Somewhere in the North Atlantic, probably alive right now, is a Greenland shark that was born before the French Revolution.
The oldest Greenland shark ever dated was put at roughly 400 years, give or take a century: a juvenile when Beethoven was born, already old by the time Napoleon rose and fell. Its kin are still out there, hunting in cold water. They are built from cells, DNA and mitochondria, the same biological toolkit that produces human aging, cancer, arthritis and cognitive decline. Yet whatever aging the shark does is stretched so thin that it looks nothing like what we mean when we use the word.
That makes the shark a problem for the usual account of aging. If aging were a physical constant, entropy made flesh and wear accumulating toward inevitable failure, the shark would be impossible. Physics doesn’t make exceptions for cold water. But the shark exists, and has existed for centuries, using the same molecular machinery that kills most mammals in under a decade. So what, exactly, is the claim about inevitability?
The evolutionary logic of dying
The intuitive model of aging is a physics model. Machines wear out; organisms, being complex machines, eventually deteriorate past the point of function. Entropy wins. This model has the appeal of inevitability, and evolutionary biology shows that inevitability to be wrong.
The correct frame is evolutionary. George Williams laid it out in 1957: antagonistic pleiotropy. Genes that are advantageous early in life, when selection pressure is high because you haven’t yet reproduced, can carry costs that manifest late in life, when selection pressure has weakened because most of your genes have already been passed on. Evolution does not program death. As your contribution to the next generations winds down, selection’s grip on you loosens, and with little force selecting against the damage at 75, aging accumulates.
The theory generates a testable prediction. In its classic form, it says that species whose adults face heavy predation and constant threat should age faster. Evolution under intense predation cannot afford to invest in long-term maintenance because most individuals die before reaching the age where that investment would pay off. Species whose adults escape those dangers should age slower. More individuals survive to old age; selection can therefore act on late-life function; maintenance becomes worth paying for.
Bats are a clean test of this prediction. By body size, a bat should live maybe five or six years; that’s what the mammalian size-longevity relationship predicts. Instead, measured by maximum recorded lifespan, bats on average outlive similar-sized ground mammals more than threefold, and in five species individuals have been recorded surviving more than thirty years in the wild. Lower predation risk, evolutionary theory predicts, should produce more investment in somatic maintenance and slower aging. Across bat species, the longest-lived turn out to be the ones that hibernate, shelter in caves and breed slowly, the habits that keep a bat out of harm’s way. Flight is the original escape route, and the rest of the lifestyle widens it. A wider comparison across mammals confirms that bats live at least three times longer than non-flying mammals of their size, and that neither a slow metabolism nor the ability to enter torpor explains it. Marsupials undercut the wear-and-tear model from the other side: despite running slower metabolisms than most mammals, they live only about four-fifths as long as placental mammals of their size.
Thomas Kirkwood’s disposable soma theory, proposed in 1977, adds the metabolic side of the same argument. Organisms allocate a finite energy budget. Continuous, high-fidelity repair for indefinite survival is metabolically expensive. Under constant predation pressure, where most individuals die before benefiting from heavy investment in long-term maintenance, evolution systematically under-invests in repair: cheap somatic maintenance, fast reproduction, early death. This explains the size-longevity correlation. Large animals such as elephants, whales and tortoises face less predation and have more time to amortize maintenance costs, so evolution has built them to last. Mice, under intense predation, are biologically disposable. Their cells are not inferior; they are optimized for a different trade-off.
These two theories, antagonistic pleiotropy and disposable soma, describe the same trade-off from different angles: one at the level of individual genes, one at the level of whole-organism energy budgets. Together they frame aging as a biological parameter, calibrated by selection in response to ecological pressure, rather than a physical inevitability.
The Greenland shark, in this frame, is the disposable soma theory extended toward its logical extreme: a cold-water predator with few natural enemies, an extremely slow metabolism, and centuries of accumulated evolutionary investment in maintenance. The theory predicts exactly this kind of animal, and the shark is built from standard biology running a different program.
The natural experiments
Evolution has run the longevity experiment many times, and the naked mole-rat, the Greenland shark and a small jellyfish are among the results that stray furthest from the wear-and-tear model.
Naked mole-rats should live around six years. Body size predicts it; every rodent metabolic rule predicts it. They live twenty-five to thirty-plus years and show near-negligible cancer rates and reproductive capacity well into old age, all while carrying more oxidative damage in their tissues than mice do, from youth onward. More strikingly, their mortality rate doesn’t increase with age.
