What SELECT Showed: Semaglutide Cuts Heart Attack Risk by 20% in Non-Diabetics with Cardiovascular Disease

Until 2023, every major cardiovascular outcomes trial of a GLP-1 receptor agonist had been done in patients with type 2 diabetes. That changed when the SELECT trial published in the New England Journal of Medicine in November 2023, showing that weekly semaglutide 2.4 mg cut major adverse cardiovascular events by 20% in non-diabetic adults with overweight or obesity and established cardiovascular disease [1]. SELECT is now the single largest piece of evidence for thinking of GLP-1 therapy as a cardiovascular disease intervention rather than just a weight loss intervention. This article walks through what the trial actually measured, the surprising finding that the cardiovascular benefit appears largely independent of how much weight patients lost, and the FDA label expansion that followed.

SELECT trial design

SELECT randomized 17,604 patients across 41 countries to weekly subcutaneous semaglutide 2.4 mg or matching placebo, with both arms receiving standard background cardiovascular therapy and lifestyle counseling [1]. The inclusion criteria were unusually specific:

• Age ≥ 45 years
• BMI ≥ 27 kg/m² (overweight or obese category)
• Established cardiovascular disease at baseline (prior MI, prior stroke, or symptomatic peripheral artery disease)
• No history of diabetes — HbA1c < 6.5% at screening and no glucose-lowering agent use in the prior 90 days

The diabetes exclusion is the key feature that makes SELECT different from every prior GLP-1 cardiovascular outcomes trial. LEADER (liraglutide), SUSTAIN-6 (semaglutide), REWIND (dulaglutide), and PIONEER-6 (oral semaglutide) all enrolled patients with type 2 diabetes — they answered the question “does GLP-1 therapy reduce cardiovascular events in diabetics?” SELECT was designed to answer the more contested question: does GLP-1 therapy reduce cardiovascular events in obese non-diabetic patients?

The primary endpoint was a three-component composite of major adverse cardiovascular events (MACE): cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke (time-to- first-event analysis) [1]. Mean follow-up was 39.8 months — nearly four years, the longest sustained GLP-1 trial in obese non-diabetic patients to date.

The primary result: 20% MACE reduction

The primary endpoint hit by a comfortable margin. The MACE composite occurred in [1]:

The relative risk reduction is 20%. The absolute risk reduction is 1.5 percentage points. The number needed to treat (NNT) over the ~3.3-year mean follow-up is approximately 67 — meaning 67 patients need to be treated for roughly three years to prevent one major cardiovascular event. For comparison, the NNT for high-intensity statin therapy in secondary prevention is in roughly the same range (~30–60 depending on the population), and that's considered one of the highest-value drug interventions in modern cardiology.

The heart failure secondary endpoint

SELECT's confirmatory secondary endpoint was a composite of cardiovascular death, hospitalization for heart failure, or urgent medical visit for heart failure. That composite also favored semaglutide [1]:

The heart failure signal is consistent with the parallel STEP- HFpEF trial (Kosiborod et al, NEJM 2023), which showed that semaglutide improves symptoms and exercise capacity in patients with heart failure with preserved ejection fraction (HFpEF) and obesity [4]. Together, the two trials make a strong case for GLP-1 therapy in obese patients with — or at risk of — heart failure, regardless of diabetic status.

The most surprising finding: weight loss didn't predict cardiovascular benefit

SELECT's most clinically important sub-analysis examined whether the cardiovascular benefit was driven by how much weight each patient actually lost on semaglutide. The trial team looked at patients' weight change at week 20 — early in the treatment course — and asked whether patients who lost more weight had bigger cardiovascular benefits.

They didn't. Early weight loss at week 20 did NOT predict subsequent cardiovascular outcomes in the semaglutide group. Greater reductions in waist circumference (a marker of visceral fat) were modestly associated with lower MACE risk, but the magnitude of weight loss itself was not. The sub-analysis authors concluded that “prescribing restrictions based on BMI thresholds or weight-loss targets may not be appropriate, as patients might benefit regardless of weight-loss response.”

Translation: in SELECT's population, the cardiovascular benefit of semaglutide appears to be substantially weight-independent. The drug did its cardiovascular work through mechanisms that aren't fully captured by the scale.

