For more than a decade, the most reliable way to extend an animal’s life has also been the least appealing: eat less, permanently. That trade-off is why longevity researchers have hunted for a calorie restriction mimetic, a pill that reproduces the biology of eating less without the hunger. A study published in Nature in September 2026 suggests that the most prescribed weight-loss drug in the world may already be one.
Researchers at the University of California, Berkeley, gave semaglutide, the active ingredient in Ozempic and Wegovy, to 20-month-old female mice, roughly the mouse equivalent of a 60-year-old human. The treated animals lived about 12 percent longer than untreated controls, held onto muscle and memory, and reversed several molecular hallmarks of aging. In a head-to-head comparison with actual calorie restriction, the drug matched most of the benefits and surpassed it in three.
This article breaks down what the study measured, why the calorie restriction comparison matters, and what the growing Ozempic aging research does and does not mean for people.
The New Finding: Semaglutide Extended Lifespan in Older Mice
The experiment was deliberately late-life. Lead author Danica Chen, a professor of metabolic biology and nutrition at UC Berkeley, and her team obtained 20-month-old female C57BL/6 mice from the National Institute on Aging and injected them daily with either 10 nmol per kg of semaglutide or saline.
For the lifespan arm, 39 control mice and 40 treated mice continued on their regimen until death. The control group reached a median lifespan of 742 days. The semaglutide group reached 834 days, a gain of 92 days, or roughly 12 percent. Measured from the start of treatment, saline mice survived a median of 4.7 more months and treated mice 7.8 more months.
The researchers also tracked what the animals died of. Across several non-tumour categories, age at death was pushed later in the treated group, which points to a broad slowdown in physiological deterioration rather than protection against any one disease.
Notably, the authors reported no adverse effects attributable to the drug at any timepoint, and on-site veterinarians monitored the animals daily throughout.
What Actually Slowed: Muscle, Memory, and the Biology of Aging
A separate cohort was treated for three months and then run through a battery of behavioral and molecular tests. Treated mice outperformed untreated controls across nearly every measure.
Physical function. Treated mice moved more in an open field, spent more time exploring the center of the arena, and entered the open arms of an elevated plus maze more often, all standard readouts of exploratory drive. On the rotarod test of motor coordination, the inverted screen test of muscle strength, and a treadmill exhaustion test, they performed significantly better. Importantly, the muscle results held up after statistical adjustment for body weight, meaning the improvement was not simply a byproduct of the animals being lighter.
Cognition. In the Barnes maze, a test of spatial memory, treated mice spent more time in the target quadrant, searched more holes in that quadrant, and found the escape hole faster.
Metabolic control. Treated mice cleared glucose more effectively and showed improved insulin sensitivity in metabolic tissues.
Molecular hallmarks. The researchers measured six well-established hallmarks of aging and found improvements in all of them:
- Stem cell attrition: bone marrow hematopoietic stem cells showed greater regenerative capacity per cell, with less myeloid bias
- Inflammation: lower expression of inflammatory cytokines and fewer pro-inflammatory macrophages
- Cellular senescence: reduced p16 and p21 expression and lower senescence-associated beta-galactosidase activity
- Genomic instability: fewer cells positive for gamma-H2AX, a marker of DNA damage
- Mitochondrial dysfunction: higher ATP content, lower reactive oxygen species, and increased expression of mitochondria-related genes
- Loss of proteostasis: reduced endoplasmic reticulum stress in the liver and reduced mitochondrial folding stress in stem cells
One finding stood out. In the dentate gyrus of the hippocampus, a brain region central to learning and memory, treated mice showed more newly born neurons. Aging depletes the neural stem cells that drive this process, and that decline is implicated in neurodegeneration. The drug appeared to restore it.
Was It Just Eating Less? The Calorie Restriction Comparison
Semaglutide suppresses appetite. In this study, treated mice ate 24 percent less. That raised the obvious question: is the drug just a complicated way of doing what a diet does?
To answer it, the team ran a separate five-month study with three groups of 20-month-old mice: untreated controls, semaglutide-treated mice, and mice on a 24 percent calorie-restricted diet, matched to the intake reduction caused by the drug. Ten mice per group, assessed at baseline, two months, and four months.
The overlap was substantial. Both interventions preserved locomotor activity, motor coordination, muscle function, and treadmill endurance at similar levels.