Most animals, including you, follow the Gompertz pattern, first described mathematically in 1825: mortality risk increases exponentially with time. In humans, your probability of dying roughly doubles every eight years after early adulthood. This is the mathematical signature of aging, with deterioration compounding on deterioration and failure rates accelerating. Naked mole-rats violate this entirely: their risk of death holds flat regardless of measured age. The researchers who tracked them described the naked mole-rat as “a non-aging mammal,” which puts it in a different category from species that merely live long.
Naked mole-rats do, however, age epigenetically. Naked mole-rat methylomes, the genome-wide patterns of chemical marks on their DNA, drift with time, and an epigenetic clock can predict their age with reasonable accuracy. The demographic non-aging and the epigenetic aging coexist, which separates two things often assumed to be the same. Epigenetic drift happens in naked mole-rats. It does not, apparently, translate into the accelerating mortality that defines the Gompertz curve in other species. The information-loss model of aging (discussed later) and the demographic evidence for naked mole-rat non-aging pull against each other, and the conflict hasn’t been fully resolved.
The Greenland shark’s credentials are more straightforward: it is the longest-lived vertebrate known. In 2016 Julius Nielsen and his colleagues dated twenty-eight female sharks by radiocarbon in the core of the eye lens, tissue that is metabolically inert after it forms and so preserves the isotope composition present at birth. Their estimate: the species lives at least 272 years, and the largest individual, at just over five meters long, was 392 ± 120 years old. Sexual maturity arrives around 150 years. They are slow in every biological sense, growing roughly a centimeter a year.
In 2026, a team in Japan sequenced the Greenland shark genome and compared it against shorter-lived sharks. They found expanded gene families associated with DNA repair, immune function and NF-κB signaling, the inflammatory pathway that also features prominently in mammalian aging research. The genome points to no alien biochemistry at all, only extra copies and distinctive variants of repair and maintenance genes that other species also carry, candidates for how the shark stretches its aging so thin.
Turritopsis dohrnii, the “immortal jellyfish,” is the most extreme case and requires the most precise language. This small hydrozoan can revert from its adult medusa form back to the juvenile polyp form through transdifferentiation: differentiated cells converting directly to a different cell type, resetting developmental age. It can, in principle, repeat this indefinitely. In practice, most individuals die to predation or disease first, so this is not immortality in the popular sense. What the jellyfish does show is that differentiated cell identity is not irreversibly fixed. Developmental state can be reset biologically, and the cellular program rewound. For the epigenetic reprogramming work now underway in laboratories, that makes the jellyfish direct evidence of principle.
The Gompertz law and the organisms that break it
Naked mole-rats violate the Gompertz law. Their mortality rate doesn't increase with age at any measured point in their lifespan, which is why the researchers who analyzed them described naked mole-rats as defying Gompertz's law entirely.Naked mole-rats show that the exponential acceleration of deterioration Gompertz described is not built into living matter, and that a different biology can flatten it. Physics doesn’t produce exceptions like this, so the curve has to be a biological setting.
What happens inside
Four mechanisms come up in every serious account of aging, and they turn out to be connected.
Start with telomeres, the protective nucleoprotein caps at the ends of chromosomes. Every time a cell divides, they get slightly shorter. When they shorten past a critical threshold, the cell reads this as DNA damage and activates one of two responses: apoptosis, in which the cell kills itself, or replicative senescence, in which it stops dividing but doesn’t die. Elizabeth Blackburn, Carol Greider and Jack Szostak won the 2009 Nobel Prize in Physiology or Medicine for identifying this system: the structure of telomeres, the enzyme telomerase that rebuilds them in certain cell types, and the mechanism linking shortening to the senescence program. Telomere erosion works as a countdown, and also as the trigger for a cascade.
The cascade is the SASP, the senescence-associated secretory phenotype. Senescent cells don’t die and they don’t rest quietly. They secrete a cocktail of pro-inflammatory cytokines, including IL-1α, IL-6 and IL-8, alongside matrix-degrading proteases and growth factors that disrupt the tissue architecture around them. At the level of the organism, the result is chronic, low-grade inflammation, “inflammaging,” that degrades tissue function, impairs resident stem cells and creates a feedforward loop in which SASP signals can induce senescence in previously healthy neighboring cells. A small number of senescent cells can corrupt an entire tissue environment out of proportion to their numbers. This is why senolytics, drugs that target senescent cells, have a mechanistic rationale that doesn’t require any speculation about pleiotropic effects elsewhere: they remove the source of the inflammatory signal.
