Longevity: what are the biggest challenges now?
In our longevity market deck, you will find everything you need to understand the market
SUMMARY
Longevity’s biggest challenge now is proving that treatments which change aging biology actually give humans more healthy years, rather than simply changing biomarkers or extending lifespan in animals.
Human lifespan has continued to rise, but healthspan has not kept pace. Across 183 countries, the average gap between life expectancy and healthy life expectancy widened from 8.5 to 9.6 years, which means longer lives are still bringing substantial periods of disease and disability.
Extending lifespan is also getting harder in the world’s longest-lived populations. Many of the easy twentieth-century gains came from preventing early deaths, while future progress increasingly depends on reducing mortality at older ages, when several diseases and biological aging processes arrive together.
Aging biology itself is clearly manipulable. The harder problem is reproducibility: lifespan benefits in mice can disappear when researchers change the dose, treatment age, sex or experimental conditions, even in unusually rigorous multi-laboratory programs.
Rapamycin remains one of the strongest drug candidates because its mouse lifespan effects have repeated more reliably than most. Human evidence is still far thinner, with small trials offering useful safety and functional clues without showing that rapamycin prevents dementia, cancer, frailty or death.
Senolytics and partial cellular reprogramming show how quickly longevity science is moving into humans while also showing how early the field remains. Senolytics have produced biological effects with inconsistent clinical outcomes, while reprogramming has only just entered a small, localized Phase 1 test.
Aging clocks may eventually make longevity trials much faster, but today they create a second validation problem. Researchers can measure treatment-related changes within months, yet nobody can reliably convert a younger epigenetic or proteomic age into a known number of extra healthy years.
The growing evidence for organ-specific aging makes the problem even more interesting. People of the same chronological age can apparently have very different biological ages across the brain, arteries, kidneys and other systems, which may eventually make longevity medicine far more personalized than the idea of one universal anti-aging drug suggests.
Trial design, regulation and safety are tightly connected. Hard healthspan outcomes take years to appear, aging itself lacks a simple drug-development pathway, and preventive treatments given to relatively healthy people may need a cleaner long-term safety profile than medicines used for serious established disease.
For now, GLP-1 drugs, exercise and ordinary cardiovascular prevention are doing more to create proven healthy years at scale than dedicated longevity biotech. The field changes decisively when a treatment can delay several major age-related diseases at once, when biomarker changes reliably predict those benefits, and when the treatment remains safe enough to use for years.
This market map, featured in our longevity market deck, highlights top companies and startups in the longevity market
Are people actually living longer in good health?
Human longevity still has a basic problem: lifespan is growing faster than healthspan.
For this article, that is the definition that matters. A longevity treatment becomes genuinely useful when it gives people more years without major disease, disability or serious loss of function.
The clearest global measurement comes from a Mayo Clinic analysis published in JAMA Network Open. Researchers compared life expectancy and health-adjusted life expectancy across 183 WHO member states over two decades. Life expectancy increased by 6.5 years, while healthy life expectancy increased by 5.4. The average gap between the two widened from 8.5 to 9.6 years, a 13% increase.
Some long-lived countries had even larger gaps. The United States reached 12.4 years, Australia 12.1, New Zealand 11.8 and the United Kingdom 11.3. Women globally spent about 2.4 more years than men in that gap.
Country comparisons need some care because a shorter lifespan can mechanically produce fewer years lived with chronic disease. Still, the broad pattern is hard to escape. Modern medicine has become very good at keeping people alive after disease appears, while preventing decades of multimorbidity, frailty and disability has proved much harder.
If you want more recent data on this point, please see our latest longevity market report.
Is human lifespan becoming harder to extend?
Human lifespan is still increasing in many places, but adding each extra year has become much harder in the world’s longest-lived populations.
A Nature Aging analysis looked at mortality from 1990 to 2019 in Australia, France, Italy, Japan, South Korea, Spain, Sweden, Switzerland, Hong Kong and the United States. The pace of life-expectancy improvement slowed across the group. Outside Hong Kong and South Korea, annual gains in the most recent decade studied fell below 0.2 years.