The divergence was the interesting part.
| Measure | Calorie Restriction | Semaglutide |
|---|---|---|
| Exploratory drive (open field) | Held near baseline | Improved above baseline |
| Spatial memory (Barnes maze) | Held near baseline | Improved above baseline |
| Glucose tolerance | Held near baseline | Improved above baseline |
| Locomotor activity | Preserved | Preserved |
| Motor coordination (rotarod) | Preserved | Preserved |
| Muscle strength (inverted screen) | Preserved | Preserved |
| Treadmill endurance | Preserved | Preserved |
| Metabolic rate | Slowed (daytime) | Largely unchanged |
| Feeding pattern | Gorging, then prolonged fast | Gradual, appetite suppressed |
| Hunger-driven foraging activity | Present | Absent |
Three measures moved above baseline on semaglutide but merely held steady on calorie restriction: exploratory drive in the open-field test, spatial memory in the Barnes maze, and glucose tolerance.
The two groups also behaved differently around food. Calorie-restricted mice gorged their daily allotment soon after feeding and then entered a prolonged fast, and they showed a spike in locomotor activity late in the light cycle, a foraging behavior driven by hunger. Semaglutide-treated mice spread their intake across the day and showed no such hunger-driven activity.
Metabolically, calorie-restricted mice showed lower daytime energy expenditure, consistent with the well-documented metabolic adaptation to chronic dieting. Semaglutide-treated mice did not. Their metabolic rate stayed largely where it started.
That last difference is the most consequential result in the paper. Calorie restriction extends lifespan, but it does so while the body actively defends itself by burning less. Semaglutide produced many of the same benefits without triggering that defense, which points to a mechanism that is not simply a downstream consequence of eating less. Chen put it directly: the differences suggest GLP-1 drugs may tap into a biological pathway independent of calorie restriction.
The Mechanism: NAD+, Sirtuins, and the IGF-1 Pathway
To explain the biology, the researchers looked at the two best-characterized nutrient-sensing systems in aging research.
The first is the NAD+/sirtuin axis. NAD+ is a coenzyme that cells use to shuttle energy, and its levels fall with age. Sirtuins are a family of seven NAD+-dependent enzymes that act as cellular maintenance crews, coordinating DNA repair, mitochondrial function, and inflammation. Calorie restriction is known to raise NAD+ and induce sirtuin expression.
In the treated mice, NAD+ levels rose in both liver and muscle, and expression of several sirtuins increased, including SIRT7 in the neural stem cells of the dentate gyrus.
The second is the insulin/IGF-1 pathway. Insulin-like growth factor 1 promotes growth in youth and appears to accelerate aspects of aging later in life. Animals with reduced IGF-1 signaling consistently live longer. Treated mice showed lower plasma IGF1.
The drug also induced Oser1, a FOXO-regulated gene that increases lifespan in multiple species, and raised expression of FOXO transcription factors, the downstream effectors of the insulin/IGF-1 pathway. An earlier anti-diabetic combination was shown to reverse biological aging through related nutrient-sensing routes, which is part of why this class of drugs has drawn longevity researchers’ attention.
RNA sequencing of liver tissue tied it together: semaglutide downregulated lipid metabolism and inflammatory responses while upregulating adaptive immune response, insulin response, and proteostasis. Genes that rise with age and were suppressed by the drug were enriched in inflammation and lipid metabolism. Genes that fall with age and were restored were enriched in DNA repair and glucose response.
In short, semaglutide appears to push the same molecular buttons that calorie restriction pushes, which is why researchers describe it as a calorie restriction mimetic. What makes the result notable is that it does so while avoiding the metabolic slowdown that comes with actual dieting.
Are GLP-1 Drugs the First Real Longevity Drugs?
The phrase “first longevity drug” gets applied loosely. Before semaglutide, the strongest candidate was rapamycin, an mTOR inhibitor that reliably extends lifespan in mice and is now in human trials. Metformin, a diabetes drug with a similar repurposing story, is another.
What distinguishes the GLP-1 result is the combination of three things:
- The intervention began late in life, at a mouse age equivalent to a human in their early sixties, when most interventions stop working
- The lifespan extension was clean and statistically significant, not a marginal trend
- The drug is already prescribed to millions of people with a well-characterized safety profile, so the translational path is unusually short
That does not make it a longevity drug yet. A mouse is not a human; the study used only one sex and one inbred strain, and no regulatory body recognizes aging as a treatable indication, which makes the clinical trial path genuinely difficult.
But the mechanistic case is stronger than it was. Earlier GLP-1 studies had shown the drugs delay the onset of individual age-related diseases. This paper offers a possible common explanation: if the drugs slow aging itself, then a benefit that spans heart disease, kidney disease, liver disease, and neurodegeneration is exactly what you would predict.
Rafael de Cabo, a senior investigator at the National Institute on Aging and author of a commentary on the study, framed it that way: most chronic diseases are deeply rooted in the aging process, so if GLP-1 agonists slow it down, a wide range of clinical benefits is what you would expect to see.
