What Does a “Longevity Supplement” Actually Have to Prove?
NOVOS Core increased median survival by about 18% in an under-peer-review study of aged male mice under metabolic stress. That does not conclude it extends human lifespan. But it raises a more important question for an industry built around the word “longevity”: how much evidence should that word require?
There may be no word in consumer health having a bigger moment than “longevity.”
Longevity clinics. Longevity diets. Longevity molecules. Longevity powders. Longevity stacks. Longevity supplements.
Some of the products behind that word contain genuinely interesting compounds. Some ingredients have decades of laboratory research. Others affect molecular pathways that biologists believe are important to aging. That science can be valuable. But there is a gap hiding inside the language.
A compound being associated with aging biology is not the same thing as a finished product demonstrating an effect on aging. And twelve ingredients, each with an impressive list of citations, do not automatically become a proven longevity formula when they are combined in a packet.
Studies have found that combinations of longevity molecules can behave differently from their parts and even shorten lifespan. An intervention can make perfect mechanistic sense and fail when tested in a living animal. Something that works in a mouse can fail in a person. Something that changes a molecular biomarker may never meaningfully change how an organ functions.
This is not a reason to dismiss longevity science. It is a reason to demand more of it.
Longevity should be an evidence standard, not just a product category.
The question, then, is not whether a supplement contains ingredients connected to longevity.
The harder question is: what happens when you actually test the intervention?
The word “longevity” has a standards problem
The evidence behind “longevity” products varies enormously
There is a common chain of reasoning in supplements. Ingredient A affects pathway B. Pathway B has been implicated in aging. Therefore, a product containing Ingredient A is marketed for longevity.
Nothing about that argument is necessarily dishonest. But each step answers a different scientific question. A cell-culture study tells us what may happen to cells under specific laboratory conditions. An ingredient trial tells us something about that ingredient. An animal lifespan experiment asks whether an intervention changes survival in a living organism. A randomized human trial asks whether the actual intervention changes a predefined outcome in people.
These are not interchangeable forms of evidence. They are layers.
-
01
Ingredient rationale
A reason to believe a compound matters to aging biology.
-
02
Cellular mechanism
A defined effect on an aging-related pathway in cells.
-
03
Finished formula in vitro
The actual combination tested under laboratory conditions.
-
04
Animal healthspan + lifespan
Survival and function measured in a living organism.
-
05
Human biomarkers / pilots
Early signals in people; exploratory, not definitive.
-
06
Randomized human function
The finished formula versus placebo, on a predefined outcome.
A product does not become scientifically stronger by blurring those layers together. It becomes stronger by climbing them. As of 2026, after 7+ years of R&D, NOVOS has achieved all six layers of evidence, with additional research in progress.
Why lifespan experiments are a high bar
Lifespan is surprisingly difficult to move
Scientists have been trying to identify interventions that extend mammalian lifespan for decades. One of the clearest examples is the National Institute on Aging’s Interventions Testing Program (ITP), which tests candidates using standardized protocols across three sites.
Candidates are selected because researchers already have a scientific reason to think they might work. And yet the current ITP record is remarkably selective. Since the program began in 2002, the NIA reports 15 agents and 2 combinations that have significantly increased median lifespan.
15Agents
2Combinations
Identified by the NIA Interventions Testing Program as significantly increasing median mouse lifespan since 2002.
That distinction is important. We cannot place one NOVOS experiment beside an ITP experiment and call it a head-to-head comparison. What the ITP does show is something more general: even interventions chosen specifically because scientists think they may extend lifespan often do not. Mechanistic promise is not the same thing as a survival result.
The successful list includes interventions such as rapamycin, acarbose, glycine, astaxanthin and several pharmaceuticals or experimental compounds. There are many more negative results in the program’s public record.
A longevity pathway is a hypothesis. Lifespan is an outcome.
Why test mice?
Why mice remain important in longevity research
Humans are the species we care about. So why spend years studying mice? Because the ultimate human longevity experiment is extraordinarily difficult — it would need very large groups of people, extraordinarily long follow-up, and decades of patience. That does not mean human longevity science is impossible. It means researchers build evidence in stages.
