with type 2 diabetes and established atherosclerotic cardiovascular disease or high cardiovascular risk, regardless of HbA1c level.^[2]

When Insulin Remains the Right Choice

Despite the advantages of GLP-1 agonists, insulin remains essential for several clinical scenarios.

Type 1 diabetes. GLP-1 agonists cannot replace insulin in type 1 diabetes. Without functional beta cells, there is no insulin-producing machinery for GLP-1 to stimulate. A review by Guyton and colleagues found that adding GLP-1 agonists to insulin in type 1 diabetes produced modest HbA1c reductions of approximately 0.4%, decreased insulin requirements, and significant weight loss of around 5 kg.^[8] This positions GLP-1 agonists as potential adjuncts, not replacements, for insulin in type 1 diabetes.

Advanced beta cell failure. Some patients with long-standing type 2 diabetes have severely depleted beta cell reserves. When beta cells can no longer respond to GLP-1 stimulation adequately, exogenous insulin becomes necessary.

Acute hyperglycemic emergencies. Diabetic ketoacidosis and hyperosmolar hyperglycemic states require insulin for immediate glucose control. GLP-1 agonists act too slowly and through too indirect a mechanism for emergency management.

Very high HbA1c at diagnosis. When HbA1c exceeds approximately 10% at diagnosis, many guidelines recommend starting insulin to achieve rapid glycemic control, then considering transition to GLP-1 agonists once glucose levels stabilize.

Pregnancy. GLP-1 agonists are not approved for use during pregnancy. Women with diabetes who become pregnant are managed with insulin.

Gastrointestinal intolerance. GLP-1 agonists cause nausea in 17% to 22% of patients, with vomiting in 6% to 10% and diarrhea in 13% to 16%.^[3] While these side effects usually diminish over weeks with dose titration, a subset of patients cannot tolerate any GLP-1 agonist. For these individuals, insulin remains an effective alternative without gastrointestinal effects.

The decision between GLP-1 agonists and insulin is not permanent. Patients who start on insulin can transition to GLP-1 agonists as their clinical situation evolves, and those who begin with GLP-1 agonists may eventually need insulin added as beta cell function declines over the natural history of type 2 diabetes.

Combining GLP-1 Agonists and Insulin

The choice is not always either/or. Fixed-ratio combinations of a GLP-1 agonist with basal insulin (such as insulin degludec/liraglutide or insulin glargine/lixisenatide) are approved therapies. These combinations exploit the complementary mechanisms of each drug class: insulin handles fasting and overnight glucose control, while the GLP-1 agonist manages post-meal spikes, limits weight gain, and reduces the insulin dose needed.^[2]

This combination approach is particularly useful for patients who have been on basal insulin but cannot achieve target HbA1c without adding mealtime insulin. Instead of progressing to multiple daily insulin injections, adding a GLP-1 agonist often achieves comparable glycemic control with the added benefits of weight reduction and lower hypoglycemia risk.

The Cost and Access Equation

GLP-1 agonists have historically been substantially more expensive than insulin, particularly generic or biosimilar insulin formulations. In the United States, monthly costs for brand-name GLP-1 agonists have ranged from $800 to over $1,000 without insurance, compared to as little as $25 for generic insulin glargine.

This cost gap has been the primary barrier to broader adoption. A cost-effectiveness analysis must weigh the higher drug cost against reduced hypoglycemia-related emergency visits, fewer cardiovascular events, and lower long-term costs associated with weight management and metabolic improvement. As patents expire and biosimilar GLP-1 agonists enter the market, the economic calculus is shifting, but access remains uneven globally.

The cost-effectiveness debate is further complicated by the fact that many patients on insulin also require glucose monitoring supplies, which adds to the total cost of insulin-based therapy.

The Bottom Line

Head-to-head trial data consistently favor GLP-1 receptor agonists over insulin for most patients with type 2 diabetes, with advantages in glycemic control, weight, hypoglycemia risk, and cardiovascular outcomes. Insulin remains essential for type 1 diabetes, advanced beta cell failure, and acute hyperglycemic emergencies. The two drug classes work well in combination, and the optimal choice depends on the individual patient's metabolic state, cardiovascular risk, and treatment goals.

Frequently Asked Questions

Can GLP-1 agonists replace insulin for type 2 diabetes?

For most people with type 2 diabetes, GLP-1 agonists can be used instead of starting insulin therapy. A meta-analysis of 19 head-to-head trials found GLP-1 agonists matched or slightly exceeded insulin for HbA1c reduction while producing 3.71 kg more weight loss and 34% fewer hypoglycemic episodes. Current guidelines from the ADA and EASD recommend GLP-1 agonists as the preferred first injectable therapy over insulin for type 2 diabetes, except when type 1 diabetes, severe hyperglycemia, or catabolic features are present.

Do GLP-1 agonists work better than insulin for blood sugar control?

GLP-1 agonists reduce HbA1c by about 0.12 percentage points more than insulin on average, based on pooled data from 19 clinical trials. However, basal insulin lowers fasting plasma glucose more effectively, by approximately 1.8 mmol/L. The newer GLP-1 agonists, particularly semaglutide and tirzepatide, produce HbA1c reductions of 1.8% to 2.3%, which exceeds what most insulin regimens achieve in practice.

Why do GLP-1 agonists cause less hypoglycemia than insulin?

GLP-1 agonists stimulate insulin release only when blood glucose is elevated. This glucose-dependent mechanism means insulin secretion automatically decreases as blood sugar approaches normal levels. Injected insulin has no such safety mechanism and continues lowering blood sugar regardless. This difference resulted in 34% fewer hypoglycemic episodes with GLP-1 agonists compared to insulin across 19 head-to-head trials.

Can you take GLP-1 agonists and insulin together?

Yes. Fixed-ratio combinations of a GLP-1 agonist with basal insulin are approved therapies. Insulin degludec/liraglutide and insulin glargine/lixisenatide combine the fasting glucose control of basal insulin with the post-meal glucose and weight management benefits of GLP-1 agonists. This approach often achieves better glycemic control than basal insulin alone while limiting weight gain and reducing total insulin dose requirements.

Do GLP-1 agonists reduce heart attack and stroke risk better than insulin?

Three large cardiovascular outcome trials, LEADER (liraglutide, 9,340 patients), SUSTAIN-6 (semaglutide, 3,297 patients), and REWIND (dulaglutide, 9,901 patients), demonstrated that GLP-1 agonists reduce major cardiovascular events by 12% to 26%. Liraglutide reduced cardiovascular death by 22%. No comparable cardiovascular benefit has been demonstrated for insulin therapy in type 2 diabetes.

Why are GLP-1 agonists preferred over insulin now?

Guidelines shifted after accumulating evidence showed GLP-1 agonists provide similar or better glycemic control than insulin with additional benefits: weight loss instead of weight gain, lower hypoglycemia risk, cardiovascular event reduction, and simpler dosing (once weekly vs daily injections with monitoring). The ADA/EASD consensus now recommends GLP-1 agonists as the preferred first injectable for type 2 diabetes over insulin in most clinical scenarios.