Incretin drugs: outcomes before weight loss

A review published in September 2026 in Diabetes, Obesity and Metabolism asks a question that rarely makes headlines: when a clinician is choosing between a GLP-1 receptor agonist, a dual GIP/GLP-1 agonist, or the emerging triple agonists, what should the hierarchy of goals actually look like^[1]? The paper by Sztanek, Lorincz, Harangi, and Nagy at the University of Debrecen proposes a framework that ranks outcomes by evidence maturity, then uses that ranking to guide drug-to-patient matching.

The short version: cardiovascular mortality sits at the top. Major adverse cardiovascular events (MACE) come next. Weight loss and glycaemic control, for all the attention they receive, occupy lower rungs. The review argues that treating the hierarchy in reverse order, chasing weight numbers before asking about hard outcomes, leads to suboptimal clinical decisions. That argument is worth unpacking.

The outcome hierarchy

The Debrecen group structures the hierarchy around the quality and completeness of available trial evidence. CV mortality reduction requires a dedicated cardiovascular outcomes trial (CVOT) with a long follow-up and hard endpoints. The SELECT trial, which enrolled 17,604 adults with obesity and established cardiovascular disease but without diabetes, provides the clearest example for semaglutide: weekly 2.4 mg semaglutide reduced the composite of cardiovascular death, non-fatal heart attack, and non-fatal stroke to 6.5% from 8.0% on placebo, a hazard ratio of 0.80^[2]. That is a hard outcome, in a well-powered trial, in patients who do not have diabetes.

MACE reduction is closely related but distinct. A drug that cuts MACE in a trial dominated by people with type-2 diabetes may not generalise to the obesity-only population, and vice versa. The review treats these as separate rungs because the underlying populations and the evidence bases diverge.

Weight loss, despite driving most prescribing conversations, is classified as an intermediate endpoint. It correlates with CV benefit, but correlation is not causation. The drugs may reduce cardiovascular risk through pathways that operate partly independently of weight change, including direct effects on inflammation, endothelial function, and blood pressure. Glycaemic control is similarly downstream: important for people with type-2 diabetes, but not a surrogate for the outcomes that determine longevity.

Receptor-network pharmacology: why the drug chosen matters

The second pillar of the framework is receptor pharmacology. The current generation of incretin drugs does not act on a single switch. GLP-1 receptor agonists such as semaglutide activate the GLP-1R, which is expressed in the pancreas, heart, vasculature, brain, and kidneys. Dual agonists such as tirzepatide add GIP receptor activation. Triple agonists in late-stage development add glucagon receptor activity. Each receptor layer contributes differently.

GLP-1R activation slows gastric emptying, reduces appetite, and augments insulin secretion in a glucose-dependent manner. The cardiovascular benefits documented in the SELECT and LEADER trials appear to be at least partly mediated through direct GLP-1R effects on cardiac and vascular tissue, not solely through weight loss. GIP receptor activity adds to insulin secretion and, in adipose tissue, appears to influence lipid handling and adiponectin release. Glucagon receptor agonism increases energy expenditure and, in the liver, promotes glycogenolysis, making it particularly relevant in conditions where hepatic fat accumulation is a driver.

The practical implication, as the review frames it, is that drug selection should reflect the receptor network most relevant to the patient's phenotype. A patient whose dominant risk is atherosclerotic cardiovascular disease has different receptor-level needs than a patient whose primary problem is hepatic steatosis or who is preparing for bariatric surgery.

Phenotype-based clinical positioning

The third pillar is matching drug to patient phenotype. The Debrecen group identifies several phenotypic clusters where the evidence base tilts clearly in one direction.

High atherosclerotic cardiovascular risk

For patients with established ASCVD and obesity but without diabetes, the review positions semaglutide as the drug with the most mature hard-outcome evidence, based specifically on SELECT^[2]. The authors note that GLP-1 receptor agonists with proven CV benefit should be given precedence when atherosclerotic risk reduction is the primary goal.

Type-2 diabetes plus obesity

In patients with both conditions, dual agonists offer meaningful advantages on glycaemia and weight, with SURPASS trial data supporting tirzepatide's superiority over semaglutide on HbA1c and body weight. The review acknowledges that the CV outcomes trial for tirzepatide (SURPASS-CVOT) was still running as of the publication date, so head-to-head hard-outcome comparisons in this population were not yet available.

Metabolic-associated steatotic liver disease

Patients with significant hepatic fat accumulation represent a phenotype where glucagon receptor activity may contribute additional benefit by promoting hepatic fat clearance. Triple agonists and, to a lesser extent, dual agonists that include some effective glucagon receptor engagement appear better positioned here than GLP-1-only drugs, although the evidence base for triple agonists remains preliminary relative to the approved class members.

What the framework does not resolve

The review is candid about its limits. The outcome hierarchy is evidence-derived but static: as new CVOT data emerge (tirzepatide's, retatrutide's), the hierarchy for those agents will update. Phenotype-based positioning assumes clean patient profiles, when in practice most patients carry multiple overlapping risk factors that cut across the neat categories. The framework also does not address cost, access, or tolerability, all of which shape real-world prescribing as much as receptor pharmacology does. Gastrointestinal adverse events, loss of lean mass with aggressive weight loss, and the long-term safety profile of newer agents remain practical constraints the hierarchy cannot resolve.

The harder limitation is that the reviewed evidence base is still largely short-horizon. Most outcome trials run two to five years. Whether the cardiovascular benefits seen at those timepoints persist, grow, or attenuate over a decade of treatment is unknown for any agent in this class.

What readers should take from this

The Debrecen framework is useful not because it tells clinicians what to prescribe, but because it makes the logic behind the choice explicit. If you are reading about incretin drugs for a family member with heart disease and obesity, the relevant question is not which drug produces the largest number on a scale, but which drug has evidence at the level of cardiovascular death and MACE. For a broader view of how semaglutide's trial programme developed over time, including the SELECT and STEP trials, our peptide page has the regulatory and evidence detail.

If the framework has a single actionable takeaway for the informed non-specialist: weight loss is not the goal. It is, at best, a marker that something useful may be happening elsewhere.