The mitochondrial story is more complicated than the textbooks once suggested. Denham Harman proposed in 1956 what became the “free radical theory of aging”: the reactive by-products of using oxygen slowly damage the cell. In 1972 he pinned the problem on mitochondria, and the idea hardened into a vicious cycle. Mitochondria generate reactive oxygen species; these damage mitochondrial DNA; damaged mitochondria generate more reactive oxygen species; the cycle compounds over time. It dominated aging research for decades. Recent work puts this in sharper perspective. Mitochondrial DNA mutations, as the Northwestern biologists Scott Budinger and Navdeep Chandel read the evidence, do not set a causal limit on physiological aging: the mutation rate is far lower than what would be needed to impair energy production. In their view, mitochondria matter primarily as signaling organelles, regulating stress responses, metabolic state and apoptosis, rather than as progressively destroyed power plants. Mitochondrial dysfunction is real in aged cells. Its causal relationship to aging is less direct than Harman assumed.
The fourth, epigenetic drift, is where the most ambitious current thinking lives. As cells age, the patterns of methylation on their DNA, and the patterns of histone modification that compact or expose different regions of the genome, progressively drift from youthful baselines.
The epigenetic program that tells each cell which genes to express and which to silence degrades. It is the program that makes a liver cell different from a neuron different from a retinal ganglion cell, though all three contain identical DNA. Steve Horvath developed epigenetic clocks that can measure this drift with striking precision. His first, built in 2013, reads methylation at 353 CpG sites and estimates age across fifty-one different tissue and cell types. The clocks work across species too, so what they measure is a general feature of the aging process.
In December 2023 David Sinclair, Yuancheng Lu and Xiao Tian laid out what they call the information theory of aging. Its core proposition: aging is primarily the loss of epigenetic information. The DNA sequence, the digital data of the genome, stays largely intact. What erodes is the epigenetic program: the analog overlay of methylation patterns and histone states that tells the genome how to behave in each specific cell context.
The other mechanisms fit within this frame. Telomere erosion causes localized epigenetic changes at the site of DNA damage. SASP cytokines alter neighboring cell epigenomes via intercellular signaling. Mitochondrial dysfunction changes the metabolic state of cells, which in turn alters the enzymatic activity of the proteins that write and read epigenetic marks. Rather than a catalog of parallel failures, the information theory proposes a single architecture: aging as the progressive corruption of a program that was working.
The practical implication is that the program might be recoverable. In 2020 Lu, Sinclair and colleagues used gene therapy to deliver three Yamanaka factors, OCT4, SOX2 and KLF4 (collectively OSK), to retinal ganglion cells in aged mice. Youthful DNA methylation patterns and transcriptomes were restored; axon regeneration after injury improved; measurable vision loss reversed in a mouse model of glaucoma and in aged mice. It was not a longevity treatment. It tested whether the epigenetic information needed for youthful cell function is still present in old cells or has been destroyed, and found that it is still there, though inaccessible.
Matt Kaeberlein, a longtime University of Washington aging biologist and co-founder of the healthspan company Optispan, and others have argued that aging is irreducibly multi-factorial, with no single mechanism deserving to be treated as primary. Epigenetic reprogramming’s safety at whole-organism scale also remains unestablished in humans: the same signals that reset cellular age in a retinal neuron might have very different consequences in a tissue with high turnover and cancer risk. The information theory is the most generative and testable current unifying hypothesis in aging biology, but the science has not converged on it.
Sinclair, resveratrol, and the problem of commercial science
David Sinclair is both the scientist behind the information theory and the most commercially prominent figure in longevity science, and the two identities are not easily separated. In 2003 Sinclair and colleagues claimed that resveratrol activates sir2 and extends yeast lifespan. The finding spawned Sirtris Pharmaceuticals, co-founded by Sinclair and Andrew Perlman, which GlaxoSmithKline acquired in 2008 for $720 million. GSK tested SRT501, a high-bioavailability resveratrol formulation, in patients with multiple myeloma. What followed was no clear benefit in human trials, a cancer trial halted after kidney failures in patients taking a high-dose formulation, and a core sirtuin-activation effect that turned out to depend on a fluorescent tag used in the original assay. GlaxoSmithKline halted all further SRT501 development, citing inadequate efficacy and unacceptable tolerability, and shuttered Sirtris in 2013 without an approved drug.
