UTCardiothoracicSurgery.com

Heart conditions, cardiac medicines and supplements: what the evidence says on safety

A GLP-1 Heart Benefit Headline: Read the Absolute Numbers

posted on

By UTCardiothoracicSurgery.com Editorial Team. Sources reviewed October 2, 2026.

Short answer: In the SELECT trial behind the FDA's March 2024 Wegovy (semaglutide) heart-risk approval, a major heart event (cardiovascular death, heart attack or stroke) happened in 6.5% of people who took Wegovy and 8% of people who took placebo. That is a gap of 1.5 percentage points, or about 15 fewer people per 1,000 over the trial. The “20% lower risk” you see in headlines describes the same result as a relative change. Both numbers are true, but they answer different questions.

This guide shows how to turn a risk-reduction headline into real event rates, with a filled example and a blank worksheet you can reuse. It describes what one trial found in one group of people. It does not predict what will happen to you.

Why the same result can sound big or small

A relative risk reduction compares two rates to each other (“one rate is 20% lower than the other”). An absolute risk difference subtracts one rate from the other (“8 in 100 versus 6.5 in 100”). A relative figure can sound large even when few people are affected, and it says nothing about how common the event was to begin with. A 2017 paper by Jager and colleagues makes the point in its title: Relative risk versus absolute risk: one cannot be interpreted without the other.

Filled example: the FDA's published Wegovy numbers

These figures come from the FDA's announcement of Wegovy's new heart-risk indication, with the follow-up window and trial name taken from the other sources listed below.

  • Product: Wegovy (semaglutide) injection, used with a reduced-calorie diet and increased physical activity.
  • Who was studied: Adults with cardiovascular disease and either obesity or overweight. The FDA says the trial randomly assigned over 17,600 participants to Wegovy or placebo. The trial's published design describes participants as aged 45 or older with pre-existing cardiovascular disease and a BMI of 27 or higher, and without diabetes.
  • Comparator: Placebo. Both groups also received standard medical care, such as blood pressure and cholesterol management, and lifestyle counseling.
  • Outcome: Major adverse cardiovascular events, defined by the FDA as cardiovascular death, heart attack and stroke combined.
  • Event rate with placebo: 8% (8 per 100, or 80 per 1,000).
  • Event rate with Wegovy: 6.5% (6.5 per 100, or 65 per 1,000).
  • Follow-up window: The FDA announcement does not state it. The trial's mean follow-up was 39.8 months, or about 3.3 years, according to an American College of Cardiology expert analysis and a published analysis of the trial.

Now the arithmetic:

  • Absolute risk difference: 8.0% minus 6.5% = 1.5 percentage points. In plain terms, about 15 fewer first events per 1,000 people over the follow-up period.
  • Relative risk reduction: 1.5 divided by 8.0 is about 19%. The trial's own analysis, which accounts for when events happened, reported a hazard ratio of 0.80 (95% confidence interval 0.72 to 0.90). That is the “20%” in headlines.
  • Number needed to treat (NNT): 100 divided by 1.5 is about 67. Because the published rates are rounded, treat this as approximate. It means roughly 67 people taking the drug for the trial's follow-up period for each first major event avoided.
  • The flip side: Most people in both groups did not have a first event of this kind during the trial: about 92 of 100 on placebo and about 93.5 of 100 on Wegovy.

Why the follow-up window matters

The same trial data give different numbers depending on how long you look. A published analysis of the trial data, funded by Novo Nordisk (the maker of Wegovy), calculated a number needed to treat of 125 at one year and 58 at four years for the same combined outcome. A headline that leaves out the time window leaves out part of the answer.

What these numbers do not tell you

  • They are group averages. A trial rate describes the whole group. A number needed to treat depends on the baseline risk of the people studied, the time period and the outcome chosen, so it can look very different for people whose risk is higher or lower than the trial average.
  • They apply to the group studied. Do not assume the same numbers apply to people outside that description, such as people without established cardiovascular disease or people with diabetes.
  • They cover one product. These rates belong to Wegovy in this trial. Do not carry them over to other GLP-1 medications.
  • They come from a company-funded trial. SELECT was funded by Novo Nordisk, which also received the FDA approval. That does not make the results wrong, but it is part of how to read them.
  • They are one side of the ledger. Wegovy's prescribing information carries a boxed warning about thyroid C-cell tumors, and the FDA lists other warnings, including pancreatitis, gallbladder problems and kidney injury. The most common side effects include nausea, diarrhea, vomiting and constipation. In the same trial, stomach and bowel complaints led 10.0% of people on semaglutide to stop the drug permanently, compared with 2.0% on placebo.
  • They combine three outcomes. A combined endpoint tells you about the group of events together, not about heart attacks alone or strokes alone.

Blank worksheet: absolute versus relative risk

Use this for any “lowers your risk by X%” headline. If you cannot fill in a line from the source, treat the headline as incomplete.

  1. Exact product and dose: ______________________
  2. Exact outcome (what counted as an event?): ______________________
  3. Who was studied (age, condition, risk level): ______________________
  4. Comparator (placebo, another drug, usual care): ______________________
  5. Follow-up window (how long were people watched?): ______________________
  6. Event rate in the comparator group: ______ %
  7. Event rate in the treatment group: ______ %
  8. Absolute difference (line 6 minus line 7): ______ percentage points, or ______ per 1,000 people
  9. Relative reduction (line 8 divided by line 6): ______ %
  10. Approximate number needed to treat (100 divided by line 8 in percentage points): ______
  11. Harms reported in the same trial (rates, not adjectives): ______________________
  12. Who paid for the study: ______________________
  13. How closely do I resemble the people studied? ______________________
  14. Source and date I checked it: ______________________

Who should talk to a professional, and what to do next

If you have heart disease and are wondering whether this kind of treatment fits your situation, the right conversation is with your cardiologist or the clinician managing your heart care. They can look at your history, medications and risks, which a trial average cannot do. Do not start, stop or change any heart medication based on this article.

Bring the completed worksheet and ask:

  • Do I resemble the people in this trial, or is my situation different?
  • What is my own baseline risk, and how would that change the absolute difference?
  • Which side effects or warnings matter most given my other conditions and medications?

For the cardiac-specific questions that come up around blood pressure, heart rate, heart failure and contraindications, see our GLP-1 safety guide for cardiac patients.

Sources

This article is editorial content for general education. It was not written or reviewed by a physician and is not medical advice, diagnosis or treatment. UTCardiothoracicSurgery.com is an independent publication and is not affiliated with any hospital, medical school or medical practice. Trial results describe groups of people and do not predict any one person's outcome. Prescribing information can change, so check current labeling and talk with your cardiologist or other licensed clinician before making decisions about your care.

Related guides