Why Do We Age? The Science of Aging and Longevity

BIOLOGY  /  LONGEVITY  /  GEROSCIENCE

Why We Age - and Whether Science Can Actually Slow It

A scientific guide to the biology of aging, longevity research, epigenetic clocks, senolytics, rapamycin and the first human trial of partial cellular reprogramming.

 

A visual overview of human aging as a biological process driven by cellular damage, mitochondrial dysfunction, senescence, inflammation and loss of resilience.

At 25, a bad night's sleep can feel like an inconvenience. At 75, the same kind of physiological stress may take longer to recover from. That shrinking ability to repair, adapt and bounce back is one of the most useful ways to think about aging. Gray hair and wrinkles are visible; the deeper story is a body slowly losing resilience.

For most of medicine's history, doctors have treated that decline one failure at a time. High blood pressure gets one drug. A tumor gets surgery or chemotherapy. A worn joint gets replaced. Geroscience asks a different question: if age is the biggest shared risk factor for cancer, heart disease, dementia, frailty and many other conditions, could we intervene farther upstream - in the biology that makes an older body vulnerable in the first place?

A few decades ago, that sounded close to science fiction. In 2026, it no longer does. That does not mean scientists have a pill that makes people young. It means something more subtle - and, scientifically, more interesting. Aging is becoming experimentally malleable. Researchers can slow parts of it in animals, shift some aging-related biomarkers in humans and, for the first time, test partial epigenetic reprogramming in people with age-related optic nerve disease.

That last sentence needs two brakes. Partial does not mean whole-body. A clinical trial does not mean a treatment works. We are nowhere near turning a 70-year-old into a 30-year-old. But we are finally asking, in real experiments rather than philosophy, which parts of aging can be changed and which cannot.

Aging Is Not One Process

There is no single switch inside the body labeled AGE. A better analogy is a city that has been running for decades. Roads are patched more often than rebuilt. Garbage collection misses things. Power stations become less reliable. Emergency services stay half-activated even when there is no crisis. Skilled workers leave faster than they can be replaced. No single failure explains the city. Together, they reduce what the city can survive.

Scientists try to map this mess with the 'hallmarks of aging' framework. A major 2023 Cell review expanded it to twelve interconnected processes: genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled macroautophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication, chronic inflammation and dysbiosis. The names are intimidating; the logic is not. In plain English, the body's information gets noisier, its cleanup systems weaken, energy and growth signals drift, damaged cells accumulate, repair capacity falls and tissues communicate less cleanly.

Worse, these problems do not stay in their own lanes. DNA damage can disturb gene regulation. Poor gene regulation can weaken protein quality control. Damaged proteins and mitochondria can trigger inflammation. Inflammation can disrupt stem cells and tissue repair. Senescent cells can push neighboring cells toward dysfunction. A disturbed gut microbiome can amplify immune signaling. One problem becomes fuel for the next.

That is why searching for one master 'cause of aging' may be the wrong quest. Aging looks less like a clock running down and more like a network gradually losing stability.

The hallmarks of aging describe the interconnected biological processes that gradually reduce the body’s ability to maintain and repair itself.

Why Would Evolution Allow This?

There is an obvious evolutionary puzzle here. If aging is so costly, why did natural selection not eliminate it?

Because evolution does not optimize a body for perfect maintenance forever. It strongly rewards traits that help an organism survive and reproduce, especially earlier in life. A biological trade-off can therefore be useful at 25 and costly at 75. Perfect repair is also expensive: energy spent endlessly fixing every molecule is energy not spent on growth, reproduction, immunity or surviving the next immediate threat. These ideas sit behind theories such as antagonistic pleiotropy and the disposable soma theory.

Neither theory explains everything, and aging biology is still debated. But they correct a common misconception. The body is not a machine built to remain pristine until a hidden warranty expires. It is a self-repairing system built under evolutionary trade-offs. Aging appears when damage, miscommunication and loss of regulation begin to outrun the systems that once kept them under control.

The Damage Is Real - but Damage Is Not the Whole Story

DNA: the instruction set gets noisier

Every cell has to protect and read an enormous instruction set while dealing with replication errors, reactive chemistry, radiation and ordinary metabolic wear. Cells have impressive DNA repair machinery, but it is not flawless. Over time, mutations and structural changes accumulate, particularly in long-lived tissues and stem-cell populations.