Kaeberlein and others have publicly argued that Sinclair regularly overclaims, moving in his public communication from peer-reviewed finding to commercial product faster than the evidence justifies. This assessment is relevant to how readers should weight the information theory, which is serious science and should be judged on its merits, distinct from the commercial history surrounding the scientist who articulated it.The evidence for intervention
The list of longevity interventions with real evidence behind them is shorter than the industry suggests, but it is not empty. Caloric restriction is the most reproducible, across species from yeast to primates.
Its mechanism, mTOR pathway inhibition and AMPK activation, slows processes that contribute to epigenetic drift and SASP accumulation. In rhesus monkeys, two long-running parallel studies produced partially divergent results. The University of Wisconsin study showed reduced age-related disease and improved survival in calorie-restricted animals. The NIA study showed health improvements but no statistically significant survival benefit, a discrepancy attributed partly to differences in diet composition and how control animals were fed. Both studies confirmed that caloric restriction does something. What, exactly, and whether it extends lifespan in long-lived primates, remains unsettled.
In humans, the CALERIE trial produced two distinct findings that get conflated in popular coverage. Volunteers were asked to cut their calories by a quarter for two years; in practice the non-obese adults managed about twelve percent, and even that improved multiple cardiometabolic biomarkers: blood pressure, plasma lipids, C-reactive protein, glucose homeostasis. A separate analysis of samples from the same participants found that caloric restriction increased thymic mass and the production of recent thymic emigrant cells (naïve T cells that the aging thymus produces in ever smaller numbers), and pointed to reduced expression of the gene PLA2G7 as a possible mediator. Caloric restriction produces real effects. The full quarter cut the trial asked for is also not a practical intervention for most people, which is where the pharmaceutical work becomes relevant.
Rapamycin is the best-evidenced pharmacological longevity intervention in mammals. It inhibits mTOR complex 1, mimicking caloric restriction’s cellular response at the molecular level. In 2009 the NIA Interventions Testing Program fed rapamycin to genetically heterogeneous mice starting at 600 days of age, roughly equivalent to a 60-year-old human, and it extended average lifespan by 9% in males and 13% in females. Starting late in life, in a species already aged, a drug extended lifespan. For years the evidence stayed in animals, with nothing translatable to humans.
Pooled results across eight vertebrate species confirm that rapamycin produces lifespan extension comparable to dietary restriction, and that metformin, heavily promoted in longevity circles, does not. Rapamycin is in a different evidence class from most of what the longevity industry sells.
Earlier placebo-controlled trials in older adults lasted weeks or a few months; the PEARL trial ran for almost a year. Run by the telehealth company AgelessRx, it gave healthy adults 5 mg or 10 mg of compounded rapamycin a week, or a placebo, for 48 weeks; 114 of them saw it through. The compounded form reaches only about a third of the blood levels of standard sirolimus, so the real doses were lower than they sound. The primary endpoint, a drop in visceral fat, was not met. Adverse events were similar across groups, and low-dose intermittent rapamycin was well-tolerated. Women on the higher dose gained lean tissue and reported less pain; the lower-dose group reported better general health and emotional well-being. None of this is evidence that rapamycin extends human life; that question is open, and answering it properly will take a decade and a very large trial.
The critical caveat: rapamycin is an immunosuppressant at therapeutic doses, used in organ transplantation. Impaired wound healing, hyperlipidemia, and infection susceptibility are documented. The PEARL doses are far below transplant protocol levels, and intermittent dosing looks meaningfully safer — but the long-term risk profile in healthy people is unknown. People currently taking rapamycin for longevity are using a drug with real science behind it, in what amounts to an uncontrolled experiment.
Senolytics round out the credible portfolio. The mechanistic logic is clear: eliminate senescent cells, remove the SASP signal, reduce inflammaging. The first-generation combination, dasatinib (a cancer drug that inhibits the BCR-ABL tyrosine kinase) plus quercetin, was developed in animal models by Kirkland and colleagues at Mayo Clinic. The first study to show that senolytics clear senescent cells in people, a small Mayo Clinic pilot in 2019, gave three days of oral dasatinib plus quercetin to nine patients with diabetic kidney disease. Within eleven days, multiple markers of senescent cell burden in adipose tissue measurably decreased: p16INK4A and p21CIP1 expression, senescence-associated β-galactosidase activity, circulating SASP factors. With nine patients and only surrogate endpoints, the pilot produced no long-term outcome data. The first placebo-controlled sign of real clinical benefit came in 2026, in a small trial in people with fibrotic fatty-liver disease (MASH): after three cycles of dasatinib plus quercetin, liver scarring improved in 47% of treated patients against 7% on placebo. Whether clearing senescent cells slows aging itself has yet to be shown.
