Can a GLP-1 medicine help someone live longer? There is credible evidence of fewer serious cardiovascular events in a defined high-risk population, and an encouraging mortality signal. There is no established number of extra years that these medicines add to a person’s life. Those statements can all be true together.
A review published on November 10 in Diabetology International brings the question back into focus. Jens Juul Holst discusses whether GLP-1 therapies might become longevity drugs, drawing partly on cardiovascular trials and survival findings after bariatric surgery. It is a synthesis and a hypothesis, not a new trial measuring lifespan. The useful question is how far the original evidence lets that hypothesis travel.
Start with the people SELECT actually studied
The original SELECT trial randomized 17,604 adults aged 45 or older who had established cardiovascular disease, a body mass index of at least 27, and no history of diabetes. It compared once-weekly injectable semaglutide, targeting 2.4 mg, with placebo alongside standard care. That is a research regimen, not a recommendation to start or change treatment.
Over a mean follow-up of 39.8 months, the primary composite of cardiovascular death, nonfatal heart attack or nonfatal stroke occurred in 6.5% of the semaglutide group and 8.0% of the placebo group. The hazard ratio was 0.80, with a 95% confidence interval of 0.72–0.90. This is important cardiovascular evidence. Because the endpoint includes nonfatal events, it is not itself a measurement of longer life.
The current US Wegovy prescribing information includes the corresponding cardiovascular-risk indication. The semaglutide profile explains why a molecule, a particular product and an approved use need to stay connected.
Fewer deaths were recorded, with an important statistical limit
A 2024 mortality analysis of SELECT examined the same participants, rather than providing a second independent trial. Across roughly 3.3 years, 833 participants died. The current label reports 375 deaths among 8,803 semaglutide recipients and 458 among 8,801 placebo recipients: 4.3% versus 5.2%, using rounded percentages.
The all-cause mortality hazard ratio was 0.81, with a 95% confidence interval of 0.71–0.93. But reading that interval alone misses a crucial qualification. In Section 14.1, Table 6 of the label, all-cause death did not achieve statistical significance under the trial’s prespecified testing hierarchy.
That hierarchy sets an order for confirming secondary findings while limiting false-positive conclusions across multiple tests. Superiority for cardiovascular death, the first confirmatory secondary endpoint, was not established. The later all-cause result therefore cannot be presented as confirmatory proof of a survival benefit. The observed difference remains informative; its formal evidential status also matters. Our guide to reading trials explains why primary outcomes and additional analyses answer different questions.
A lower hazard is not a percentage increase in lifespan
A hazard ratio of 0.81 describes an estimated 19% lower hazard of death during the observed follow-up. It does not mean participants lived 19% longer. Nor does the approximately 0.9-percentage-point difference between the rounded death proportions reveal how many years anyone gained.
A lifespan estimate asks a different question: how long people survive across a much longer horizon. Extending a trial’s curves into the future requires assumptions about later treatment, changing risks and what happens after observation ends. SELECT’s registered outcomes and posted results describe events during the trial, not lifetime follow-up. A calculator that turns these findings into a personal number of extra years would need evidence the trial did not supply. The relative-versus-absolute-risk explainer offers another way to keep a striking percentage in context.
Why the longevity hypothesis still has limits
The mortality analysis also found that the lower noncardiovascular death rate was predominantly associated with fewer infectious deaths during the COVID-19 pandemic. That context matters when considering whether the same pattern will recur in other periods. It does not establish semaglutide as an infection-prevention treatment.
The new review’s comparison with bariatric surgery is a reason to investigate longevity, not a way to borrow surgery’s estimated years of survival for medication users. Different interventions, populations and follow-up cannot be exchanged simply because both produce weight loss. Likewise, SELECT cannot establish the same mortality effect for every GLP-1-based product, younger adults without cardiovascular disease, or people seeking treatment solely to slow aging.
What this changes in a treatment conversation
The practical question is which benefit has been demonstrated for someone with a similar clinical history, using the relevant product. A proven reduction in serious cardiovascular events can matter greatly without an anti-aging promise. Ask which outcome treatment is intended to improve, how benefit will be reviewed, and what safety or tolerability issues need monitoring.
Adverse events led to permanent trial-product discontinuation in 16.6% of semaglutide recipients versus 8.2% with placebo in the original report. Benefits and treatment burden belong in the same discussion. Our withdrawal-evidence guide explains why longer-term care also needs a plan for interruptions.
The November review raises a worthwhile research question. The answer available today is narrower: cardiovascular benefit is established in the studied setting, mortality findings need their statistical qualifications, and a promise of extra years of life goes beyond the evidence inspected here.