But aging is not simply a story of DNA filling up with typos. An old cell can still have a recognizable genome and yet behave differently from a young one. The deeper problem is that the genome becomes harder to maintain and read correctly: damage responses stay active, chromosomes become less stable and the molecular systems that decide which genes should be on or off drift away from youthful patterns.

Telomeres: important, but not the master countdown

Telomeres are protective stretches of repetitive DNA at the ends of chromosomes. In many human cells they become shorter as cells divide. When they become critically short, a cell may stop dividing or enter senescence. That made telomeres famous as a supposed internal countdown clock.

Reality is messier. Telomere shortening matters, but it is only one part of aging, and different tissues handle telomeres differently. Nor is simply turning telomerase back on an obvious cure. Endless cell division is exactly what cancer cells want. Any useful longevity therapy has to preserve repair without disabling the safeguards that stop damaged cells from growing forever. Biology rarely gives a free upgrade.

Proteins, autophagy and the cellular cleanup problem

A cell is not just DNA. It is also a crowded factory filled with proteins that must be built, folded, moved, repaired and destroyed at the right time. With age, that quality-control system becomes less reliable. Damaged or misfolded proteins can linger and aggregate - famously in neurodegenerative disease, but also as part of a much broader loss of cellular housekeeping.

One of the cleanup systems is autophagy, literally 'self-eating.' The name sounds dramatic; the job is practical. Cells dismantle worn-out components and recycle the useful material. This is one reason longevity researchers care so much about nutrient-sensing pathways such as mTOR and AMPK: these systems help a cell decide whether conditions favor growth and construction, or conservation, repair and recycling.

Mitochondria: more than worn-out batteries

Mitochondria are usually described as the cell's batteries. They are closer to power plants that also act as metabolic sensors and alarm systems. They generate usable energy, but they also influence stress responses, inflammation and programmed cell death. Aging mitochondria can become less efficient and more erratic. The older theory that 'free radicals cause aging' captured one piece of the story but made it too simple: reactive oxygen species can damage cells, yet they also carry useful signals. The real problem is loss of control over a deeply connected energy system.

Senescent Cells: When a Safety Mechanism Turns Toxic

Cellular senescence is one of the easiest aging mechanisms to picture. A senescent cell has largely stopped dividing. At first, that can be protective: it prevents a damaged cell from multiplying, contributes to wound healing and even plays roles in development. The trouble begins when these cells accumulate and do not get cleared.

They are often nicknamed 'zombie cells.' The label is memorable, but slightly wrong. Senescent cells are very much alive. Many release inflammatory signals, growth factors and tissue-remodeling molecules - a chemical cloud known as the senescence-associated secretory phenotype. In the wrong tissue and in sufficient numbers, that cloud can change how nearby healthy cells behave.

In mice, clearing selected senescent cells can improve multiple age-related functions and, in some experimental settings, extend healthy lifespan. That result created the field of senolytics: therapies designed to selectively kill senescent cells, or senomorphics that suppress their harmful signaling without necessarily killing them.

This is where the mouse-to-human gap matters. In laboratory animals, the results can look spectacular. In people, we still do not know whether clearing senescent cells will produce meaningful, durable anti-aging benefits. A 2025 Nature Aging commentary captured the challenge well: the field must identify which senescent cells are actually harmful, in which tissue, in which patient and at what point in disease. 'Kill zombie cells' is a simple slogan. The biology is not simple at all.

Senescent cells can protect tissues at first, but when they accumulate they begin releasing inflammatory signals that contribute to aging and disease.

The Body Also Ages as an Information System

Some of the most provocative aging research focuses not on physical damage, but on lost biological information. A liver cell and a neuron contain essentially the same DNA. They behave differently because each cell reads a different set of instructions. Think of DNA as the same enormous library in every cell; epigenetic marks and chromatin decide which shelves are open, which books are locked away and which pages are read repeatedly.

With age, that access system becomes less orderly. Patterns of DNA methylation - small chemical tags associated with gene regulation - change in reproducible ways. Researchers can feed those patterns into statistical models called epigenetic clocks and estimate aspects of aging from blood or tissue.