The evidence hierarchy: caloric restriction, rapamycin, and senolytics have the best-supported mechanistic rationale and the best animal-to-early-human trajectory in the field. None of them is ready to be sold as a longevity product. That distinction matters because the industry hasn’t been waiting for the evidence.
The industry that got there first
The longevity industry built its business model, without manufacturing fraudulent science, out of the gap between preliminary results and clinical outcomes. In biomedical research that gap reliably spans a decade or more, and the industry found it commercially productive.
NMN is the current exemplar. Nicotinamide mononucleotide is a precursor to NAD+, which declines with age and plays a documented role in sirtuin function and DNA repair. Human trials have confirmed NMN raises blood NAD+ levels. That’s an intermediate endpoint, the molecular step before any change in lifespan or health. The best small trials have found modest effects, such as better muscle insulin sensitivity in prediabetic women; none has shown a lasting, clinically meaningful health benefit. In November 2022, the FDA ruled NMN could not be sold as a dietary supplement because Metro International Biotech, a company co-founded by David Sinclair, had filed an Investigational New Drug application for it, triggering the drug exclusion clause. Following a Natural Products Association lawsuit and regulatory review, the FDA reversed course in September 2025, reinstating NMN as lawful while maintaining premarket notification requirements. The regulatory flip-flop resolved the marketing question without touching the scientific one.
Resveratrol’s trajectory, traced in the Sinclair box, is the instructive precursor. The market moved on the yeast result before the clinical evidence arrived, and by the time it did, the commercial cycle had already completed: the money had changed hands on findings that didn’t hold up at clinical scale.
The pattern is not unique to resveratrol. A mechanism appears in yeast or cell culture; a scientist with a platform communicates it broadly; a market willing to pay for hope funds the company; and the window between initial publication and clinical failure, reliably measured in years, is wide enough to build and exit before the results arrive.
Bryan Johnson’s Blueprint protocol, a $2 million annual personal regimen including dozens of supplements, aggressive biomarker monitoring and therapeutic interventions, is the reductio ad absurdum of this ecosystem. “Biological age” testing companies have proliferated, some using legitimate Horvath-derived epigenetic clocks, others selling proprietary tests of disputed validity. Longevity clinics prescribe rapamycin off-label, which is the use case where the scientific distinction matters most. Unlike NMN, rapamycin rests on real underlying science, and the difference between those two categories is exactly the information the industry has a financial interest in obscuring.
The arithmetic of more time
Aging research is usually framed as a medical question, but success would also raise economic and structural ones: who pays for much longer retirements, and who gets the treatments first.
Across OECD countries, people who leave the labor market today can expect, on average, about nineteen more years of life if they are men and twenty-three if they are women. That is the window pay-as-you-go pension systems, in which today’s workers pay for today’s retirees, are built to fund. Adding twenty years of healthy, productive life without moving retirement ages would break that design.
Around forty years between retirement and death would be a different kind of pension problem. The calculation that keeps current systems solvent, in which contributions from working-age people fund a relatively short retirement, stops closing. Systems that must serve forty years of retirement require higher contributions, longer working lives, lower benefits, or some combination of the three. The European Commission already projects EU age-related public spending rising from roughly 24.4% of GDP in 2022 to about 25.6% by 2070. That projection assumes current demographic trends, including gradual increases in life expectancy. It does not assume a successful longevity intervention.
Work through what even a partial success would mean. If a fifteen-year healthspan extension reached broad adoption across OECD economies within two or three decades, and retirement ages and benefits stayed where they are, the retirement liability would far exceed anything current demographic projections contemplate. The pension systems that face solvency questions on current projections would face them sooner and with larger deficits. Several countries already tie the retirement age to life expectancy, and how quickly such rules could keep pace with a sudden jump in lifespan would decide how large the gap became.
Access inequality is the second structural problem, and it’s worse than the pension arithmetic because it compounds across generations. Generic rapamycin costs little. The barrier to a longevity intervention is access: who is connected to a physician willing to prescribe an immunosuppressant off-label, and who has the metabolic monitoring infrastructure to manage side effects responsibly. Those who take part in the trials that generate the evidence in the first place also build the professional relationships and information access that make such use rational rather than reckless. Money, education and the right professional networks make every one of those doors easier to open.
