SELECT Trial: Semaglutide Cut Major Cardiovascular Events by 20% in Adults Without Diabetes

What the SELECT Trial Was and Why It Was Designed

SELECT was a double-blind, randomized, placebo-controlled outcomes trial that ran from October 2018 through June 2023. The question was specific: does semaglutide 2.4 mg reduce MACE in people who are overweight or obese, have established atherosclerotic cardiovascular disease, but have never been diagnosed with type 2 diabetes? That last condition mattered. Every major GLP-1 cardiovascular outcomes trial before SELECT, including LEADER (liraglutide) and SUSTAIN-6 (semaglutide 0.5/1 mg), enrolled patients with type 2 diabetes. SELECT was the first large-scale cardiovascular outcomes trial to test a GLP-1 receptor agonist in a non-diabetic population at high cardiac risk.

Eligibility required age 45 or older, BMI of 27 or above, and documented CVD (prior myocardial infarction, stroke, or symptomatic peripheral artery disease). Patients with a current or prior diagnosis of type 2 diabetes were excluded. The 17,604 participants were randomized 1:1 to weekly subcutaneous semaglutide 2.4 mg or matched placebo, titrated over 16 weeks to the full dose per the standard Wegovy escalation schedule [1].

The primary endpoint was time to first occurrence of cardiovascular death, nonfatal MI, or nonfatal stroke. The trial was event-driven, requiring at least 1,225 adjudicated MACE events to reach the pre-specified 80% power for a hazard ratio of 0.84.

Primary MACE Result: 20% Relative Risk Reduction

Semaglutide reduced the composite MACE endpoint by 20% versus placebo (HR 0.80; 95% CI 0.72, 0.90; P<0.001) [1]. MACE occurred in 6.5% of the semaglutide group versus 8.0% of the placebo group across the median 39.8-month follow-up. That absolute risk reduction of 1.5 percentage points works out to a number needed to treat of roughly 67 patients over about three years to prevent one MACE event.

Breaking down the composite: nonfatal MI drove much of the effect (HR 0.72; 95% CI 0.61, 0.85). Cardiovascular death trended in the same direction (HR 0.85; 95% CI 0.71, 1.01), though that component alone did not cross the conventional significance threshold. Nonfatal stroke showed a directionally consistent reduction (HR 0.93; 95% CI 0.74, 1.15) that was also not significant individually, which the trial authors attributed to lower statistical power for that endpoint alone.

The trial investigators, led by A. Michael Lincoff, MD, reported that semaglutide was superior to placebo for reducing this composite endpoint in this population [1]. The Kaplan-Meier curves separated within the first six months and continued to diverge through month 40, suggesting a benefit that accumulates over time rather than appearing early and flattening out. Readers who want the exact published wording of the conclusion should consult the primary paper directly rather than a secondhand quotation.

Was the Benefit From Weight Loss or a Direct Drug Effect?

This is the most debated mechanistic question in the SELECT data. Mean body weight fell 9.4% in the semaglutide arm at 104 weeks versus 0.9% in the placebo arm. The question is whether the cardiovascular benefit is simply downstream of losing roughly 8 to 9 kg more than placebo, or whether semaglutide also acts on the vasculature and myocardium directly.

The SELECT authors ran exploratory mediation analyses and found that weight loss accounted for only a minority of the observed risk reduction. Blood pressure reductions (systolic BP fell approximately 3.3 mmHg more with semaglutide), favorable lipid shifts, and reduced systemic inflammation all contributed. Waist circumference dropped 7.7 cm with semaglutide versus 1.3 cm with placebo, and hsCRP, a marker of inflammation, fell more in the semaglutide group.

GLP-1 receptors have been identified in cardiac and vascular tissue in human and animal studies, and some animal work suggests a degree of direct myocardial protection. SELECT was not designed to isolate that mechanism in humans, and the trial itself does not prove a direct cardiac drug effect. What the mediation data do suggest is that weight loss alone is unlikely to be the whole explanation for a 20% MACE reduction, since the effect on hard events was larger than would be expected from the degree of weight change by itself.

Who Qualifies: The SELECT Population Profile

The enrolled population was predominantly male (72%), white (77%), with a mean age of 61.6 years. Mean BMI at baseline was 33.3 kg/m2. All participants had at least one qualifying CVD event: prior MI (72%), prior stroke (23%), or symptomatic peripheral artery disease (12%). Baseline HbA1c was below 6.5% for all participants, confirming no diabetes at entry.

Patients with eGFR <15 mL/min/1.73m2 or active cancer were excluded. Personal or family history of multiple endocrine neoplasia type 2 (MEN2) or medullary thyroid carcinoma remained an absolute contraindication, consistent with the existing Wegovy label [3].

The FDA updated the Wegovy prescribing label in 2024 to add a cardiovascular risk reduction indication matching the SELECT population: adults with BMI ≥27 and established CVD [3]. That label update means a clinician can prescribe Wegovy for cardiac risk reduction as a standalone indication, separate from the weight management label.