Does Ozempic Age Your Face? Separating Two Conversations
Search interest in “ozempic aging” splits into two very different questions, and it is worth being precise about which one this study answers.
The first is the one the study addresses: does the drug slow the biological process of aging?
The second is the one dominating search engines: does Ozempic age your face? The term “Ozempic face” describes the gaunt, hollowed appearance some people develop after rapid weight loss on GLP-1 drugs. It is a real and well-documented cosmetic effect, and dermatologists attribute it to the loss of facial fat volume that comes with fast, substantial weight reduction, particularly in people over 40.
The two are not the same thing and are not in conflict. “Ozempic face” is a volume effect. It is what happens when subcutaneous fat disappears quickly, and it can occur with any method of rapid weight loss, including bariatric surgery and crash dieting. The new study is about something different: the cellular and molecular processes that drive functional decline with age.
The new data does, however, complicate the muscle-loss concern that has trailed these drugs. In the mice, semaglutide reduced body weight, but the loss came predominantly from fat. Lean mass as a percentage of body weight actually increased. That said, the mouse dose and the human experience are not directly comparable, and human studies have reported meaningful muscle loss alongside fat loss on GLP-1 drugs, a concern independent researchers have raised repeatedly.
What This Means for Humans: The Caveats That Matter
The gap between this result and a human prescription is wide, and the paper is explicit about it.
- It is a mouse study. Mice are the standard model for aging research, but they are not humans, and the translation rate for lifespan interventions is poor.
- Only female mice were studied. The authors chose females deliberately to remove the confounding effects of male aggression and injury in group housing. Male data does not exist yet, and independent researchers have flagged this as a priority.
- One strain, one dose. The animals were inbred C57BL/6 mice receiving 10 nmol per kg daily. Human genetic diversity is vastly greater.
- No human lifespan endpoint exists. The closest evidence comes from a recent post-hoc analysis of the SLIM LIVER trial, which looked at liver-related outcomes rather than aging itself.
- Regulatory reality. Aging is not an approved indication, so the trial infrastructure for testing a longevity claim in humans is still being built. A small number of trials are now underway, including one at the University of Texas Medical Branch testing tirzepatide, a dual GLP-1 and GIP agonist. One anti-aging treatment targeting old cells has already cleared a first human trial, which shows the field is moving.
Chen has said future clinical studies may investigate these drugs in healthy older adults, without obesity or diabetes. If benefits appear in that population, the use case for GLP-1 drugs would broaden considerably. UC Berkeley has published a summary of the findings, and the full paper is available through PubMed.
Frequently Asked Questions
Does semaglutide speed up aging?
The evidence currently points the other way. In the September 2026 Nature study, semaglutide slowed multiple hallmarks of aging and extended median lifespan in older female mice. Human data on aging endpoints does not yet exist.
Is semaglutide linked to dementia?
No causal link has been established, and the new mouse data is reassuring rather than worrying. Treated mice showed improved spatial memory and increased neurogenesis in the hippocampus, the brain region most affected in Alzheimer’s disease. Human trials of GLP-1 drugs in neurodegenerative disease are ongoing.
Why do I feel so good on semaglutide?
Clinical reports of improved wellbeing on GLP-1 drugs are common and not fully explained. The new study offers a possible framework: the drugs reduced inflammation and improved metabolic control in mice, both of which influence how people feel. This remains a hypothesis, not a clinical finding.
Is 2 years too long to be on Ozempic?
Long-term use is already common, and trial data now extends several years for approved indications. The aging study does not change the safety picture, and the researchers observed no adverse effects in the treated mice. Any decision about duration should be made with a prescriber, since the risk and benefit profile is individual.
What organ is Ozempic hard on?
The known risk profile centers on the gastrointestinal system, with nausea, vomiting, and constipation being the most common side effects. Rare risks include pancreatitis and gallbladder problems. The drug has shown protective effects on the heart, kidney, and liver in outcome trials.
Does Ozempic cause your face to age?
Not in the biological sense. “Ozempic face” refers to the gauntness that follows rapid facial fat loss, not accelerated cellular aging. It is a volume effect that can occur with any form of rapid weight loss.
Conclusion
The most interesting thing about the Berkeley result is not that a weight-loss drug extended mouse lifespan. It is that it did so without reproducing the metabolic cost of dieting and that it produced functional gains, particularly in memory and glucose control, that calorie restriction did not.
Whether that translates to humans is an open question that will take years and purpose-built trials to answer. But the paper moves GLP-1 drugs out of the category of interventions that treat individual age-related conditions and into the category of interventions that might act on the underlying process.
That is a different claim, and for the first time it is a testable one.