The time lag in preventive health access is measured in decades, and wealth already concentrates at the top of the distribution. Twenty extra healthy years of compound returns on the largest fortunes would widen that gap fast.
The negotiation in progress
The Greenland shark has no longevity protocol. It takes no supplements and has no subscription to a biological age testing service. It is doing exactly what evolution built it to do in a cold, dark environment where a slow metabolism and heavy investment in molecular repair machinery made more sense than fast living and early death. Its genome carries expanded DNA repair gene families, most plausibly the trace of evolution working for long enough on a creature with its particular ecological pressures.
What makes the shark instructive is that evolution, through ordinary selection operating on standard biological components in a particular niche, produced the result that longevity researchers are spending billions of dollars trying to engineer. The machinery exists, in several very different versions across the tree of life. The question facing the research is whether that machinery can be understood well enough to be tuned in species (including one in particular) where evolution invested less heavily in maintenance.
The research has begun to answer this, slowly and not in the directions the industry prefers, and it stands closer to the salesmen’s promises than the fatalists assume and farther from them than the industry admits. The evidence for epigenetic reprogramming as a working principle is real. Rapamycin’s efficacy in animal models is as well established as any result in the field, and first-generation senolytics have reduced senescent cell burden in humans, in preliminary form, in a handful of small pilot studies.
Set against this, no candidate longevity drug has yet been shown to extend human life or to hold off the diseases of aging in healthy people, while an industry that has been selling hope in advance of evidence for twenty years shows no particular sign of stopping.
Gen AI Disclaimer
Some contents of this page were generated and/or edited with the help of a Generative AI.
Media
Key Sources and References
Williams, G.C. “Pleiotropy, natural selection, and the evolution of senescence.” Evolution 11(4): 398–411, 1957. The foundational antagonistic pleiotropy paper establishing the evolutionary logic of aging.
Kirkwood, T.B.L. “Evolution of ageing.” Nature, 1977. doi: 10.1038/270301a0. The disposable soma theory: energy trade-offs between maintenance and reproduction.
Wilkinson, G.S., and South, J.M. “Life history, ecology and longevity in bats.” Aging Cell 1: 124–131, 2002. https://onlinelibrary.wiley.com/doi/abs/10.1046/j.1474-9728.2002.00020.x. Comparative analysis across 64 bat species finding the longest-lived species are those that hibernate, use caves and reproduce slowly, consistent with low extrinsic mortality; records above 30 years in the wild exist for five species.
Austad, S.N., and Fischer, K.E. “Mammalian aging, metabolism, and ecology: evidence from the bats and marsupials.” Journal of Gerontology 46: B47–B53, 1991. Comparison across 580 mammal species: bats live at least three times as long as non-flying placental mammals of similar size, while marsupials live only about 80% as long despite lower metabolic rates.
Ruby, J.G., Smith, M., and Buffenstein, R. “Naked mole-rat mortality rates defy Gompertzian laws by not increasing with age.” eLife 7: e31157, 2018. https://elifesciences.org/articles/31157. Analysis of over 3,200 naked mole-rats demonstrating flat mortality rates in violation of the Gompertz pattern.
Andziak, B., and Buffenstein, R. “Disparate patterns of age-related changes in lipid peroxidation in long-lived naked mole-rats and shorter-lived mice.” Aging Cell, 2006. Naked mole-rats show higher lipid-peroxidation damage than mice, even when young.
Kerepesi, C., Meer, M.V., Ablaeva, J., et al. “Epigenetic aging of the demographically non-aging naked mole-rat.” Nature Communications 13: 355, 2022. https://www.nature.com/articles/s41467-022-27959-9. Bisulfite sequencing of over 100 naked mole-rats demonstrating epigenetic age drift coexisting with demographic non-aging.
Nielsen, J., et al. “Eye lens radiocarbon reveals centuries of longevity in the Greenland shark (Somniosus microcephalus).” Science 353(6300): 702–704, 2016. doi: 10.1126/science.aaf1703. Radiocarbon dating of 28 female Greenland sharks; species lifespan of at least 272 years; the largest individual estimated at 392 ± 120 years.