How SELECT Compares to the STEP Trials

SELECT addressed cardiovascular outcomes. The STEP program addressed weight reduction. Both used semaglutide 2.4 mg weekly but answered different questions, and mixing up their results leads to prescribing confusion.

STEP-1 (N=1,961, non-diabetic adults): Semaglutide 2.4 mg produced 14.9% mean body weight loss at 68 weeks versus 2.4% with placebo, with roughly 86% of participants losing at least 5% of body weight [4].

STEP-2 (N=1,210, adults with type 2 diabetes): Mean weight loss was 9.6% with semaglutide 2.4 mg versus 3.4% placebo at 68 weeks, with HbA1c falling 1.6 percentage points in the semaglutide group [5]. Weight loss was smaller than in STEP-1, a pattern that has also been observed with other GLP-1 therapies in people with type 2 diabetes, though the exact reasons are not fully settled.

STEP-3 (N=611, semaglutide plus intensive behavioral therapy): Adding a 30-session intensive behavioral intervention to semaglutide 2.4 mg produced 16.0% mean weight loss at 68 weeks versus 5.7% with placebo plus behavioral therapy [6].

STEP-5 (N=304, 104-week extension): At two years, semaglutide 2.4 mg maintained 15.2% mean body weight reduction versus 2.6% with placebo [7], the longest randomized controlled trial data available for semaglutide weight loss.

The STEP program established the weight-loss efficacy profile. SELECT took the same drug and dose and asked whether it reduces hard cardiovascular events. Together they describe a compound with evidence for both.

Safety Profile in SELECT

The adverse event profile in SELECT matched the known GLP-1 class profile, with a large enough sample to quantify rates precisely. Gastrointestinal events were the most common reason for discontinuation: nausea occurred in 19.3% of the semaglutide group versus 11.6% placebo; diarrhea in 16.6% versus 10.7%; vomiting in 10.6% versus 5.4% [1]. These rates are broadly consistent with what STEP-1 reported over its shorter 68-week window.

Serious adverse events occurred in 33.4% of the semaglutide group versus 36.4% placebo, a lower rate on the drug arm driven largely by the reduction in MACE.

Gallbladder disease was somewhat elevated: cholelithiasis occurred in 2.8% of the semaglutide group versus 2.3% placebo. Acute pancreatitis was infrequent in both arms (0.3% vs. 0.2%). Diabetic retinopathy complications, a signal seen in some prior diabetic-population GLP-1 trials, were not a concern in the non-diabetic SELECT population, occurring at 0.7% in both arms.

Discontinuation due to adverse events was 16.6% for semaglutide versus 8.2% for placebo across the trial, a meaningful gap worth raising directly with patients considering long-term treatment. Anyone starting a GLP-1 agonist should expect nausea to be most noticeable in the first several weeks of dose escalation and to often ease as the body adjusts, though not everyone tolerates the medication well enough to continue.

Comparing SELECT to LEADER and SUSTAIN-6

Before SELECT, cardiovascular evidence for GLP-1 agonists came from diabetic populations. LEADER (N=9,340) showed liraglutide 1.8 mg reduced MACE by 13% (HR 0.87; 95% CI 0.78, 0.97) in patients with T2D and high CV risk [2]. SUSTAIN-6 (N=3,297) showed semaglutide 0.5/1 mg reduced MACE by 26% (HR 0.74; 95% CI 0.58, 0.95) in T2D patients over 104 weeks, though SUSTAIN-6 was powered for non-inferiority rather than superiority.

SELECT extended this evidence to people without diabetes for the first time. The relative risk reduction (20%) was similar in magnitude to LEADER in a population that had no diabetes to begin with, which suggests the cardiovascular benefit is not primarily explained by improved blood glucose control, since glucose-lowering was not a factor for SELECT participants.

Implications for Prescribing Semaglutide 2.4 mg Today

The 2024 Wegovy label change created a distinct prescribing pathway. A patient with BMI ≥27, prior MI or stroke, and no diabetes now has a labeled cardiovascular indication for semaglutide 2.4 mg, independent of whether weight loss is the clinical goal [3]. Cardiologists, not only obesity medicine specialists or endocrinologists, are appropriate prescribers for this indication.

For patients who have both T2D and established CVD, the decision involves comparing semaglutide 2.4 mg (Wegovy) against semaglutide 1 mg (Ozempic) and against SGLT-2 inhibitors such as empagliflozin or dapagliflozin, which also carry cardiovascular outcome data in diabetic populations. SELECT excluded T2D, so applying its result to that group requires clinical judgment rather than a direct read-across.

Tirzepatide (Zepbound, Mounjaro) does not yet have a completed dedicated cardiovascular outcomes trial. SURMOUNT-1 (N=2,539) showed up to 22.5% mean weight loss with tirzepatide 15 mg at 72 weeks [9]. SURMOUNT-2, in patients with type 2 diabetes, also reported weight loss with tirzepatide, though the exact figure varies by dose and analysis method in the published paper and should be confirmed against the primary source before being cited precisely [10]. Neither SURMOUNT trial used MACE as a primary endpoint, and a cardiovascular outcomes trial for tirzepatide has not yet reported results as of this writing; readers should check for updated status before relying on that point.

