In November 2023, Lincoff and colleagues published SELECT in the New England Journal of Medicine — a 17,604-patient randomized trial showing that semaglutide 2.4 mg weekly reduced the composite of cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke by 20% versus placebo in adults with overweight or obesity. That result moved semaglutide from "weight-loss drug with incidental metabolic benefits" into cardiovascular-risk-reduction territory.
What SELECT did not establish is whether GLP-1 receptor agonists prevent first cardiovascular events in obese or overweight patients who do not already have established cardiovascular disease. That is the primary-prevention question, and it is distinct from what SELECT answered. The field in 2026 is still working through it. This piece walks through what SELECT showed, why the primary-prevention extrapolation is not automatic, and where the data currently sits.
What SELECT actually studied
SELECT enrolled adults aged 45 and older with BMI ≥27 kg/m² and established atherosclerotic cardiovascular disease — defined as prior myocardial infarction, prior stroke, or symptomatic peripheral artery disease. Diabetes was an exclusion. Over a mean follow-up of approximately 39.8 months, a primary endpoint event occurred in 6.5% of the semaglutide group versus 8.0% of the placebo group (hazard ratio 0.80; 95% CI 0.72–0.90; P<0.001).
That is a secondary-prevention result. Every participant had at least one prior CV event or established symptomatic atherosclerosis at enrollment. The population the trial was powered to inform is the population with known disease — and the result in that population is durable, unambiguous, and clinically meaningful.
Why primary prevention is a separate question
Primary prevention — preventing a first event — is a different statistical and clinical problem. Event rates in an obese population without prior CVD are lower, which means a trial needs either much larger enrollment, much longer follow-up, or a much higher-risk baseline population to achieve the same statistical power. It also means the absolute-risk reduction for any given hazard ratio is smaller, which bears on the cost-benefit calculation from a public-health perspective.
There is a mechanistic argument that GLP-1 benefits should extend to primary prevention — weight loss, blood-pressure reduction, improved glycemic control, and direct anti-inflammatory effects all plausibly reduce first-event risk. The argument is reasonable. It is not the same thing as trial evidence.
What the 2026 evidence base actually looks like
Three data strands bear on the primary-prevention question as of April 2026.
First, SURPASS-CVOT — the tirzepatide cardiovascular outcomes trial — published in NEJM in December 2025 (Sattar et al., DOI 10.1056/NEJMoa2505928). Over a median 4-year follow-up, the primary composite endpoint of cardiovascular death, MI, or stroke occurred in 12.2% on tirzepatide versus 13.1% on dulaglutide. The trial met noninferiority but not superiority. A pre-specified expanded MACE endpoint (adding coronary revascularization) significantly favored tirzepatide. This was a secondary-prevention design in type 2 diabetes patients with established ASCVD, not a primary-prevention trial.
Second, SURMOUNT-MMO is Lilly's ongoing event-driven trial studying tirzepatide in adults with obesity and without diabetes who have cardiovascular disease or are at risk — a population that includes patients without prior CV events. As of April 2026, the trial is ongoing; no primary readout has been reported.
Third, observational analyses of real-world cohorts (type 2 diabetes and obesity) have shown cardiovascular benefits in populations that extend beyond established CVD, but observational data cannot substitute for a powered randomized primary-prevention trial on this question.
The honest 2026 summary: the evidence base for secondary prevention with GLP-1 and dual-agonist drugs in obese patients with established CVD or T2D-plus-ASCVD is strong. The evidence base for primary prevention in obese patients without prior CVD is mechanistic, extrapolated, and not yet supported by a published, powered, placebo-controlled primary-prevention readout.
What this means for readers
For a reader evaluating whether semaglutide or tirzepatide is the right tool for a specific clinical situation, the SELECT evidence is load-bearing if there is prior ASCVD. It is suggestive but not confirmatory if there is not. Our semaglutide-vs-tirzepatide comparison walks through the head-to-head weight-loss and safety profile; our GLP-1 pillar covers the cardiovascular story in more depth; and our peptide-safety reference walks through the general framework for weighing trial-population applicability against individual clinical context.
Waiting for SURMOUNT-MMO and any subsequent primary-prevention-specific readouts is the right call for patients and clinicians who want the primary-prevention question answered with trial-grade data rather than mechanistic extrapolation.