The Twelve Hallmarks of Aging

In 2013, five researchers published a paper in Cell that did for aging what a good filing system does for a chaotic desk. Aging research had accumulated decades of findings (telomeres, free radicals, stem cells, inflammation) without a shared structure to hold them. Carlos López-Otín, María Blasco, Linda Partridge, Manuel Serrano, and Guido Kroemer proposed nine “hallmarks”: distinct, measurable biological processes that together describe how a body ages.

A decade later, in January 2023, the same authors returned with an update titled “Hallmarks of Aging: An Expanding Universe.” The list grew to twelve. More importantly, the picture of how they relate to one another sharpened considerably, and that relational picture, not the list itself, is where the practical insight lives.

The twelve are sorted into three groups, and the grouping is the whole point.

Tier 1: Primary Hallmarks, the Original Damage

Primary hallmarks are unambiguously bad. There is no dose at which they are helpful. They are the upstream insults from which everything downstream follows.

1. Genomic Instability: Every cell in your body sustains tens of thousands of DNA lesions per day: from ultraviolet light, environmental toxins, and radiation on the outside, and from replication errors and reactive oxygen species generated by your own metabolism on the inside. You have elaborate repair machinery for this, but the machinery itself degrades with age. The result is a slowly accumulating mutational burden across every tissue. This is the most upstream hallmark of all. It directly triggers cellular senescence, drives cancer, and degrades stem cell function.

2. Telomere Attrition: Telomeres are protective caps on the ends of chromosomes, functionally similar to the plastic tips on shoelaces. They shorten a little with every cell division. When they reach a critical length, the cell is forced into permanent retirement, called senescence, or destroys itself outright. Elizabeth Blackburn shared the 2009 Nobel Prize for establishing how telomeres and the enzyme telomerase work, and the research that followed showed telomere length behaves like a biological clock, one that chronic stress, smoking, and poor sleep can measurably speed up.

3. Epigenetic Alterations: Your DNA sequence is largely fixed; the annotations layered on top of it are not. DNA methylation patterns, histone modifications, and chromatin architecture all drift with age. Tumor suppressor genes get quietly switched off. Inflammatory genes get switched on. This drift is regular enough to be read as a clock. Steve Horvath’s 2013 epigenetic clock, built on 353 methylation sites and validated across 51 human tissue and cell types, estimates biological age independently of the number on your birth certificate. It has since become one of the most important measurement tools in longevity research, the closest thing the field has to a scoreboard for whether an intervention is actually working.

4. Loss of Proteostasis: Proteins have to fold into precise three-dimensional shapes to do their jobs. Three systems maintain that precision: heat shock proteins that chaperone folding, the ubiquitin-proteasome system that shreds defective proteins, and autophagy that clears the rest. All three decline with age. What accumulates is misfolded protein aggregates: amyloid-beta and tau in Alzheimer’s disease, alpha-synuclein in Parkinson’s, and damaged structural proteins nearly everywhere else.

5. Disabled Macroautophagy: This is one of the three additions in 2023, and it was promoted out from under proteostasis for a specific reason. Autophagy, literally “self-eating,” is the cell’s recycling program, but macroautophagy handles cargo the proteasome cannot: entire worn-out organelles, large aggregates, cellular debris at scale. Losing it is not the same problem as losing chaperone function. Its failure is directly implicated in neurodegeneration, cancer, cardiovascular disease, and the inflammatory drift discussed below.

Tier 2: Antagonistic Hallmarks, Protection That Overstays

Antagonistic hallmarks are the interesting ones. Each began as an adaptive response to the damage above. At low intensity, each is protective. Sustained chronically, each becomes more destructive than the insult it was answering.

6. Deregulated Nutrient-Sensing: The mTOR, IGF-1, and insulin signaling pathways evolved as a metabolic thermostat: build and grow when food is plentiful, conserve and repair when it is scarce. Modern life, with its abundant calories and minimal movement, leaves the thermostat jammed on “grow.” Chronic overactivation suppresses autophagy, accelerates senescence, and degrades mitochondrial quality control. This is the formal hallmark under which insulin resistance and most metabolic dysfunction properly sit.

7. Mitochondrial Dysfunction: Damaged mitochondria produce reactive oxygen species instead of efficient ATP, and the failure is self-amplifying: ROS damages mitochondrial DNA, damaged mitochondrial DNA impairs the electron transport chain, and an impaired chain generates more ROS. The consequence is not only an energy deficit. Failing mitochondria leak signals that activate inflammatory pathways directly, feeding the tier below.