Dose escalation for Wegovy follows the labeled 16-week titration: 0.25 mg weekly for weeks 1 to 4, 0.5 mg for weeks 5 to 8, 1.0 mg for weeks 9 to 12, 1.7 mg for weeks 13 to 16, then 2.4 mg from week 17 onward [3]. Patients who cannot tolerate 2.4 mg may remain at 1.7 mg per the label.

Semaglutide is contraindicated in patients with personal or family history of medullary thyroid carcinoma, MEN2 syndrome, or active or recent pancreatitis. Prescribers should consult the current Wegovy prescribing information prior to treatment initiation [3].

What SELECT Did Not Show

The trial enrolled 72% male participants, which limits direct applicability to women, though subgroup analyses showed directionally consistent results across sexes. The population was also 77% white, underrepresenting groups that carry a disproportionate obesity and cardiovascular burden in the United States.

The trial did not show a statistically significant reduction in cardiovascular death on its own, only in the composite three-component MACE endpoint. A broader secondary endpoint that added hospitalization for heart failure and coronary revascularization also favored semaglutide (HR 0.81; 95% CI 0.71, 0.93), but the pre-specified primary endpoint remained the three-component composite.

SELECT did not randomize patients to different doses. Only the 2.4 mg dose was studied for the cardiovascular indication. Whether the 0.5 mg or 1.0 mg semaglutide doses used for glycemic management would produce similar MACE reductions is not answered by this trial.

A Decision Framework: Does This Evidence Change What You Should Do

This is not medical advice for an individual case, but a structured way to sort the SELECT evidence into what is actually decision-relevant before a conversation with a prescriber.

Step 1: Check the four population facts. SELECT enrolled people who were age 45 or older, had a BMI of 27 or above, had a documented cardiovascular event (prior MI, stroke, or symptomatic peripheral artery disease), and had no diagnosis of type 2 diabetes. If all four are true, the trial's population, and the 2024 FDA cardiovascular indication built on it, applies fairly directly. If one or more is false, the evidence below needs adjustment.

Step 2: Know the exceptions that override everything else. A personal or family history of medullary thyroid carcinoma or MEN2 syndrome is an absolute contraindication to any semaglutide product, regardless of cardiovascular history. Active or recent pancreatitis is also disqualifying. Severe kidney disease (eGFR under 15) was excluded from the trial, so the data do not speak to that group.

Step 3: If type 2 diabetes is present, use different evidence. SELECT excluded T2D by design. A person with both diabetes and cardiovascular disease should have the conversation anchored in LEADER, SUSTAIN-6, or SGLT-2 inhibitor cardiovascular trials instead, since those trials, not SELECT, studied that population.

Step 4: Weigh the size of the benefit against the burden of side effects. The absolute MACE reduction was 1.5 percentage points over roughly 3.3 years (NNT approximately 67). Against that, about 1 in 5 people on semaglutide had nausea, and people were roughly twice as likely to stop the drug for side effects as those on placebo (16.6% vs 8.2%). Neither number makes the decision automatically; both belong in the conversation.

Step 5: Separate the indication from the goal. The same drug and dose can be prescribed under a weight-management indication or a cardiovascular risk-reduction indication, and insurance coverage rules can differ between the two even though nothing about the prescription itself changes. Ask explicitly which indication is being used, since that affects both coverage and what outcome the treatment is meant to be judged against.

Step 6: Plan for the titration period, not just the destination dose. The labeled schedule takes 16 weeks to reach the full 2.4 mg dose. Expect GI symptoms to be most noticeable during escalation. A follow-up check around week 16 to confirm tolerance of the full dose is a reasonable default, not a fixed rule.

Step 7: If tirzepatide is also on the table, note the evidence gap. Tirzepatide has strong weight-loss data but does not yet have a completed cardiovascular outcomes trial. The MACE-reduction evidence described on this page belongs to semaglutide 2.4 mg specifically and should not be assumed to transfer to tirzepatide.

A Worked Example

Consider a 58-year-old man with a BMI of 31, a myocardial infarction two years ago, HbA1c of 5.9%, and no diabetes, currently on a statin and aspirin. Running him through the framework above: Step 1 is satisfied on all four points. Step 2 has no red flags assuming no MEN2 or pancreatitis history. Step 3 does not apply since he has no diabetes. Step 4 means his clinician should discuss an absolute risk reduction on the order of roughly 1.5 percentage points over a few years, against a real chance of GI side effects during titration. Step 5 means confirming whether the prescription is being written under the cardiovascular indication, which is what his history supports. Steps 6 and 7 are logistics for the visit.

The SELECT cardiovascular benefit was broadly consistent across subgroups defined by BMI, sex, age, region, and baseline lipid levels, and the treatment groups' outcome curves separated early and stayed separated regardless of how much individual weight patients lost. That is useful context for a patient who is more interested in cardiovascular protection than in the number on the scale, but it does not remove the need to weigh the side-effect burden described above.