Yang, K., et al. “The Greenland shark genome: Insights into lifespan extremes and population dynamics.” Proceedings of the National Academy of Sciences 123(21): e2601272123, 2026. PMID 42154556. doi: 10.1073/pnas.2601272123. Chromosome-level genome assembly; gene-family expansion and positive-selection analyses point to DNA repair, immune function, NF-κB signaling and cancer resistance as potential longevity mechanisms.
Coppé, J.-P., et al. “Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor.” PLoS Biology 6(12): 2853–2868, 2008. doi: 10.1371/journal.pbio.0060301. Characterization of the SASP and its role in promoting malignancy in neighboring tissue.
Budinger, G.R.S., and Chandel, N.S. “Mitochondria dysfunction: cause or consequence of physiologic aging?” Genes & Development 39: 917–919, 2025. PMC12315864. https://pmc.ncbi.nlm.nih.gov/articles/PMC12315864/. Outlook (opinion) arguing that the evidence does not support mtDNA mutations as a causal limit on physiological aging, and repositioning mitochondria as signaling organelles.
Horvath, S. “DNA methylation age of human tissues and cell types.” Genome Biology 14: R115, 2013. doi: 10.1186/gb-2013-14-10-r115. Development of the multi-tissue epigenetic clock using 353 CpG methylation sites across 51 tissue and cell types; trained to estimate chronological age (DNA-methylation age).
Lu, Y., Tian, X., and Sinclair, D.A. “The Information Theory of Aging.” Nature Aging 3(12): 1486–1499, 2023. PMID 38102202. doi: 10.1038/s43587-023-00527-6. https://www.nature.com/articles/s43587-023-00527-6. Proposition that aging is primarily the loss of epigenetic information and that this information is in principle recoverable.
Lu, Y., et al. “Reprogramming to recover youthful epigenetic information and restore vision.” Nature 588: 124–129, 2020. doi: 10.1038/s41586-020-2975-4. OSK Yamanaka factor delivery to retinal ganglion cells restoring youthful methylation patterns and reversing vision loss in aged mice.
Howitz, K.T., Bitterman, K.J., Cohen, H.Y., et al. (including Sinclair, D.A.) “Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan.” Nature 425: 191–196, 2003. doi: 10.1038/nature01960. The foundational resveratrol-sir2 paper; spawned Sirtris Pharmaceuticals.
Venable LLP. “FDA Declares Nicotinamide Mononucleotide Is a Lawful Dietary Supplement.” October 2025. https://www.venable.com/insights/publications/2025/10/fda-declares-nicotinamide-mononucleotide-is. Legal analysis documenting FDA citizen petition response letters of September 29, 2025, reinstating NMN as a lawful dietary supplement under DSHEA, following reversal of the 2022 exclusion decision.
Yoshino, M., et al. “Nicotinamide mononucleotide increases muscle insulin sensitivity in prediabetic women.” Science, 2021. Ten-week placebo-controlled trial; the clearest physiological benefit of NMN in humans to date.
Colman, R.J., et al. “Caloric restriction delays disease onset and mortality in rhesus monkeys.” Science 325: 201–204, 2009. doi: 10.1126/science.1173635. Wisconsin National Primate Research Center 20-year CR study showing reduced age-related disease and mortality.
Colman, R.J., et al. “Caloric restriction reduces age-related and all-cause mortality in rhesus monkeys.” Nature Communications 5: 3557, 2014. https://www.nature.com/articles/ncomms4557. Follow-up to Colman 2009 extending the longevity findings.
Mattison, J.A., et al. “Impact of caloric restriction on health and survival in rhesus monkeys from the NIA study.” Nature 489: 318–321, 2012. doi: 10.1038/nature11432. NIA parallel study; caloric restriction improved some health measures but did not significantly improve survival.
Kraus, W.E., et al. “2 years of calorie restriction and cardiometabolic risk (CALERIE): exploratory outcomes of a multicentre, phase 2, randomised controlled trial.” Lancet Diabetes & Endocrinology 7(9): 673–683, 2019. https://www.thelancet.com/journals/landia/article/PIIS2213-8587(19)30151-2/abstract. 25% caloric restriction prescribed to non-obese adults over two years, about 12% (11.9%) achieved; improvements in multiple cardiometabolic biomarkers.
Spadaro, O., et al. “Caloric restriction in humans reveals immunometabolic regulators of health span.” Science 375(6581): 671–677, 2022. doi: 10.1126/science.abg7292. CALERIE participants showing CR-induced thymic mass increase and more recent thymic emigrant cells; CR lowered PLA2G7 expression, and deleting Pla2g7 in mice improved thymic and metabolic health, suggesting its reduction may mediate CR’s immunometabolic effects.