A mouse lifespan study can ask a question a six-month human study cannot: did the treated animals actually survive longer?
Cell
Days to weeks
Mouse
Months to years
Human
Years to decades
Translation becomes more relevant as evidence moves toward humans, but direct lifespan experiments become dramatically harder.
The tradeoff is obvious. Mice are not small humans. Their metabolism differs. Their diseases differ. And an intervention that extends mouse lifespan may do nothing of the sort in people. Mouse lifespan is therefore meaningful preclinical evidence. It is not human proof.
The NOVOS Core mouse study
So researchers put the complete formula into an aged-mouse lifespan experiment
The new study was conducted by researchers including scientists at Newcastle University and is currently available as a preprint while undergoing peer review. The intervention was not one isolated Core ingredient. It was the full, exact, patent-pending, twelve-component formulation that is sold to consumers:
The researchers compared three groups of naturally aged male C57BL/6J mice, beginning at 20 months of age: a control group with no active intervention; a nutraceutical group receiving continuous NOVOS Core in food; and a senolytic positive control group receiving two injected courses of Navitoclax plus BAM15.
- What was studied?
- Whether late-life intervention with the complete twelve-ingredient NOVOS Core formulation, or an experimental senolytic regimen, could affect survival and healthspan in naturally aged mice.
- In what model?
- Naturally aged male C57BL/6J mice, interventions beginning at 20 months under a soaked-food protocol associated with caloric excess and metabolic stress.
- What was measured?
- Survival to endpoint, plus repeated frailty index scoring, body weight and short-term memory (Y-maze).
- What did it find?
- Median survival of 703 days in controls versus 830 days with the nutraceutical (about 18% greater; log-rank p = 0.045) and 834 days with the senolytic.
- What does it not prove?
- It does not establish human lifespan extension, human healthspan effects, or equivalence between the supplement and senolytic drugs.
- Conflicts / funding:
- Funded through a UK Medical Research Council grant to Thomas von Zglinicki and a donation from NOVOS Labs to Satomi Miwa. Diogo Barardo is an employee of NOVOS Labs. Funders had no role in study design, data collection, analysis, interpretation, manuscript writing, or the decision to publish.
This was also a late-life intervention. Treatment began at 20 months of age — approximately 60 years old in human terms. The researchers were asking whether intervention beginning after substantial biological aging had already occurred could still matter.
The result
Median survival increased by about 18%
The primary survival result was straightforward. Median survival in the control mice was 703 days. Median survival in mice receiving the multi-ingredient nutraceutical was 830 days: a difference of 127 days, or approximately 18%. The survival curves were significantly different by log-rank analysis, p = 0.045.
Median survival, in days
703
Control, days
830
Core, days
+127
Days difference
~18%
Animal study
Mouse results do not establish human lifespan extension.
There is an obvious temptation to stop the article here. A longevity formula. A lifespan experiment. Eighteen percent longer median survival. Done. But that would leave out the most scientifically important part of the paper.
A model of real-world metabolic stress
The mice were in a more “real world” model of metabolic stress
At 20 months, the mice were switched to soaked food to ensure consistent delivery of the intervention. The soaked-food format substantially increased food consumption, and the mice gained considerable body weight. The authors therefore interpret the study as a model involving caloric excess, obesity and metabolic stress, rather than an optimized aging environment.
That is not a flaw. It is a feature that may make the model more relevant to human aging. Most people do not grow old in pathogen-free laboratories on perfectly controlled diets. They age amid metabolic stressors: dietary caloric excess, inadequate exercise, disrupted sleep, and the accumulated metabolic load of modern life.
In this context, the question is not whether NOVOS Core can extend maximum lifespan in a pristine environment. The question is whether a late-life intervention can preserve survival when aging is already occurring under metabolic stress.
In this experiment, the answer was yes. The intervention increased median survival by roughly 18% under these conditions. But the translational relevance needs to be studied in humans; it is not evidence that Core will add 18% to a human lifespan.
Mouse lifespan is not human lifespan. But it is still a much harder test than a marketing claim.