The phrase 'epigenetic clock' can be misleading. There is no tiny stopwatch inside a cell. A clock is a mathematical model trained to recognize patterns that correlate with age, mortality risk or physiological decline. That distinction becomes crucial the moment someone claims a treatment has made a person 'five years younger.'

Different clocks are built for different jobs, and they can disagree about the same person or the same intervention. A lower clock reading may be encouraging, but it is not automatically proof that muscles became stronger, arteries became healthier, dementia risk fell or lifespan increased.

That warning became especially important in August 2026. A Nature Medicine team reanalyzed data from 51 human intervention studies using the same panel of 16 prominent epigenetic clocks. Some pharmacological and lifestyle interventions pushed several biomarkers in a younger direction, but the clocks did not respond uniformly. The important result was not 'science has reversed aging.' It was that researchers now have a better way to ask which clocks are consistent enough to become useful trial endpoints. Nature Medicine had already summed up the mood in an editorial that month: the field needs less hype and more clinical evidence.

Epigenetic clocks estimate biological age by reading age-related patterns in DNA methylation, but different clocks can measure different aspects of aging.

Can We Actually Slow Aging?

So, can aging actually be slowed? The most defensible answer in 2026 is yes - probably in parts, and probably by modest amounts with interventions we can already test. That sounds less exciting than the longevity industry's promises. Scientifically, it is a much bigger statement than saying aging is completely untouchable.

The easiest way to get fooled in longevity science is to treat all evidence as equal. It is not. Making a worm live longer is a clue. Extending lifespan in a mouse is stronger. Moving a blood biomarker in a person is useful. Preserving strength, cognition or independence is better. Reducing cancer, cardiovascular disease, dementia or death in randomized human trials is the level of evidence that ultimately matters.

Exercise: boring, powerful, and still undefeated

Exercise does not sound futuristic, which may be why it is easy to underrate. Yet if the goal is to stay functional for longer, few interventions have stronger human evidence. Resistance training helps preserve muscle and strength. Aerobic fitness is strongly linked with lower cardiovascular and all-cause mortality. Regular physical activity also affects insulin sensitivity, inflammation, mitochondrial function, vascular health and the brain - several systems that overlap directly with aging biology.

Does that mean exercise literally makes every cell 'younger'? We cannot make that claim. But healthspan is not a laboratory score. If an intervention keeps people mobile, metabolically healthier and more resilient for longer, that matters more than whether one fashionable biomarker moves in the desired direction.

Calorie restriction: the strongest classic experiment

Calorie restriction - eating less without becoming malnourished - is one of the oldest and most reproducible lifespan interventions in animal research. It can extend life in many laboratory species, although the size of the effect depends on species, genetics, sex, diet and when restriction begins. Humans are harder: no one can run a perfectly controlled 70-year feeding experiment.

The CALERIE trial is therefore unusually valuable. It randomized 220 healthy adults without obesity for two years. The restriction group was asked to cut calorie intake by 25%, but in practice averaged about 12% - a useful reminder that real humans are not laboratory mice. A later DNA-methylation analysis found a roughly 2-3% slowing in DunedinPACE, a measure designed more like a speedometer for the pace of aging. Two other widely used clocks, PhenoAge and GrimAge, did not significantly change.

That small, messy result is almost a perfect snapshot of modern geroscience. There was a randomized signal. It was modest. Different biomarkers disagreed. And nobody can yet say that this means healthy people will live dramatically longer. That may sound anticlimactic, but this is what real progress often looks like before it becomes medicine.

Rapamycin: spectacular in mice, unresolved in humans

Rapamycin blocks mTOR, one of the cell's main growth-and-nutrient sensing systems. When nutrients are abundant, mTOR helps favor growth and protein production; when it is restrained, cells can shift toward maintenance and recycling. In multiple mouse studies, rapamycin extends lifespan, sometimes even when treatment starts relatively late in life. Few drugs have produced as much excitement in experimental gerontology.

The catch is obvious but important: rapamycin is a real drug with real biological costs. It is used as an immunosuppressant in transplant medicine, and dose and schedule matter enormously. Human studies of related drugs have produced interesting signals - including better responses to influenza vaccination in some older adults - but we do not have proof that healthy people can take rapamycin to live longer. For now it is a serious research candidate, not a validated longevity prescription.

