Semaglutide mechanism: the missing pieces

Semaglutide's basic mechanism is well mapped. It mimics a gut hormone, binds a receptor found on cells in the pancreas, stomach and brain, and produces measurable effects on insulin release, appetite and digestion. What is not fully mapped is how much of that mechanism explains semaglutide's biggest cardiovascular result: a 20% reduction in heart attack, stroke and cardiovascular death, reported in the SELECT trial of adults with obesity but no diabetes^[3]. A September 2026 editorial in the European Heart Journal, written by cardiologists Subodh Verma, Tomasz Guzik and Francesco Cosentino, names that gap directly. Its title is "Semaglutide mechanisms of action: the missing pieces."^[1]

What the established mechanism explains

Semaglutide is a synthetic analogue of GLP-1 (glucagon-like peptide-1), a hormone the gut releases after eating. It binds the GLP-1 receptor, the same receptor native GLP-1 activates, on cells in the pancreas, stomach and brain^[4]. Three effects follow from that binding: the pancreas releases more insulin, but only when blood sugar is already elevated; the liver gets a weaker signal to release stored glucose; and the stomach empties more slowly, which is part of why nausea is the most common early side effect. In the brain, the same receptor sits in the hypothalamus and brainstem, the regions that govern hunger. The EMA's authorisation for Wegovy describes the appetite effect in plain terms: semaglutide increases fullness while reducing food intake, hunger and cravings^[4]. None of this is in dispute, but on its own it is an incomplete explanation for what the SELECT trial found.

The EMA authorises semaglutide under the Wegovy brand for adults with a BMI of 30 or above, or 27 or above with a weight-related condition such as high blood pressure or elevated cholesterol, and separately for adolescents aged 12 and up with obesity, with treatment expected to stop if a minimum weight response is not met within the first 12 weeks^[4]. The cardiovascular-risk-reduction indication that SELECT supports sits inside that same approved population: overweight or obese adults, with or without diabetes, who already have cardiovascular disease.

The gap the mediation analysis exposed

That mechanism accounts well for why semaglutide lowers blood sugar and body weight, but it does not fully account for SELECT's cardiovascular result. A separate analysis of the same trial data, led by Helen Colhoun at the University of Edinburgh, tested that assumption directly, calculating how much of the 20% reduction in major cardiovascular events can be statistically traced back to the risk factors semaglutide is known to improve^[2].

The published results, in the same journal, found that semaglutide significantly improved every mediator the researchers measured: body weight, waist circumference, blood pressure, cholesterol, blood sugar, kidney function markers, and high-sensitivity CRP (a marker of inflammation). But when the researchers calculated how much of the cardiovascular benefit each factor could explain on its own, the estimates were partial and inconsistent. Body weight alone explained an estimated 19.5% of the benefit. Waist circumference explained 64.0%. CRP explained 42.1%. HbA1c (a blood sugar marker) explained 29.0%. Combined in a single model, the measured factors accounted for an estimated 31.4%, with a confidence interval wide enough to run from negative to well over 100%, a sign of how much statistical noise sits inside this kind of analysis^[2].

The paper's own conclusion is blunt: the mediation analyses "could not fully ascribe the effects of semaglutide on MACE to known risk factors in SELECT with any certainty."^[2] Medical Xpress, reporting on the paper, quoted Colhoun making the same point in plainer language and naming anti-inflammatory action and direct effects on the heart and blood vessels as candidates that still need dedicated study^[6].

What this is not

The numbers do not add up to one tidy story. Waist circumference alone plausibly channels most of the benefit under one reading, body weight a much smaller share under another, and the honest answer sits somewhere the trial was not designed to isolate cleanly. Weight loss still matters here, but it is not sufficient on its own, and no single alternative mechanism has taken its place. What the analysis rules out is the simplest version of the story: that semaglutide protects the heart purely by making people thinner. Verma and his co-authors made a related point in their companion editorial, which Medical Xpress summarised as arguing that the incomplete mechanistic picture should not be treated as a barrier to prescribing semaglutide to reduce vascular events^[6]. That is a judgment about clinical practice, not a claim that the mechanism is solved.

Why this matters

For someone deciding whether to start semaglutide, the unresolved mechanism does not change the trial results already on the label: a documented reduction in cardiovascular events in a specific population, alongside a well-characterised side-effect profile. The most common adverse reactions are gastrointestinal (nausea, diarrhoea, vomiting, constipation and abdominal pain), and the US label carries a boxed warning for thyroid C-cell tumours seen in rodent studies, with medullary thyroid carcinoma and MEN 2 syndrome listed as contraindications^[5]. What the mediation analysis changes is how confidently anyone, researcher, prescriber or drug company, can claim to know why the drug works. Readers considering semaglutide for any indication should read the full regulatory picture on our semaglutide page and talk to a clinician about their personal cardiovascular risk profile, not just the trial's topline percentage.