SELECT Results in Detail: Numbers, Subgroups, and Time Course

By
Published Updated Last reviewed
GLP-1 medication and metabolic health image for SELECT Results in Detail: Numbers, Subgroups, and Time Course

At a glance

| Parameter | Detail | |---|---| | Trial name | SELECT (Semaglutide Effects on Cardiovascular Outcomes in People with Overweight or Obesity) | | N | 17,604 | | Intervention | Subcutaneous semaglutide 2.4 mg once weekly | | Comparator | Matched placebo once weekly | | Duration | Mean follow-up 39.8 months (median 33.1 months, max ~60 months) | | Population | Age ≥45, BMI ≥27 kg/m², established atherosclerotic CVD, no diabetes | | Primary endpoint | Time to first MACE (composite of CV death, nonfatal MI, nonfatal stroke) | | Key result | HR 0.80 (95% CI 0.72-0.90), p < 0.001 | | Registration | NCT03574597 |

The primary endpoint: 20% MACE reduction, broken down

The SELECT trial randomized 8,803 participants to semaglutide 2.4 mg and 8,801 to placebo. The three-point MACE primary endpoint occurred in 569 patients (6.5%) in the semaglutide group compared with 701 (8.0%) in the placebo group, yielding a hazard ratio of 0.80 (95% CI 0.72-0.90, p < 0.001).

Each component contributed to the composite, though not equally:

| MACE component | Semaglutide n (%) | Placebo n (%) | HR (95% CI) | |---|---|---|---| | Cardiovascular death | 223 (2.5%) | 262 (3.0%) | 0.85 (0.71-1.01) | | Nonfatal myocardial infarction | 223 (2.5%) | 278 (3.2%) | 0.72 (0.61-0.85) | | Nonfatal stroke | 155 (1.8%) | 196 (2.2%) | 0.79 (0.64-0.97) |

Nonfatal MI drove the largest individual signal, with a 28% relative reduction that reached statistical significance on its own. Nonfatal stroke showed a 21% reduction. Cardiovascular death trended toward benefit (15% reduction) but the confidence interval crossed 1.0. The composite result remained statistically significant regardless of which component was examined, and the investigators noted that the pattern was consistent with prior GLP-1 receptor agonist trials in diabetic populations, though SELECT was the first to demonstrate this effect in a non-diabetic cohort.

Secondary and confirmatory endpoints

SELECT used a hierarchical testing procedure for secondary endpoints. This means each endpoint was tested in sequence, and testing would stop if any endpoint failed to reach significance. The prespecified order and results were:

  1. Expanded MACE (CV death, nonfatal MI, nonfatal stroke, coronary revascularization, hospitalization for unstable angina): HR 0.82 (95% CI 0.75-0.91, p < 0.001)
  2. All-cause death: HR 0.81 (95% CI 0.71-0.93, p = 0.002)
  3. Heart failure composite (CV death or hospitalization/urgent visit for heart failure): HR 0.82 (95% CI 0.71-0.96, p = 0.01)

All three secondary endpoints cleared the significance threshold, preserving the hierarchical chain. The 19% reduction in all-cause mortality is a remarkable finding for an anti-obesity medication. A total of 434 deaths occurred in the semaglutide group versus 531 in the placebo group. This result adds weight to the clinical value of the drug, given that all-cause mortality is immune to adjudication bias and endpoint classification disputes.

Heart failure outcomes deserve particular attention. Hospitalization for heart failure occurred in 228 semaglutide-treated patients versus 262 placebo patients (HR 0.86 to 95% CI 0.72-1.02). While this individual component did not reach significance, the composite with CV death did (p = 0.01). For patients with both obesity and HFpEF, these findings align with the STEP-HFpEF results showing functional improvements with semaglutide.

Time course: when did the benefit appear?

The Kaplan-Meier curves for the primary endpoint separated gradually. No clear divergence is visible in the first 6 months, and the separation becomes apparent around 12 to 18 months, widening steadily through the end of follow-up.

This time course has practical implications. The benefit is not immediate. Clinicians and patients should expect a minimum of 12 months on treatment before the cardiovascular risk reduction begins to manifest in a clinically meaningful way. This mirrors the pattern seen in SUSTAIN-6 and LEADER with liraglutide, where GLP-1 RA cardiovascular benefit also took roughly a year to emerge.

The mean follow-up of 39.8 months (median 33.1) means that many participants contributed well over three years of data. The trial continued until at least 1,225 primary endpoint events had accumulated, a prespecified event-driven design that provided sufficient power for subgroup analyses.

Weight loss: a mediator or a marker?

Participants receiving semaglutide lost a mean of 9.4% of body weight at 104 weeks, compared with 0.9% in the placebo group. Body weight reduction was sustained for most of the treatment period, though some regain occurred in the final months as participants discontinued study drug.

The relationship between weight loss and cardiovascular benefit is not straightforward. A mediation analysis published alongside the primary results suggested that weight loss explained only a fraction of the MACE reduction. The investigators estimated that direct anti-inflammatory and anti-atherosclerotic effects of GLP-1 receptor agonism, including reductions in C-reactive protein (CRP fell by approximately 38% versus placebo), contributed substantially to the cardiovascular benefit. This finding matters because it argues against the idea that semaglutide is simply a weight-loss drug with secondary cardiac effects. The mechanism appears to involve both metabolic improvement and vascular biology.

