A FormBlends network publication
The Science

Cardiovascular outcomes: what SELECT proved and what it left open

The SELECT trial's design and numbers, the candidate mechanisms for a 20% reduction in major cardiovascular events, why the early separation of the curves matters, and why the mechanism is graded debated on this site.

By FormBlends editorial teamUpdated September 4, 2026Educational, not medical advice

SELECT is the trial that moved GLP-1 medicines from weight drugs to cardiovascular drugs in the eyes of regulators. It is also a clean example of a proven effect with an unproven mechanism.

The trial

17,604 adults aged 45 or over with a BMI of 27 or higher and established cardiovascular disease (prior heart attack, stroke or peripheral artery disease), without diabetes, were randomised to semaglutide 2.4 mg weekly or placebo and followed for a mean of 39.8 months. The primary endpoint was a composite of cardiovascular death, non-fatal heart attack and non-fatal stroke. It occurred in 6.5% of the semaglutide group and 8.0% of the placebo group, hazard ratio 0.80 (95% CI 0.72 to 0.90), P less than 0.001 (PubMed 37952131).

Secondary results pointed the same way: death from any cause, hazard ratio 0.81 (95% CI 0.71 to 0.93); the heart failure composite, 0.82 (0.71 to 0.96); cardiovascular death, 0.85 (0.71 to 1.01), which did not reach significance under the trial's hierarchical testing. Mean weight change was -9.4% versus -0.9%. Adverse events leading to discontinuation were more common on semaglutide, 16.6% versus 8.2%, mostly gastrointestinal. In March 2024 FDA added the reduction of major adverse cardiovascular events to the Wegovy label.

Why the mechanism is unresolved

Two features of the data matter. The Kaplan-Meier curves separated early, within the first months, when average weight loss was still modest. And the treatment effect was consistent across subgroups defined by baseline BMI. Both observations make "the weight loss did it" incomplete as an explanation, without ruling weight out as a contributor.

Candidate mechanisms for the cardiovascular benefit seen in SELECTA semaglutide box on the left connects through six intermediate boxes to an outcome box on the right labelled MACE down 20 percent, grade A. The intermediates are: body weight down 9.4 percent; systolic blood pressure down; LDL and triglycerides down; HbA1c down; C-reactive protein down; and direct vascular or sinoatrial receptor effects. Each intermediate is annotated with a grade for the link to outcome, all D, and the outcome box is annotated grade A. A note says that the curves separated within months, before most weight was lost, and that the effect was consistent across BMI subgroups.Semaglutide2.4 mg weeklyBody weightSystolic blood pressureLDL, triglyceridesHbA1cInflammation (CRP)Direct receptor effects-9.4% vs -0.9% (measured, A)lower on treatment (measured, A)lower on treatment (measured, A)lower on treatment (measured, A)lower on treatment (measured, A)SA node, arterioles; no ventricular receptor (B)each linkto outcome: DMACE down 20%HR 0.80 (0.72 to 0.90)grade ATwo cluesCurves separated within months,before most weight was lost.Effect consistent across baselineBMI subgroups.Lincoff 2023, figure and subgroups
Figure 15. Candidate pathways from semaglutide to fewer cardiovascular events, with this site's evidence grade for each link. The outcome is grade A; every arrow is a hypothesis about how the outcome arose. Drawn from Lincoff 2023, Drucker 2016 and Pyke 2014.

The candidates

Weight. A 9.4% loss improves nearly every risk factor. Against it as the sole explanation: the timing and the BMI-subgroup consistency above.

Blood pressure, lipids, glucose. All moved favourably. Each is an established cardiovascular risk factor, and each moved less than would be expected to produce a 20% event reduction on its own over three years.

Inflammation. C-reactive protein fell substantially in SELECT and in STEP-HFpEF. Whether it is a mediator or a marker of losing inflamed adipose tissue is unknown.

Direct receptor effects. The heart's pacemaker tissue and some arterioles carry the receptor; ventricular muscle and, on validated mapping, most of the vasculature do not (PubMed 24467746). Drucker's review sets out the animal evidence for effects on endothelium, plaque and ischaemic injury, and is frank that much of it has not been shown in people (PubMed 27423522).

This site grades the mechanism D and the outcome A. In the mechanism map the heart node shows both.

Context from diabetes

SUSTAIN-6, in 3,297 people with type 2 diabetes, found a hazard ratio of 0.74 (95% CI 0.58 to 0.95) for the same composite with injectable semaglutide over about two years (PubMed 27633186). It was a noninferiority trial that happened to show superiority, and the benefit was driven largely by fewer strokes. SELECT confirmed the direction in a population without diabetes and with three times the follow-up. Together they establish the class effect for semaglutide; they still do not name the mechanism.

Where to go next

Questions people ask

Does SELECT mean the weight loss is what protects the heart?

It cannot show that. The event curves separated early, before most of the weight was lost, and the benefit was consistent across baseline BMI subgroups, which points to effects beyond weight. But the trial randomised a drug, not a mechanism, and it cannot apportion credit between weight, blood pressure, lipids, glucose, inflammation and direct vascular effects.

Does SELECT apply to people with diabetes, or to compounded semaglutide?

SELECT excluded diabetes; the diabetes cardiovascular evidence for semaglutide comes from SUSTAIN-6 and, for the tablet, PIONEER 6. Compounded semaglutide was not studied in any of these trials, is not FDA approved and cannot be assumed to reproduce the result.

Sources

  1. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes (SELECT). N Engl J Med 2023. PubMed 37952131 Accessed September 4, 2026.
  2. Marso SP, Bain SC, Consoli A, et al. Semaglutide and cardiovascular outcomes in patients with type 2 diabetes (SUSTAIN-6). N Engl J Med 2016. PubMed 27633186 Accessed September 4, 2026.
  3. Drucker DJ. The cardiovascular biology of glucagon-like peptide-1. Cell Metab 2016. PubMed 27423522 Accessed September 4, 2026.
  4. Pyke C, Heller RS, Kirk RK, et al. GLP-1 receptor localization in monkey and human tissue. Endocrinology 2014. PubMed 24467746 Accessed September 4, 2026.
  5. Wegovy (semaglutide) prescribing information. Drugs@FDA, NDA 215256 Accessed September 4, 2026.

Canonical URL: https://formblendsscience.com/outcomes/cardiovascular-select. Written by the FormBlends editorial team. This page is educational and is not medical advice; see the medical disclaimer.