That slowdown makes sense when we look at where twentieth-century longevity gains came from. Public health, vaccines, sanitation, antibiotics, safer childbirth, lower smoking rates and better cardiovascular treatment removed huge numbers of deaths at young and middle ages. Once most people survive those risks, further gains depend increasingly on lowering death rates at old ages, where several diseases and biological aging arrive together.
The Nature Aging researchers estimated that repeating the roughly 30% improvement in mortality after age 65 seen between 1990 and 2019 would add only around another 2.5 years to average life expectancy.
Survival to 100 also remains unusual. Using mortality patterns in the populations studied, the researchers estimated an average probability of reaching 100 of around 5.1% for females and 1.8% for males. Hong Kong was the strongest outlier, at 12.8% and 4.4% respectively.
Those figures do not create a permanent biological ceiling. A treatment that genuinely slows aging could still bend the mortality curve. Current mortality trends simply give us little reason to expect widespread centenarian survival from ordinary medical progress alone.
As this slide shows, and as featured in our longevity market deck, online search interest in longevity has been steadily increasing
Do scientists actually understand aging well enough to target it?
Aging biology is now understood well enough to manipulate experimentally, although we still cannot predict which manipulation will produce a meaningful human healthspan benefit.
The updated Hallmarks of Aging framework describes 12 interconnected processes associated with aging, including genomic instability, epigenetic changes, impaired autophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem-cell exhaustion, chronic inflammation and dysbiosis.
Those mechanisms can be altered experimentally. The US National Institute on Aging’s Interventions Testing Program is particularly useful because it tests candidate drugs in genetically diverse mice at three independent laboratories under standardized conditions.
The program currently lists 15 individual agents and two drug combinations that have significantly increased median mouse lifespan under at least one tested condition. Rapamycin, acarbose, canagliflozin, glycine, 17α-estradiol, epicatechin and halofuginone are among them.
The same program also keeps revealing how fragile some effects are. A recent ITP experiment retested 11 interventions, including different doses or starting ages for previously promising compounds. None significantly extended lifespan in either male or female mice. Astaxanthin, mitoglitazone and meclizine had produced positive results under earlier conditions and lost the effect when the protocol changed.
So mammalian lifespan clearly responds to pharmacology. Predicting which intervention will keep working across different biological conditions remains much harder.
Has any longevity drug actually worked in humans?
No drug has yet proved that it broadly slows human aging and gives healthy people several extra years free from major age-related disease.
Human trials have produced interesting fragments of evidence. Rapamycin has shown some encouraging safety and functional results. Senolytics can reduce markers of cellular senescence and may help selected patients. Metformin has decades of epidemiological and disease-treatment data. GLP-1 drugs reduce major cardiovascular and metabolic outcomes. Partial cellular reprogramming has now reached its first patients.
None of those results yet demonstrates a generalized slowing of human aging.
The scale difference tells the story. The NIA has accumulated decades of controlled lifespan work in mice. By comparison, the longest randomized rapamycin study in healthy aging followed 114 completers for 48 weeks. A prominent randomized dasatinib-plus-quercetin senolytic study involved 60 women. The first partial-reprogramming trial plans to enroll only a small number of patients because it is primarily testing safety.
Metformin was supposed to produce a much larger test. The proposed TAME study was designed around more than 3,000 people aged 65 to 79, followed across multiple sites to see whether metformin could delay a group of major age-related diseases. AFAR’s current public page still says funding is needed to launch the study.
| Approach | Best human evidence so far | What we can really say |
|---|---|---|
| Rapamycin | PEARL followed 114 completers for 48 weeks; the primary visceral-fat endpoint did not improve, although some secondary outcomes did | Interesting enough to justify larger trials |
| Senolytics | Small human studies show biological activity; a 60-person randomized bone study missed its primary endpoint | The mechanism works biologically; broad healthspan benefit remains unclear |
| Metformin | Huge real-world disease data, but the proposed 3,000+ person TAME longevity trial has yet to deliver results | Strong candidate, weak direct geroscience proof |
| Partial reprogramming | ER-100 has entered a first-in-human Phase 1 study in optic neuropathy | Human testing has started; efficacy against aging remains unknown |
| GLP-1 drugs | Large randomized trials show fewer major cardiovascular and metabolic events in high-risk people | Strong health-outcome evidence in defined populations, without proof of generalized aging slowdown |
If you want more recent data on this point, please see our latest longevity market report.