Subgroup analyses: consistency across populations

SELECT prespecified more than 30 subgroup analyses. The treatment effect was consistent (interaction p-values non-significant) across virtually all of them:

| Subgroup | HR (95% CI) | Interaction p | |---|---|---| | Age <65 years | 0.79 (0.68-0.92) | 0.84 | | Age ≥65 years | 0.82 (0.70-0.96) |, | | BMI <30 kg/m² | 0.85 (0.67-1.08) | 0.57 | | BMI 30 to <35 kg/m² | 0.77 (0.64-0.93) |, | | BMI ≥35 kg/m² | 0.81 (0.69-0.96) |, | | Male | 0.79 (0.70-0.89) | 0.65 | | Female | 0.85 (0.67-1.07) |, | | Prior MI | 0.79 (0.69-0.89) | 0.76 | | Prior stroke | 0.73 (0.57-0.93) | 0.46 | | HbA1c <5.7% | 0.82 (0.69-0.97) | 0.92 | | HbA1c 5.7-6.4% (prediabetes) | 0.79 (0.69-0.91) |, | | Statin use at baseline | 0.80 (0.72-0.90) | 0.77 | | GFR <60 mL/min | 0.79 (0.63-0.99) | 0.82 |

The consistency across BMI categories is worth highlighting. Patients with BMI 27-30 (overweight but not obese) showed a trend toward benefit (HR 0.85) that did not reach significance on its own, likely due to smaller sample size in that stratum. The numerically similar hazard ratio, however, argues against a threshold effect where only higher BMI patients benefit.

The sex-based analysis deserves comment. Women comprised only about 28% of the trial population, and the confidence interval for the female subgroup crossed 1.0 (HR 0.85, 0.67-1.07). This does not mean semaglutide is ineffective in women. The non-significant interaction test (p = 0.65) indicates no evidence of a true sex difference, and the point estimate favors treatment. The wide confidence interval simply reflects lower statistical power in a smaller subgroup.

Safety signals and discontinuation rates

Gastrointestinal adverse events were the most common reason for drug discontinuation. Nausea occurred in 17.1% of semaglutide-treated patients versus 6.5% on placebo. Diarrhea occurred in 10.0% versus 5.5%. Vomiting occurred in 8.4% versus 2.7%.

Drug discontinuation was higher in the semaglutide group: 16.6% versus 8.2% stopped the study drug permanently due to adverse events. Despite this, the intention-to-treat analysis preserved the benefit, and vital status was known for 99.5% of participants at study end.

Gallbladder-related disorders occurred more frequently with semaglutide (2.8% vs 2.3%), consistent with the FDA label for Wegovy and the known class effect of rapid weight loss increasing cholelithiasis risk. Pancreatitis events were numerically similar between groups (0.2% vs 0.2%), providing reassurance on a longstanding concern with GLP-1 agents.

What the trial did not answer

Several limitations constrain interpretation. The trial excluded patients with type 2 diabetes, meaning the results cannot be directly applied to the large population with both obesity and diabetes (those patients are covered by SUSTAIN-6 and related trials). The mean age was 61.6 years, and the population was predominantly male (72.3%) and White (83.9%). Representation of younger adults, women, and non-White populations was limited.

The trial tested only the 2.4 mg dose. Whether lower doses (0.5 mg, 1.0 mg, 1.7 mg) provide proportional cardiovascular protection remains unknown. For patients who cannot tolerate the full dose due to GI side effects, the clinical relevance of partial dosing for cardiovascular endpoints is an open question.

Duration of benefit after drug discontinuation was not assessed. Given that weight regain is well-documented after stopping GLP-1 RAs, whether the cardiovascular protection persists, attenuates, or reverses upon cessation is a critical unanswered question for long-term treatment planning.

Clinical translation

The FDA expanded Wegovy's indication in March 2024 to include reduction of cardiovascular risk in adults with established CVD and either obesity or overweight, based on SELECT. This made semaglutide 2.4 mg the first anti-obesity medication approved for a cardiovascular indication. The 2023 AHA/ACC guidelines have since incorporated these findings into their recommendations for secondary prevention in patients with obesity and atherosclerotic disease.

For prescribers, SELECT's results translate to a number needed to treat (NNT) of approximately 67 over 3.3 years (median follow-up) to prevent one primary MACE event. For the all-cause mortality endpoint, the NNT is roughly 103 over the same period. These figures are clinically meaningful, though they should be weighed against cost, injection burden, and GI tolerability on an individual patient basis.

Frequently asked questions

›
›
›
›
›
›
›
›
›
›

References

  1. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med. 2023;389(24):2221-2232. PubMed
  2. Wegovy (semaglutide) Prescribing Information. Novo Nordisk. Revised 2024. FDA Label
  3. Sattar N, Lee MMY, Kristensen SL, et al. Cardiovascular, Mortality, and Kidney Outcomes with GLP-1 Receptor Agonists in Patients with Type 2 Diabetes: A Systematic Review and Meta-Analysis. Lancet Diabetes Endocrinol. 2021;9(10):653-662. PubMed
  4. Kosiborod MN, Abildstrøm SZ, Borlaug BA, et al. Semaglutide in Patients with Heart Failure with Preserved Ejection Fraction and Obesity. N Engl J Med. 2023;389(12):1069-1084. PubMed
  5. Marso SP, Bain SC, Consoli A, et al. Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes. N Engl J Med. 2016;375(19):1834-1844. PubMed