This is a big deal because the dominant clinical narrative until SELECT was “GLP-1 drugs help your heart by helping you lose weight.” The actual mechanism appears to be a combination of:

1. Weight loss — yes, contributing, but smaller than expected. SELECT participants lost a mean 10.2% of starting body weight at 208 weeks of follow-up, which is actually modest compared with the obesity-focused STEP-1 trial (15–17% at 68 weeks). The CV benefit accumulated across the entire follow-up period, not just during the weight loss plateau.
2. Direct anti-inflammatory effects. GLP-1 receptor agonists reduce systemic inflammatory markers (CRP, IL-6) by mechanisms only partially explained by weight loss. Mechanistic studies in animal models show that controls paired for weight loss (without GLP-1 therapy) are not protected from atherosclerosis the way GLP-1-treated animals are.
3. Direct vascular effects. Improved endothelial function, modest blood pressure reduction, natriuresis, and plaque stabilization. These are weight-independent.
4. Improved cardiac filling. The HFpEF benefit (STEP-HFpEF) appears partly related to reductions in epicardial fat and improvements in left ventricular filling pressures that are not fully explained by total body weight change [4].

Adverse events: the longer follow-up exposed a gallbladder signal

SELECT's 39.8-month mean follow-up was substantially longer than any prior obesity-focused GLP-1 trial. That extra follow-up time exposed a gallbladder safety signal that the shorter obesity trials had not made prominent. Gallbladder-related adverse events were more frequent with semaglutide than placebo in SELECT, primarily driven by cholelithiasis (gallstones) rather than cholecystitis (acute inflammation) [1].

The signal is consistent with what's known about rapid weight loss in general — see our GLP-1 side effects investigation for the broader gallbladder discussion. The Wegovy label notes that “acute gallbladder events were associated with weight reduction,” which is consistent with this being a consequence of any major weight loss intervention rather than a drug-specific toxicity [5].

Notably absent from SELECT's safety reporting:

• No increased pancreatitis signal. Acute pancreatitis occurred at similar rates in both arms, consistent with the obesity-only trials and the recent tirzepatide meta-analyses.
• No retinopathy signal. Diabetic retinopathy was a flagged safety signal in SUSTAIN-6 (semaglutide in diabetics), but SELECT's non-diabetic population experienced no equivalent retinal signal.
• Net mortality numerically favored semaglutide.Cardiovascular death rate was 2.5% on semaglutide vs 3.0% on placebo (HR 0.85, 95% CI 0.71–1.01, P=0.07) — close to statistical significance but not crossing the threshold.

How SELECT compares with the diabetic GLP-1 cardiovascular trials

The hazard ratios are remarkably consistent across all three trials despite very different populations and dose levels. This consistency strengthens the case that the GLP-1 cardiovascular effect is real, robust, and not specific to the diabetic population.

FDA label expansion: March 2024

On March 8, 2024, the FDA expanded the Wegovy (semaglutide 2.4 mg) label to include cardiovascular risk reduction. The new indication permits Wegovy “to reduce the risk of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke in adults with established cardiovascular disease and either obesity or overweight” [5]. The label expansion did not require a separate diabetes diagnosis — patients qualify based on the SELECT inclusion criteria.

The label expansion has practical implications for insurance coverage. Many commercial plans previously excluded GLP-1 medications for weight loss but covered them for diabetes. SELECT created a third covered indication (cardiovascular risk reduction in non-diabetics with established CVD) that some plans now reimburse where they previously denied. See our GLP-1 insurance coverage audit for the per-insurer breakdown.

What SELECT means for the average patient

For most patients without established cardiovascular disease, SELECT's direct clinical implications are limited — the trial enrolled a high-risk secondary-prevention population, and the cardiovascular benefit shouldn't be assumed to apply to everyone with overweight or obesity.

For patients with established cardiovascular disease and overweight/obesity, SELECT changes the calculus substantially. GLP-1 therapy is now backed by Phase 3 randomized evidence in this exact population, with an effect size comparable to the highest-tier secondary prevention drug classes. That's a different kind of recommendation than “lose weight for general health benefit” — it's a specific, measurable, NNT-quantifiable cardiovascular intervention.

For patients considering compounded semaglutide for weight loss outside the SELECT inclusion criteria, the cardiovascular evidence is suggestive but not directly applicable. The compounded telehealth market sells the same active molecule that SELECT studied — see our bioequivalence investigation — but the SELECT cardiovascular benefit was demonstrated only in a specific high-risk population with established disease.

Most importantly: SELECT's weight-independent finding reframes how to think about GLP-1 therapy generally. The question is shifting from “how much weight will I lose?” to “what is GLP-1 therapy actually doing for my cardiometabolic health?” The answer, as of 2026 and increasingly, is “more than the scale shows.”