A second survival analysis
A senolytic-related complication to note for full transparency
The experiment encountered one technical problem. The senolytic protocol involved injections, and a small cluster of deaths occurred shortly after the first treatment course. Postmortem findings suggested some deaths resulted from abdominal infection following accidental puncture of enlarged seminal vesicles, common in aged male C57BL/6J mice.
Because those deaths were considered unrelated to biological aging, the researchers performed a censored sensitivity analysis: median lifespan was 719 days for controls, 860.5 days for the nutraceutical group, and 852 days for the senolytic group — approximately 19.7% greater survival for the nutraceutical and 18.5% for the senolytic, meaning NOVOS Core outperformed the positive control.
However, the primary uncensored analysis remains the more conservative headline: approximately 18% longer median survival for NOVOS Core versus control. The sensitivity analysis is critical because it shows the direction of the result was not created by those procedure-related deaths.
- Primary analysis:
- All deaths included. 703 / 830 / 834 days. ~18% greater survival, p = 0.045.
- Censored analysis:
- Procedure-related deaths removed. 719 / 860.5 / 852 days. ~19.7% and ~18.5% greater survival.
- Why censoring:
- A cluster of deaths followed the first injection course, implicated to abdominal infection rather than biological aging.
- Why both are disclosed:
- The uncensored analysis is the more conservative headline; the censored analysis confirms the direction wasn't an artifact of procedure-related mortality.
The senolytic comparison
The other treatment makes the experiment even more interesting
The second intervention was an experimental senolytic strategy using Navitoclax plus BAM15. Senolytics aim to remove senescent cells — damaged or stressed cells that have stopped dividing but remain metabolically active, releasing inflammatory signals. Cellular senescence is one of the recognized Hallmarks of Aging.
The survival result was strikingly similar between the two active groups: 830 days with the continuous nutraceutical versus 834 days with the two-course senolytic regimen.
The similar survival curves created an interesting mechanistic question: if Core was producing a comparable survival pattern, was it working by killing senescent cells too?
Was NOVOS Core behaving as a senolytic?
The cell experiments suggested something subtler
The investigators tested the complete formulation and its individual components in senescent human fibroblasts, looking for senolytic activity. Fisetin, one of the Core ingredients, is frequently discussed as a potential senolytic. Under the conditions used here, neither the complete formulation nor the tested individual ingredients showed meaningful selective killing of senescent fibroblasts.
So the researchers looked at something different: senostatic activity.
Senolytic versus senostatic
A senolytic attempts to remove senescent cells. A senostatic (or senomorphic) intervention attempts to reduce the harmful behavior associated with those cells without necessarily killing them.
Senolytic
Senescent cell → selective elimination. Remove the cell.
Senostatic
Senescent cell → reduced harmful phenotype. Change what the cell is doing.
When the water-dissolved complete formulation was applied to senescent human fibroblasts, researchers observed reductions in nuclear size, reactive oxygen species (ROS), and IL-6 secretion, without comparable effects in non-senescent cells.
Those are laboratory outcomes; they do not demonstrate a senostatic effect in the human body. But this finding confirms NOVOS Core’s formulation does address senescent cells, offering a possible mechanistic explanation for why the benefit from the continuous nutraceutical accumulated gradually rather than appearing suddenly.
- Senescent cells stop dividing but remain metabolically active, secreting inflammatory signals collectively called the SASP. A senolytic removes them outright; a senostatic dampens their harmful behavior while leaving them alive. In this study, the formulation did not selectively kill senescent fibroblasts but reduced nuclear size, ROS and IL-6 secretion — laboratory findings that do not establish a senostatic effect in humans.
What happened to frailty and memory?
Living longer is more interesting if function is preserved too
Lifespan is easy to understand. Healthspan is harder. The investigators repeatedly measured frailty and cognition. The nutraceutical group showed a gradual attenuation in frailty progression over time; by 30 months, the difference was approaching statistical significance rather than meeting a conventional threshold. That wording matters — this was a promising directional finding, not a definitive statistically significant result.
Short-term memory was assessed using spontaneous alternation in a Y-maze. The nutraceutical-treated animals maintained performance over time while controls declined, though baseline differences and dwindling numbers of surviving control animals limited later comparisons. The senolytic group’s strongest effects appeared shortly after treatment and diminished later; the continuous nutraceutical showed a slower, more gradual trajectory.