Metformin: the anti-aging favorite with a growing question mark

Metformin became the celebrity drug of the longevity world for understandable reasons. It is cheap, widely prescribed for type 2 diabetes and affects pathways linked to energy sensing, mitochondrial metabolism and inflammation. Some observational studies also suggested unusually good outcomes in people taking it.

But longevity research is full of a dangerous pattern: a plausible mechanism becomes an association, the association becomes a headline, and the headline becomes a supplement-store certainty. A 2025 review in Ageing Research Reviews revisited the metformin story and found more uncertainty than the early enthusiasm implied. Some influential findings had methodological limitations, and trials in people without diabetes have not established broad anti-aging effects.

Metformin may still turn out to be useful for particular groups or particular age-related pathways. What we should retire is the stronger claim that it is already a proven general longevity drug for healthy humans.

Senolytics: a rational target waiting for decisive human evidence

The senolytic idea is beautifully simple on paper: remove chronically dysfunctional, inflammatory cells and give the tissue a chance to recover. Animal data make that idea hard to ignore. Human biology makes it hard to execute. Researchers still need better markers for the dangerous senescent cells, better ways to target them and trials that show actual improvements in function or disease - not merely a molecular change after treatment.

NAD boosters, resveratrol and the supplement problem

NAD metabolism, sirtuins and mitochondrial biology are legitimate research fields. That does not mean every capsule marketed around them slows human aging. Resveratrol, nicotinamide riboside and NMN have produced interesting mechanistic or biomarker results in some studies, but there is an enormous distance between 'changes a pathway in a study' and 'extends healthy human life.'

The longevity market is very good at hiding that distance. Science has to keep putting it back.

The Most Radical Idea: Reprogram the Cell Instead of Repairing It

Then there is the idea that sounds most like science fiction: instead of repairing an old cell piece by piece, reset part of its operating state. The discovery of induced pluripotent stem cells showed that adult cells can be pushed back toward an embryonic-like condition using a small set of transcription factors. During full reprogramming, many molecular signs of age are erased: telomeres can lengthen, epigenetic clocks can reset and old gene-expression patterns can disappear.

Full reprogramming is useless as a rejuvenation therapy if the cell forgets what it is. A neuron that becomes pluripotent is no longer doing the job of a neuron. Uncontrolled reprogramming also carries a serious tumor risk. The real goal is therefore partial reprogramming: restore some youthful regulation without wiping the cell's identity.

Animal and cell studies suggest that this separation may be possible. One widely studied strategy uses three of the Yamanaka factors - OCT4, SOX2 and KLF4, usually shortened to OSK - while leaving out c-MYC, a factor with strong oncogenic associations. The dream is not to turn an old cell into a stem cell. It is to make the old cell better at being itself.

That dream entered a new phase in 2026. The FDA cleared an investigational application for ER-100, an AAV-based therapy that delivers OSK instructions to retinal cells. The Phase I study began on March 2, and Life Biosciences announced the first participant had been dosed on June 9. Up to 18 people with open-angle glaucoma or non-arteritic anterior ischemic optic neuropathy are planned for enrollment. The main purpose is safety and tolerability; participants will be followed for years.

This is a genuine milestone, but not the one the headlines may tempt us to imagine. ER-100 is not a whole-body anti-aging treatment. It is a localized experimental gene therapy for serious optic nerve disease, and there are no published human efficacy results yet. If approaches like this succeed, the first real 'rejuvenation medicine' may not be an old person becoming young. It may be an old retina, liver or muscle recovering some function one tissue at a time.

Modern longevity research ranges from proven healthspan tools such as exercise to experimental interventions like senolytics, rapamycin and localized partial cellular reprogramming.

Is Aging a Disease?

Whether aging should be called a disease sounds like semantics until money, regulation and clinical trials enter the room. Medicine is built around named conditions: define the disease, find a target, measure an outcome. Aging is harder because it is universal, gradual and tangled into almost every major chronic disease of later life.

Geroscience does not actually require everyone to agree on the label. Its practical claim is enough: if several diseases share upstream mechanisms of aging, then those mechanisms may be legitimate therapeutic targets. A treatment that modestly reduces the risk of heart disease, frailty and dementia at the same time could be more important than one that moves a single biomarker dramatically.

