SELECT: When a Weight Drug Became a Cardiac Drug
Last verified May 2026 · 3 · Finished; headline results out November 2023 · NCT03574597 ↗
Weight-loss drugs have historically been judged on weight, and judged suspiciously, because the field's history includes drugs that reduced weight and harmed hearts. SELECT inverted that. It asked whether semaglutide prevents cardiovascular events in people who already have cardiovascular disease and do not have diabetes — a population where the drug had no approved indication at all — and found that it does. The effect emerged early and persisted across a mean 34.2 months of exposure. Because the trial excluded diabetes, the result cannot be explained as glucose control, and because the benefit exceeded what the weight loss alone would predict, it is not simply a weight effect either.
- Enrollment
- 17,604
- Duration
- Participants took the drug for a mean 34.2 months and were followed a mean 39.8 months, out to 240 weeks
- Drug
- Semaglutide 2.4 mg (Wegovy)
- Population
- Adults 45 and over with a BMI of at least 27 who had already had a cardiac event — a heart attack, a stroke, or symptomatic peripheral artery disease. Diabetes was an exclusion, and so was an HbA1c of 6.5% or above at screening. The people enrolled averaged a BMI near 33 and an age of 61.6, and 27.7% were women, 71.0% white.
Primary endpoint
Cardiac death, a nonfatal heart attack, or a nonfatal stroke — whichever came first
Treatment arm
569 of 8,803 (6.5%)
Comparator
701 of 8,801 (8.0%)
Treatment difference: Hazard ratio 0.80 (95% CI 0.72–0.90); P<0.001 for superiority
A fifth fewer of these events on semaglutide 2.4 mg than on placebo. ★ Translate the relative number into the absolute one before quoting it: the gap is 1.5 percentage points across roughly 40 months, meaning about 67 people have to be treated for that long for one to avoid an event.
Secondary endpoints
| Endpoint | Treatment | Comparator | Difference |
|---|---|---|---|
| Death from a cardiac cause (one of the three components above) Fewer deaths on the drug, but the confidence interval touched 1.0. ⚠ That single result is why several impressive-looking numbers below carry an asterisk: the trial's prespecified testing order stopped here, so everything after it is descriptive rather than proven. | 223 of 8,803 (2.5%) | 262 of 8,801 (3.0%) | Hazard ratio 0.85 (95% CI 0.71–1.01) |
| Heart-failure events combined with cardiac death An 18% reduction that would be a headline in most trials and cannot be claimed as one here, because cardiac death alone did not clear its testing boundary first. | 290 of 8,803 (3.3%) | 339 of 8,801 (3.9%) | Hazard ratio 0.82 (95% CI 0.71–0.96) |
| Death from any cause at all A 19% reduction pointing the same way as the main result, and formally undeclarable for the same reason. ⚠ Anyone citing SELECT as proof these drugs make you live longer is quoting a number the trial itself declined to claim. | 375 of 8,803 (4.3%) | 414 of 8,801 (4.7%) | Hazard ratio 0.81 (95% CI 0.71–0.93) |
| Heart attacks that were not fatal ★ Down 28%, and this is where the headline result actually came from. Of the three things the main endpoint counted, heart attacks moved most. | 234 of 8,803 (2.7%) | 322 of 8,801 (3.7%) | Hazard ratio 0.72 (95% CI 0.61–0.85) |
| Strokes that were not fatal Essentially unchanged. The benefit was in heart attacks and cardiac deaths, not strokes — worth knowing if stroke is the risk that concerns you. | 138 of 8,803 (1.6%) | 151 of 8,801 (1.7%) | Hazard ratio 0.93 (95% CI 0.74–1.17) |
| Kidney decline: protein in the urine, half of kidney function gone, kidney failure, dialysis, or death from kidney causes A 22% reduction in the kidney measure, in a trial that was not looking at kidneys. It is consistent with the dedicated kidney trial that later won Ozempic a kidney indication. | 1.8% | 2.2% | Hazard ratio 0.78 (95% CI 0.63–0.96) |
| Weight change after two years Weight kept coming off for about 65 weeks and then mostly stayed off. ⚠ Note the size against the weight trials: this is roughly 9%, not the 15% figure that gets quoted from STEP-1. | −9.4% | −0.9% | Estimated treatment difference −8.5 percentage points (95% CI −8.8 to −8.2); P<0.001 |
| Weight change after four years ★ The longest weight follow-up that exists for this drug: about 10% below starting weight at four years, holding across every BMI band, both sexes, and every region enrolled. It is the best answer available to whether the weight comes back while you keep taking it. | −10.2% | −1.5% | Difference −8.7 percentage points; P<0.0001 |
| Waist measurement after four years Belly fat stayed down across four years, which matters more for cardiac risk than the number on the scale does. | −7.7 cm | −1.3 cm | P<0.0001 |
Adverse events
| Event | Treatment rate | Comparator rate |
|---|---|---|
| Serious adverse events of any kind ★ Fewer on the drug than on placebo — mostly because there were fewer cardiac events and fewer infections. That is an unusual direction for a drug safety table and worth registering. | 33.4% | 36.4% |