This chart, featured in our longevity market deck, illustrates yearly VC funding for longevity startups
Why do longevity drugs look better in mice — and why does rapamycin still stand out?
Longevity drugs often lose strength when experimental conditions change, while rapamycin stands out because its mouse evidence has repeated more reliably than most.
The NIA’s mouse program makes the broader problem unusually visible. Its mice are genetically heterogeneous, experiments run independently at three laboratories, and both male and female animals are tested. Even under those controlled conditions, many lifespan effects differ sharply between sexes. Some drugs help males much more than females. Changing the dose can erase a benefit. Starting treatment later can do the same.
A recent 11-intervention ITP study sharpened that point. None of the compounds significantly extended lifespan under the new protocols. Three had previously looked successful under other conditions.
Human variation is much larger. A 70-year-old woman with diabetes, sarcopenia and high inflammation may respond very differently from a healthy 55-year-old man. Genetics, microbiome, diet, other medications, kidney function, disease history and treatment duration all add more uncertainty.
Rapamycin still stands out because its lifespan effects have appeared across multiple ITP experiments, and combinations with metformin or acarbose have also produced positive results. Analyses of age-specific mortality suggest rapamycin belongs to the smaller group of interventions whose effect can persist into very old age in mice.
Human evidence currently sits several levels below that.
The PEARL trial followed generally healthy adults receiving placebo, 5 mg or 10 mg of compounded rapamycin weekly for 48 weeks. Of 125 people who entered the study, 114 completed it.
The primary endpoint was visceral fat. Rapamycin did essentially nothing there: the statistical result was p=0.942. Researchers did see improvements in lean tissue mass and self-reported pain among women receiving the higher dose, along with some quality-of-life changes in another treatment group. Moderate and serious adverse events were broadly similar across groups.
Researchers also discovered during the study that the compounded rapamycin formulation produced roughly one-third of the blood concentration seen with commercial formulations after 24 hours, making the nominal dose harder to interpret.
PEARL gives us a useful one-year safety signal and some reasons to keep testing rapamycin. It tells us very little about whether long-term treatment reduces dementia, cancer, cardiovascular disease, frailty or death.
Are senolytics actually working in people?
Senolytics are biologically active in humans, but the clinical benefits so far are small, inconsistent and probably highly dependent on who receives them.
The basic idea remains attractive. Senescent cells stop dividing and can release inflammatory molecules that disrupt surrounding tissue. Their accumulation is one of the established hallmarks of aging, and removing them improves several age-related outcomes in animals.
Researchers have already shown that dasatinib plus quercetin can reduce senescence markers in people. In a tiny early study involving nine patients with diabetic kidney disease, three days of treatment reduced several measures of senescent-cell burden in adipose tissue and skin.
The harder tests have looked less impressive.
A randomized Phase 2 study gave intermittent dasatinib plus quercetin to 60 postmenopausal women. The primary bone-resorption endpoint showed no significant difference after 20 weeks. A bone-formation marker improved by 16% early in treatment, then the difference disappeared by week 20.
The most interesting result appeared inside a subgroup. Women starting with the highest senescent-cell burden showed stronger changes in bone formation, bone resorption and radius bone density. Those analyses were exploratory, so they need confirmation.
Several fisetin studies are still underway, including small Phase 2 programs looking at frailty, inflammation and mobility in older adults.
The key question now is whether senolytics mainly work in people carrying a high enough senescent-cell burden.
This chart, featured in our longevity market deck, looks at Function Health’s strategy in longevity
Has cellular reprogramming finally reached humans?