This is why the fight over biomarkers matters. Waiting decades for lifespan data makes trials painfully slow. But accepting an epigenetic clock too early as a substitute for real health could be worse: a company might make a clock look younger without making a person healthier. The future of anti-aging medicine may depend as much on measuring aging correctly as on finding drugs that change it.

What We Can Say With Confidence in 2026

·         Aging is driven by multiple interacting biological processes, not a single cause.

·         Those processes are modifiable in laboratory organisms, sometimes dramatically.

·         In humans, lifestyle interventions can preserve function and reduce disease risk, and some interventions can shift molecular aging biomarkers.

·         No drug has yet been proven to broadly reverse human aging or reliably extend the lifespan of healthy people.

·         Rapamycin, senolytics, metformin and related approaches remain research programs with very different levels of evidence and risk.

·         Epigenetic clocks are useful research tools, but a younger clock reading is not yet synonymous with rejuvenation.

·         Partial epigenetic reprogramming has entered human testing in a localized eye-disease trial, marking a genuine milestone without proving organism-wide age reversal.

The Next 2, 5 and 10 Years

Next 2 years: better measurements, less patience for weak claims

Over the next two years, the most important fight will probably be about measurement, not immortality. Researchers are increasingly comparing epigenetic clocks with proteomic age, metabolomic signatures, immune aging, organ-specific markers and ordinary functional outcomes. The 2026 Nature Medicine analysis of 51 intervention studies points in that direction: stop asking whether one clock moved and start asking whether different measures agree - and whether that agreement predicts something a patient can actually feel or survive.

The ER-100 trial will be watched for the same reason. A reassuring safety profile would not prove rejuvenation. It would simply clear one of the first barriers to testing controlled, tissue-specific reprogramming more seriously.

Next 5 years: aging therapies may arrive disguised as disease therapies

The first convincing geroscience drugs may never arrive with 'ANTI-AGING' printed on the label. They are more likely to be approved for fibrosis, immune decline, osteoarthritis, neurodegeneration, metabolic disease or loss of vision while quietly acting on mechanisms shared with aging. Senolytics and mTOR-modulating drugs may find narrow, useful clinical niches long before anyone proves a general lifespan effect.

Combination therapy may also become more important. If aging is a network problem, changing one node may help only so much. Future studies could combine metabolic control, inflammation reduction, senescence targeting and tissue regeneration. The scientific challenge will be to gain more benefit without stacking risks just as quickly.

Next 10 years: the real question is whether rejuvenation can be made controllable

Ten years is enough time for several glamorous ideas in today's longevity field to disappear. It is also enough for one or two to become ordinary medicine. Partial reprogramming is the biggest wild card. If researchers can repeatedly restore function in old tissues without erasing cell identity, triggering tumors or provoking immune problems, the line between 'treating disease' and 'rejuvenating tissue' will become much harder to draw.

Whole-body age reversal is a far higher bar. Organs age at different rates. Delivering therapies throughout the body is difficult. Cancer risk is unforgiving. And a 70-year-old human is biologically far more complicated than a laboratory mouse. The plausible path is therefore uneven: first longer healthspan, then restoration of selected tissues, and only much later - if it proves possible at all - systemic rejuvenation.

The Real Goal Is Not Immortality

Longevity conversations often collapse into the least useful question: how long can a human live? Maximum lifespan is fascinating, but medicine has a nearer target with more immediate value - compress morbidity. Delay frailty. Keep the brain, heart, muscles, immune system and metabolism working for longer. Make the last decade of life contain fewer years of dependency and disease.

Imagine a treatment that adds no years to the calendar but gives a person five extra years of good mobility and independent cognition. It would still be an extraordinary anti-aging therapy in every way that matters to the person taking it.

That is the real scientific shift. We are not close to defeating death. What has changed is that aging is no longer treated only as the background against which diseases happen. It is increasingly treated as biology that can be measured, perturbed and tested.

And once something becomes testable, the question gets sharper. Not 'Can we live forever?' but: Which parts of aging can we change? By how much? In which organs? At what cost? And, most importantly, does the intervention make a human life healthier rather than merely making a laboratory number look younger?

Note: The drugs and experimental therapies discussed here are described for science reporting, not as medical recommendations. Rapamycin, senolytics, metformin and gene therapies have risks and should not be used for self-directed anti-aging treatment.