Harrison, D.E., et al. “Rapamycin fed late in life extends lifespan in genetically heterogeneous mice.” Nature 460: 392–395, 2009. doi: 10.1038/nature08221. NIA Interventions Testing Program; mean lifespan extended 9% in males, 13% in females, starting at 600 days of age.
Ivimey-Cook, E.R., Sultanova, Z., and Maklakov, A.A. “Rapamycin, not metformin, mirrors dietary restriction-driven lifespan extension in vertebrates: a meta-analysis.” Aging Cell, 2025. doi: 10.1111/acel.70131. https://onlinelibrary.wiley.com/doi/10.1111/acel.70131. Meta-analysis of 911 effect sizes from 167 papers; rapamycin mirrors CR effects; metformin does not.
Moel, M., et al. (AgelessRx). “Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results.” Aging (Albany NY) 17(4): 908–936, April 4, 2025. PMID 40188830. doi: 10.18632/aging.206235. https://www.aging-us.com/article/206235/text. 114 participants completed, 48 weeks, 5 or 10 mg/week compounded rapamycin (about one-third the blood concentration of commercial sirolimus at 24 hours) vs placebo; safe and well-tolerated; primary endpoint not met; lean mass and pain improved only in women on 10 mg, general health and emotional well-being in the 5 mg group.
Kraig, E., et al. “A randomized control trial to establish the feasibility and safety of rapamycin treatment in an older human cohort: Immunological, physical performance, and cognitive effects.” Experimental Gerontology 105: 53–69, 2018. doi: 10.1016/j.exger.2017.12.026. https://pmc.ncbi.nlm.nih.gov/articles/PMC5869166/. Earlier placebo-controlled trial: 1 mg/day rapamycin for about 8 weeks in 25 analyzed adults aged 70–95.
Hickson, L.J., et al. “Senolytics decrease senescent cells in humans: Preliminary report from a clinical trial of Dasatinib plus Quercetin in individuals with diabetic kidney disease.” EBioMedicine 47: 446–456, 2019. doi: 10.1016/j.ebiom.2019.08.069. First direct evidence that senolytics reduce senescent cells in humans; 9 subjects; measurable reduction in adipose tissue senescent cell burden within 11 days of 3-day oral D+Q.
Justice, J.N., et al. “Senolytics in idiopathic pulmonary fibrosis: results from a first-in-human, open-label, pilot study.” EBioMedicine, 2019. The first clinical trial of senolytics; 14 patients.
Koning, M., et al. “Senolytics dasatinib and quercetin in metabolic dysfunction-associated steatohepatitis: a proof-of-principle randomized, controlled trial.” Nature Metabolism, published online October 1, 2026. doi: 10.1038/s42255-026-01643-4. https://www.nature.com/articles/s42255-026-01643-4. Phase 2, double-blind; 31 participants with fibrotic MASH; fibrosis improvement without MASH worsening in 47% on D+Q vs 7% on placebo; MASH resolution 53% vs 7%; adverse events more frequent but self-limiting.
Harman, D. “Aging: a theory based on free radical and radiation chemistry.” Journal of Gerontology 11(3): 298–300, 1956. The foundational free radical theory of aging.
Harman, D. “The biologic clock: the mitochondria?” J Am Geriatr Soc, 1972. Harman’s extension of the free radical theory to mitochondria.
European Commission. 2024 Ageing Report: Economic and Budgetary Projections for the EU Member States (2022–2070). Publications Office of the European Union, 2024. https://economy-finance.ec.europa.eu/publications/2024-ageing-report-economic-and-budgetary-projections-eu-member-states-2022-2070_en. Projects EU aggregate age-related public spending at approximately 25.6% of GDP by 2070, rising from approximately 24.4% in 2022.
OECD. Pensions at a Glance 2025: OECD and G20 Indicators. OECD Publishing, Paris, 2025. doi: 10.1787/e40274c1-en. https://www.oecd.org/en/publications/pensions-at-a-glance-2025_e40274c1-en.html. Expected life years after labor market exit in 2024: OECD average 18.7 for men, 22.8 for women; Denmark, Estonia, Finland, Greece, Italy, the Netherlands, Portugal, the Slovak Republic and Sweden link the retirement age to life expectancy.