Partial cellular reprogramming has now reached human patients, which makes it one of the biggest genuine changes in longevity research lately.
Life Biosciences received FDA authorization for ER-100 and announced in June 2026 that the first participant had been dosed.
ER-100 uses controlled expression of three reprogramming factors — OCT4, SOX2 and KLF4 — with the goal of shifting old or damaged cells toward a more youthful pattern of gene expression. The Phase 1 study targets people with open-angle glaucoma or non-arteritic anterior ischemic optic neuropathy.
Researchers are beginning with the eye, a localized organ where treatment can be targeted and visual function can be measured. The study is primarily about safety and tolerability, with additional measures looking for early signs of visual benefit.
Partial reprogramming is powerful enough to create unusual risks. Push cells too far and they can lose their identity. Full reprogramming can create pluripotent cells and tumors. Researchers also worry about genomic instability and the possibility of activating harmful growth programs.
The engineering problem is finding a safe middle ground: enough reprogramming to restore function without pushing cells toward uncontrolled growth or loss of identity.
ER-100 will give us the first meaningful human evidence on that question. Whole-body rejuvenation remains much further away.
If you want more recent data on this point, please see our latest longevity market report.
Can biological age clocks really tell us whether a longevity treatment works?
Biological age clocks are getting better at detecting treatment effects, but we still cannot translate a younger score into extra healthy years.
A recent Nature Medicine analysis called TranslAGE combined 51 human intervention studies and recalculated 16 major epigenetic clocks plus 94 additional DNA-methylation biomarkers using the same pipeline. That gave researchers 3,128 samples and a much cleaner comparison than reading dozens of isolated studies.
Some clocks clearly performed better. Newer models trained around mortality or pace of aging, including DunedinPACE and PCGrimAge, responded more consistently than older clocks designed mainly to predict chronological age.
Population health also mattered. People with disease generally showed larger biomarker responses than healthy participants.
Then came an unusually fresh example from Insilico Medicine’s rentosertib trial.
Researchers reanalyzed proteomic data from 42 people with idiopathic pulmonary fibrosis who had participated in the drug’s Phase 2a study. They ran six independently developed proteomic aging clocks across 2,841 measured proteins. All six showed shifts toward younger predicted biological age in treated patients.
Across each treatment regimen there were 54 treatment-versus-placebo clock comparisons. In the 30 mg twice-daily group, nine were statistically significant after correction. Week four produced the strongest pattern, with 11 of 18 comparisons significant. Some organ-specific mortality clocks also moved sharply.
There is another complication: one biological age number may itself be too crude.
Large organ-aging studies now show that different systems inside the same person can age at very different rates. A 65-year-old might have unusually old arteries and kidneys with a relatively young brain, while another person of the same chronological age shows the opposite pattern. Recent work using blood proteins, imaging and DNA methylation increasingly points toward organ-specific aging maps rather than a single whole-body score.
That could also help explain why treatments behave differently across patients. Rapamycin affects nutrient sensing and immune function. Senolytics target senescent cells. GLP-1 drugs strongly affect metabolic and cardiovascular risk. Reprogramming acts much closer to cell identity and gene expression.
The main problem remains clinical validation. Rentosertib was tested for 12 weeks in people with a serious lung disease. A proteomic clock moving by three, four or six predicted years cannot yet be converted into three, four or six extra healthy years.
| Longevity endpoint | Why researchers like it | What is still missing |
|---|---|---|
| Mortality | Completely clear and hard to manipulate statistically | Trials become huge and slow |
| Major diseases | Clinically meaningful and easier to observe than death | Different diseases appear at different rates |
| Physical and cognitive function | Directly tells us whether people are aging well | Measures need careful standardization |
| Epigenetic clocks | Can change within months | We cannot yet convert a clock change into a healthspan gain |
| Proteomic / organ clocks | Can capture several systems and organs separately | Recent results such as rentosertib still need long-term clinical validation |
This chart, featured in our longevity market deck, illustrates yearly funding for longevity startups
Why are longevity trials so slow and expensive?