What the mouse study does not show
The most important sentence in this article may be this one
NOVOS Core has not been shown to extend human lifespan.
The mouse study cannot answer that question. It also cannot establish that Core adds 18% to human life expectancy, prevents human frailty, prevents cognitive decline, acts as a senostatic in people, or reproduces the senolytic comparison in humans. There are additional limitations:
- Only male mice were studied.
- The model involved a specific inbred mouse strain.
- The animals experienced unusual metabolic stress associated with the feeding protocol.
- The control lifespan was shorter than historical expectations.
- The study experienced procedure-related mortality.
- Some late healthspan comparisons had limited surviving sample sizes.
- The manuscript has not yet completed peer review.
Status: Preprint · manuscript under peer review
These are not details to hide. They are the difference between communicating science and advertising a result.
Why the result still matters
Scientific limitations and scientific significance can coexist
There is a tendency in supplement marketing to present every positive result as definitive, and an equal tendency among skeptics to treat any limitation as a reason to discard the entire experiment. Neither is how science works.
This study asks a legitimate question. A defined twelve-ingredient formulation was administered continuously to naturally aged mice beginning late in life. Median survival was significantly greater than controls, and the direction remained similar after censoring likely procedure-related deaths. Healthspan measures showed potentially favorable trajectories, and cellular experiments provided a plausible mechanistic signal distinct from simple senolytic activity.
That is meaningful preclinical evidence. It simply belongs in the category of animal evidence, not human lifespan proof.
Good science does not become weaker when you state what it cannot prove. It becomes easier to trust.
The formula matters
A formula is not simply the sum of its citations
NOVOS Core contains twelve active ingredients, each selected in part because of evidence connecting it to biological processes implicated in aging — but that was the beginning of the research strategy, not the endpoint. The formulation was designed around the interconnected Hallmarks of Aging described in modern geroscience:
These processes interact. Aging does not occur because one molecule becomes too low — that is the rationale for a systems approach. But a systems rationale is still only a hypothesis until the system is tested.
From cells to animals
The mouse study did not appear out of nowhere
Before the lifespan experiment, NOVOS formulations had already been evaluated in laboratory models of aging biology. In a peer-reviewed study using 2D and 3D human keratinocyte models, researchers examined NOVOS formulations under chemical stress, measuring markers of oxidative stress and DNA damage. NOVOS treatment reduced several laboratory signals associated with DNA strand damage in stressed human skin cells.
These are in vitro findings — they do not tell us that someone who drinks Core experiences the same effect in their skin. But they answer a mechanistic question: can the formulation alter aging-relevant cellular responses under controlled conditions?
Human cells
DNA integrity and stress-response research. In vitro; does not establish effects in the body.
Senescent human fibroblasts
Senostatic signals in laboratory conditions. Not a demonstrated human effect.
Aged mice
Survival and healthspan in a living organism. Animal study; not human proof.
Those experiments do not prove longevity in people. That is precisely why the research program needed to move beyond cells — and why the animal experiment asked a harder question: cells cannot become frail, and a petri dish cannot develop cognition. A living mouse integrates metabolism, circulation, immune function, brain function, muscle, organs, hormones, and the microbiome — all the biological interactions cell culture removes.
NOVOS Core
The formula behind the research program
NOVOS Core is the same twelve-ingredient formulation discussed throughout this article — evaluated across cellular, animal and now human clinical research, including a randomized, double-blind, placebo-controlled human trial measuring cardiovascular aging markers.
References
- Mouse lifespan and healthspan preprint — preprints.org/manuscript/202512.2725
- NIA Interventions Testing Program — nia.nih.gov/research/dab/interventions-testing-program-itp
- Hallmarks of aging — López-OtÃn et al., Cell 2023 — doi.org/10.1016/j.cell.2022.11.001
- University of Bologna skin-cell study — Punzo et al., Nutrients 2024;16(16):2770
NOVOS Core is a dietary supplement and is not intended to diagnose, treat, cure or prevent disease. These statements have not been evaluated by the Food and Drug Administration.