What the study did
Most aging studies in mice begin early in the animals’ lives. That is scientifically convenient and practically worthless, because no human starts taking a drug against aging at the age of twenty.
A research group at the University of California, Berkeley therefore took a different approach. They began treatment in 20-month-old female mice, roughly equivalent to a 60-year-old human. The animals received semaglutide or saline daily. [1]
The dose was chosen to correspond to a clinically common exposure. Mice break down semaglutide considerably faster than humans and therefore need more frequent doses. [1] Converted to human doses, the amount used falls roughly between the dose for diabetes treatment and the dose for weight loss.
The study consisted of three parts: a lifespan experiment with 40 treated and 39 untreated animals, smaller groups for functional and tissue tests after three months, and a comparison with animals that were simply given less food. [1]
The result making the headlines
The treated mice lived an average of 834 days, compared with 742 days for the untreated mice. That is about twelve percent longer. [1]
A second point is also noteworthy. The diseases from which the animals ultimately died were distributed similarly in both groups. Semaglutide therefore did not simply prevent one specific cause of death, such as tumors, but delayed several forms of age-related decline. [1] That is exactly what one would expect from a drug that genuinely targets aging rather than a single disease.
The functional tests added to this finding. The treated animals performed better in spatial memory, explored unfamiliar environments for longer, showed better motor coordination and endurance, and processed sugar more effectively. [1]
Why the control group is a problem
The control animals lived an average of 742 days. That is not outside the published range, but it is shorter than what is often reported for this mouse strain. In a large comparative study by the Jackson Laboratory, female animals of the same strain lived an average of 866 days. [2]
At 834 days, the treated animals were therefore roughly where one would have expected them to be anyway, while the control animals fell below that level. This raises an uncomfortable alternative explanation: Semaglutide may not have slowed aging, but instead offset an unidentified factor that shortened survival in this cohort.
That does not invalidate the study. The experiment was randomized and properly controlled, and comparisons between studies are unreliable because diet, housing, and environment affect lifespan. However, it does mean that a replication in an independent cohort with normally long-lived control animals is still missing, ideally at another institute.
There is a second limitation as well: The study included only female animals. That is understandable for an initial investigation, but whether the same applies to male animals remains unknown.
Was it just eating less?
The treated mice ate about one-quarter less than the control animals and consequently lost weight, mostly fat. [1] This raises an obvious explanation: Calorie restriction reliably extends life in mice, and semaglutide could simply be a pharmaceutical version of it.
The researchers tested this by establishing a third group whose food intake was reduced by the same 24 percent. Body weight and fat mass developed similarly in the semaglutide group and the reduced-ration group, while their eating behavior was completely different. [1]
The semaglutide animals ate throughout the day, only in smaller amounts. The animals on the reduced ration devoured their portion shortly after feeding and then fasted until they were fed again. In other words, one group received smaller portions spread throughout the day, while the other received one large portion per day.
The two groups developed similarly in terms of movement and endurance. The semaglutide group performed better in spatial memory, exploratory behavior, and sugar processing. [1] The most honest interpretation of the findings is that a substantial part of the effect probably resulted from lower calorie intake. In addition, semaglutide appears to create a distinct physiological state that differs from classic fasting.
What happened in the tissues
Lifespan and functional tests show what happened, but they do not explain why. The research group therefore also examined tissue.
In the hippocampus, the region responsible for learning and memory, semaglutide-treated animals had more cells with characteristics of newly developing nerve cells. In addition, inflammatory messenger substances and markers of senescent, no longer dividing cells were less pronounced. [1]
Both findings fit the overall picture and make the results biologically more coherent. [3] These markers accompany aging, but they are not aging itself, and the study cannot determine whether they caused the longer lifespan or were merely a consequence of weight loss.
What this means for humans
For now, nothing. No study has yet examined whether GLP-1 medications affect human life expectancy, and such a study would take decades.
What we do have in humans are data on individual disease events. In a large study of overweight people with pre-existing cardiovascular disease but no diabetes, severe cardiovascular events occurred less often with semaglutide than with placebo. [4] This is an important finding, but it is not the same as slowing aging.
Another question is more practically relevant. Anyone who takes a weight-loss injection loses weight, and some of that weight is muscle.
In the mouse study, most of the weight loss came from fat, and the calf muscle was actually larger relative to body weight than in the control animals. [1] Nothing can be inferred from this for humans, and the measures for preserving muscle remain the same: resistance training and sufficient protein.
State of the evidence
This table distinguishes what the study shows from what is being inferred from it.
| Area | Level of evidence |
|---|---|
| Longer lifespan in female mice | Demonstrated (RCT, n=79) |
| Better function after three months | Preliminary evidence (small groups) |
| Effect independent of reduced food intake | Partial (inconsistent) |
| Effect in male animals | Not studied |
| Independent replication | Still pending |
| Human life expectancy | Not studied |
| Cardiovascular events in humans | Demonstrated (large RCT) |
This is the first study of its kind and was methodologically more rigorous than many earlier studies. It makes the possibility that GLP-1 medications do more than reduce weight a serious subject for discussion.