Longevity trials are slow because the outcomes we trust take years to appear, while the outcomes we can measure quickly still lack enough clinical validation.
Suppose researchers want to know whether a drug keeps healthy 60-year-olds alive longer. Mortality rates are low enough at that age that the study needs a huge population or very long follow-up.
Move the endpoint earlier to heart attacks, cancer, dementia, frailty and other age-related events and the study becomes more manageable. It also becomes statistically complicated because those outcomes differ in severity, timing and frequency.
Move even earlier to an aging clock and a trial might produce an answer within months. Regulators and clinicians then need evidence that the clock predicts something people actually care about.
A recent Nature Aging paper proposed hierarchical trial endpoints that can combine outcomes such as death, disease and functional decline while giving more weight to the most important events.
TAME shows how difficult the practical version can become. The planned metformin study proposes more than 3,000 participants aged 65 to 79 across multiple centers, followed for years for a composite of major age-related diseases. The design has existed for years, yet AFAR’s current project page still lists funding as the missing piece before launch.
Regulation adds another layer. Aging itself still lacks the kind of standard drug indication that cancer, diabetes or hypertension have. Companies therefore usually enter the clinic through specific diseases.
Life Biosciences is testing epigenetic reprogramming in optic neuropathy. Senolytic developers study osteoporosis, kidney disease, frailty or pulmonary disease. Insilico developed rentosertib for pulmonary fibrosis and then added aging-clock analysis around the disease trial.
That disease-first route may be the most practical one for now: prove a clinical benefit in an accepted indication, then use aging biomarkers to see whether something broader is happening.
How safe would a real longevity drug need to be?
A longevity drug for healthy people would need an unusually clean safety profile because even a small side effect can become serious after years of preventive use.
The comparison with cancer medicine makes the problem obvious. A person with aggressive cancer may reasonably accept substantial toxicity for a chance of survival. A healthy 50-year-old considering a drug that might reduce disease risk 20 years later has a very different risk tolerance.
Aspirin offers a useful historical warning.
The ASPREE trial randomized more than 19,000 generally healthy older adults to low-dose aspirin or placebo. After a median follow-up of 4.7 years, aspirin failed to improve the trial’s main outcome of disability-free survival. Major hemorrhage occurred at 8.6 events per 1,000 person-years with aspirin and 6.2 with placebo, around a 38% relative increase.
A fairly small harm was enough to wreck the preventive trade-off.
Rapamycin now faces the same logic over a much longer hypothetical treatment period. PEARL found broadly similar serious and moderate adverse events over 48 weeks, which is useful. A person might eventually take longevity-dose rapamycin for 20 years. We have almost no randomized information across that timescale.
Cellular reprogramming raises the stakes again because the mechanism reaches deep into cell identity and gene expression. ER-100’s first human trial is consequently small, localized and followed carefully.
Starting treatment earlier could theoretically prevent more accumulated damage, but it also means decades more exposure. That trade-off will be central to any preventive longevity drug.
If you want more recent data on this point, please see our latest longevity market report.
This chart, featured in our longevity market deck, compares the main business model options for longevity clinics
Are GLP-1 drugs already doing more for longevity than longevity drugs?
GLP-1 drugs currently have stronger evidence for preventing serious age-related health problems than any drug developed specifically around human longevity.
SELECT randomized 17,604 people with overweight or obesity and established cardiovascular disease, without diabetes, to semaglutide 2.4 mg or placebo. Major cardiovascular events occurred in 6.5% of the semaglutide group and 8.0% of the placebo group, a 20% relative reduction.
Other large GLP-1 programs have since expanded the evidence across obesity, diabetes, kidney disease, heart failure and other conditions closely tied to unhealthy aging.
The population still matters. SELECT participants already had cardiovascular disease and overweight or obesity. Semaglutide has never shown that a metabolically healthy 40-year-old can take it and slow whole-body aging.
Still, the comparison tells us what a serious longevity claim eventually has to beat.