| Stopped the drug for good because of side effects Twice as many quit on semaglutide, and stomach and bowel problems were the reason. ⚠ Read this next to the row above: the drug caused fewer serious events and drove far more people to give up on it. | 16.6% (1,461) | 8.2% (718) |
| Stomach and bowel problems Nausea, diarrhea, vomiting and constipation — the same complaints this whole class produces, and the single biggest reason people stop. | 10.0% leading to discontinuation | 2.0% leading to discontinuation |
| Gallbladder problems Gallstones and gallbladder inflammation ran modestly higher. Losing weight quickly does this regardless of how the weight comes off, so it is not specific to the drug. | 2.8% | 2.3% |
| Pancreatitis, confirmed by independent review No increase. This is the risk patients ask about most, and the largest trial in the class found nothing. | 0.2% | 0.3% |
| Cancers of any kind Identical in both arms across roughly 40 months. ⚠ Forty months is not a cancer-latency window, so read this as reassuring rather than as settled. | 5.9% | 5.9% |
| Sudden kidney injury Lower on the drug, despite the fluid losses that vomiting and diarrhea cause — consistent with the kidney benefit further up the table. | 1.7% | 2.2% |
| Serious psychiatric events ★ No signal for suicidal thinking or self-harm in the independently reviewed events. This is the largest randomized dataset bearing on the 2023 scare. | 1.4% | 1.5% |
Subgroup analyses
- Had already had a heart attack (about 70% of those enrolled): Cardiac events reduced by about a fifth, matching the trial as a whole
The benefit showed up whether or not someone had had a heart attack before, with no sign the two groups responded differently.
- Women (27.7%) against men (72.3%): Similar MACE hazard ratios across sexes
Similar benefit in both. ⚠ Note the imbalance though: obesity is at least as common in women, and barely a quarter of this trial was female.
- Starting BMI: 27 to under 30, 30 to under 35, 35 to under 40, and 40 or above: MACE benefit consistent across all BMI strata
★ The benefit did not require severe obesity — people merely in the overweight range got the same relative reduction. That undercuts the common coverage rule that ties access to a BMI cutoff.
- Blood sugar at entry: normal against prediabetic: MACE hazard ratios similar across glycemic strata
Same benefit either way, which is direct evidence the effect is not simply better blood sugar.
- Under 65 against 65 and over: Consistent MACE reduction across age groups
No difference by age.
- North America, Europe, Asia and elsewhere: Directionally consistent MACE benefit across regions
Consistent across 41 countries, so the result is not an artifact of one health system's standard of care.
Clinical significance
SELECT is the strongest argument in this class for treating obesity as a cardiovascular condition, and it is the reason coverage arguments increasingly turn on event reduction rather than on body weight. Practically, it is also the clearest asymmetry between semaglutide and tirzepatide: one has a completed outcomes trial in this population and the other does not until 2027. ⚠ It says nothing about compounded preparations, and nothing about people without established cardiovascular disease — the population was selected for prior events, and extending a 20% relative reduction to a healthier population is an assumption rather than a finding.
Who sells semaglutide, and for how much
This trial tested Semaglutide 2.4 mg (Wegovy). Across the 556 sellers on this register, 283 publish a standing monthly cash price for compounded semaglutide, from $79 a month, with a median of $174.
⛔ None of those sellers was in this trial. A compounded preparation is not the product studied here: no agency has reviewed that specific preparation for safety, effectiveness or manufacturing quality, and its concentration is set by whichever pharmacy filled it. The result above is the strongest evidence available for semaglutide and it was generated on something else. Every seller, with its price record.
Frequently Asked Questions
References
- 1.Lincoff AM, Brown-Frandsen K, Colhoun HM, et al.; SELECT Trial Investigators Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes N Engl J Med. 2023. PMID: 37952131.
- 2.Ryan DH, Lingvay I, Deanfield J, et al. Long-term weight loss effects of semaglutide in obesity without diabetes in the SELECT trial Nat Med. 2024. PMID: 38740993.
- 3.U.S. Food and Drug Administration FDA approves first treatment to reduce risk of serious cardiovascular problems specifically in adults with obesity or overweight (Wegovy) FDA Press Announcement. 2024. https://www.fda.gov/news-events/press-announcements/fda-approves-first-treatment-reduce-risk-serious-cardiovascular-problems-specifically-adults-obesity
- 4.ClinicalTrials.gov Semaglutide Effects on Cardiovascular Outcomes in People With Overweight or Obesity (SELECT) ClinicalTrials.gov registry. 2024. https://clinicaltrials.gov/study/NCT03574597