A drug that changes an epigenetic clock by one year may be biologically interesting. A drug that prevents heart attacks, kidney failure, diabetes and loss of mobility is already creating healthier years, whatever label we attach to the mechanism.
GLP-1 drugs may therefore be showing the more realistic route into longevity medicine: first produce hard benefits in a defined disease or risk group, then see whether those benefits extend across several age-related outcomes.
Are exercise and basic prevention still beating longevity biotech at scale?
Exercise and conventional prevention still beat today’s longevity biotech on both human evidence and global reach.
A 2026 systematic review searched nearly 2,000 publications for randomized trials that measured multidimensional human healthspan through outcomes such as intrinsic capacity and quality of life. Only 15 articles, covering 4,656 participants, met the criteria.
Seven studies tested exercise alone. Six more tested multidomain programs that all contained an exercise component. Eleven exercise or exercise-containing studies reported improvements in intrinsic capacity or quality of life.
The researchers could not reach a solid conclusion for caloric restriction or supplementation because the evidence was too sparse and heterogeneous.
Then we have ordinary disease prevention.
WHO’s latest global hypertension assessment estimates that 1.4 billion adults aged 30 to 79 have high blood pressure. About 600 million do not know they have it. Only roughly 320 million — 23% — have it under control.
WHO estimates that achieving effective treatment in half of affected people could prevent about 76 million deaths by 2050.
That gap also shows why scale belongs inside the longevity debate. WHO expects the number of people aged 60 or older to reach 2.1 billion by 2050, and around two-thirds of them will live in low- and middle-income countries.
A cheap oral medicine could scale quickly if it worked. Metformin would be attractive partly for that reason. Rapamycin is also an existing small molecule, although broad preventive use would require much stronger long-term evidence.
Gene therapies and cellular reprogramming start from a far harder cost and delivery base. A future system that depends on proteomic clocks, whole-body imaging and organ-specific treatment could also struggle in places where basic cardiovascular screening is still incomplete.
For now, exercise, blood-pressure control, lipid management, vaccination and smoking prevention remain far ahead on proven population-level healthspan.
This chart, featured in our longevity market deck, illustrates how revenue is distributed across customer segments in the longevity market
So what are the biggest challenges in longevity now?
The biggest challenge in longevity today is proving that our growing ability to manipulate aging biology actually gives humans more healthy years.
The field has made real progress. We can alter mammalian lifespan pharmacologically. Researchers can measure aging through DNA methylation, proteins, imaging and individual organs. Senescent cells can be targeted in humans. Partial cellular reprogramming has reached its first patient. A new rentosertib analysis has even shown that six independently developed proteomic clocks can move in the same direction during a clinical drug trial.
The missing layer is the one people ultimately care about.
We still cannot say that lowering an epigenetic age score by one year gives someone one more healthy year. Rapamycin has far stronger mouse evidence than human outcome evidence. Senolytics show biological activity with inconsistent clinical effects. Reprogramming has just entered Phase 1. TAME, one of the clearest attempts to run a proper multidisease geroscience trial, still has not produced the definitive experiment the field has talked about for years.
Trial design therefore sits near the center of the problem. The outcome we trust most — longer healthy survival — arrives painfully slowly. The measurements we can collect quickly still need validation.
Safety follows immediately behind it. Longevity treatments may eventually be given to people decades before severe disease appears, leaving very little room for chronic harm.
Then comes heterogeneity. Drug responses differ between male and female mice. Senolytic benefits may concentrate in people with high senescent-cell burden. Biological age can differ sharply from one organ to another inside the same person.
Finally, even a scientifically successful longevity treatment has to survive the real world. As seen above, WHO’s hypertension numbers show how much healthy life is currently lost despite having cheap, proven therapies available.
Longevity research today is scientifically credible, clinically early and increasingly testable.
The breakthroughs that would change the field are now fairly concrete: a validated surrogate endpoint, a trial showing that one treatment delays several human diseases at once, reliable identification of responders, or long-term safety data on a therapy people could realistically take for years.
| Biggest challenge now | What is holding longevity back | What would genuinely change the field |
|---|---|---|
| Proving human healthspan gains | Human trials remain tiny compared with the claims being made | A large randomized trial delaying several age-related diseases or functional decline |
| Validating aging biomarkers | Clocks respond to treatment, but their clinical meaning is still unclear | Showing that a given biomarker change reliably predicts later healthspan |
| Long-term safety | Preventive treatment could last decades | Multi-year evidence showing a clear benefit-risk advantage |
| Finding the right patients | Aging and drug response vary by sex, organ, disease and biological state | Biomarkers that reliably match people to the intervention most likely to help them |
| Building workable trials and regulation | Hard outcomes take years and aging lacks a simple development pathway | Accepted multidisease endpoints and clinically validated surrogates |
| Making longevity scalable | Many advanced approaches remain difficult or expensive to deliver | Treatments that normal healthcare systems can afford, prescribe and monitor |
If you want more recent data on this point, please see our latest longevity market report.
OUR METHODOLOGY
This analysis asks whether human longevity science can already extend healthy human life and, just as importantly, what is still stopping promising biology from turning into meaningful healthspan gains. We separated the problem into population longevity trends, aging biology, animal reproducibility, human translation, clinical outcomes, biomarkers, trial design, safety and real-world scalability.
We did not treat every form of longevity evidence as equivalent. Hard human outcomes carried more weight than changes in aging clocks, randomized trials carried more weight than observational associations, and replicated multi-site animal results carried more weight than isolated experiments. Exploratory subgroup findings were also treated more cautiously than prespecified primary endpoints.
Recent evidence was prioritized when it materially changed the state of the field, while older landmark trials were kept when they provided a stronger benchmark. That is why very recent work on partial cellular reprogramming, proteomic aging clocks and geromedicine trial design sits alongside older evidence such as ASPREE and SELECT.
For biological age clocks, we treated a treatment-related shift as evidence that a measurable biological feature had changed. We did not assume that a one-year change in an epigenetic or proteomic age score equals one additional healthy year, because that clinical conversion has not yet been established.
We also looked for reproducibility across sex, dose, treatment timing, population and measurement method. The National Institute on Aging’s Interventions Testing Program was especially useful here because candidate interventions are tested in genetically diverse mice across three independent laboratories, making failed replications informative rather than something to hide.
Population-level evidence came from large international analyses of lifespan, healthspan and preventable disease. Clinical evidence came primarily from randomized trials, peer-reviewed human studies and official trial records. Company disclosures were used only for very recent operational developments, such as the first participant being dosed in the ER-100 study, rather than as proof of efficacy.
Key sources include the updated Hallmarks of Aging framework in Cell, the GeroScience report on the Interventions Testing Program’s retesting of longevity interventions, the PEARL randomized rapamycin trial, the EBioMedicine study of dasatinib plus quercetin and human senescent-cell burden, and the Nature Medicine randomized senolytic trial in postmenopausal women.
For aging biomarkers and newer clinical approaches, we used the Nature Biotechnology analysis of rentosertib across six proteomic aging clocks, the Nature study of organ-specific aging through the plasma proteome, and the Nature Aging work on hierarchical endpoints for geromedicine trials.
For the safety and real-world healthspan benchmarks, we used the ASPREE disability-free-survival trial, the companion ASPREE analysis of major bleeding, the SELECT cardiovascular-outcomes trial of semaglutide, the systematic review of randomized interventions measuring multidimensional human healthspan, and World Health Organization data on hypertension and global population aging.
The final conclusions come from convergence across those different layers of evidence. A finding becomes much more persuasive when biological activity, repeated animal results, human biomarkers and clinically meaningful outcomes all point in the same direction. Longevity science currently has pieces of that chain, but very few interventions have connected all of them.
This chart, featured in our longevity market deck, shows how longevity plan technology has evolved over time
Related blog posts
- Longevity: what are the biggest unsolved problems?
- The most recent funding news in the longevity market
- What are the top startups in the longevity market?
- The startups that have raised the most funding in the longevity market