8. Cellular Senescence: When a cell sustains dangerous DNA damage, forcing it to stop dividing permanently is a sensible defense, and it is a large part of why we do not all develop cancer in our thirties. The trouble is disposal. Senescent cells that the immune system fails to clear accumulate with age and secrete SASP, the senescence-associated secretory phenotype: a cocktail of inflammatory cytokines, proteases, and growth factors that damages neighboring cells and drives essentially every age-related disease. Senolytics, drugs designed to selectively clear these cells, are among the most active therapeutic frontiers in the field.

Tier 3: Integrative Hallmarks, the Aging You Can See

Integrative hallmarks emerge when damage overwhelms the body’s capacity to compensate. This is where aging stops being a molecular abstraction and becomes something visible at the level of organs and whole organisms.

9. Stem Cell Exhaustion: Every tissue renews itself from resident stem cell pools. Those pools absorb the cumulative genotoxic, epigenetic, and mitochondrial damage of Tiers 1 and 2, while the inflamed SASP-saturated environment poisons the niches they live in. Self-renewal capacity falls. The visible results: a thinning gut lining, muscle that rebuilds poorly, reduced immune cell output, and slower wound healing everywhere.

10. Altered Intercellular Communication: Cells coordinate through hormones, cytokines, extracellular vesicles, and direct gap junctions. With age, this network degrades in a characteristic way: hormonal output falls (growth hormone, IGF-1, testosterone, estrogen), pro-inflammatory signaling rises, and the precise cross-talk between tissues that maintains homeostasis becomes noisy. Neuroendocrine aging and much of what clinicians recognize as frailty live here.

11. Chronic Inflammation (“inflammaging”): In 2023, inflammaging was promoted to a hallmark in its own right rather than treated as a symptom of communication failure. The reasoning is that persistent low-grade sterile inflammation, driven substantially by SASP from uncleared senescent cells, turns out to be the single most connected node in the entire network. It does not merely result from the other hallmarks; it feeds back into and amplifies all of them at once.

12. Dysbiosis: Also new in 2023. The gut microbiome degrades with age: short-chain-fatty-acid-producing species are lost, pro-inflammatory taxa expand, and overall diversity contracts. A decade of accumulated evidence now ties microbiome composition to inflammation, immune competence, neurodegeneration, and metabolic health. Studies of centenarians repeatedly find a distinctively diverse, SCFA-rich gut community that looks considerably younger than its owner.

The Insight That Actually Matters

The tiers are not a one-way conveyor belt. They form a dense network with bidirectional feedback running in every direction. Tier 3 inflammaging generates reactive oxygen species that damage DNA, feeding straight back to Tier 1. Senescent cells in Tier 2 exhaust the stem cell niches of Tier 3. Dysbiosis sustains the inflammation that accelerates all twelve.

But directionality still matters, and this is the practical payoff. Intervening upstream, at Tiers 1 and 2, has more leverage than intervening downstream, because it reduces load on the entire system simultaneously rather than mopping up one consequence at a time.

That is the founding logic of geroscience: target the biology of aging itself, and you compress the window of multimorbidity more effectively than managing diseases one at a time, each after it has already arrived.

What This Means Monday Morning

The most striking feature of the hallmarks framework is how unglamorous its best-supported interventions turn out to be. Nothing here requires a clinic.

Exercise is the broadest-acting intervention known. Resistance training and aerobic work together improve nutrient-sensing (Tier 2), stimulate mitochondrial biogenesis (Tier 2), induce autophagy (Tier 1), and lower systemic inflammatory markers (Tier 3). No drug currently touches that many hallmarks at once.
Sleep is when the brain’s glymphatic clearance of protein aggregates runs, when growth hormone is released, and when inflammatory tone resets. Chronic short sleep measurably accelerates telomere attrition and epigenetic aging.
Dietary quality (fiber, polyphenols, fermented foods, adequate protein) feeds the SCFA-producing bacteria that dysbiosis depletes, while limiting the metabolic load that keeps nutrient-sensing pathways jammed open.
Caloric modulation, whether through moderate restriction or time-restricted eating, is the intervention with the longest track record across model organisms. It works by releasing the brake on autophagy and dialing down mTOR and insulin signaling, hitting Tiers 1 and 2 at their most leveraged points.

None of this is new advice. What the hallmarks framework supplies is the mechanism: an account of why these four things work, which hallmarks each one touches, and why they remain more broadly effective than anything currently available in a bottle.

Going Further

Once the four foundations are genuinely in place, not aspirationally but actually, a reasonable next question is whether anything else adds meaningful leverage. There is real science here, but the signal-to-noise ratio in the consumer longevity market is poor, and the gap between “biologically plausible” and “demonstrated to help humans” is wide. This is the point at which a clinician earns their keep.

What to Look For in a Clinician

The right partner is someone trained in integrative or functional medicine who nonetheless thinks like an evidence-based physician: comfortable with off-label and preventive strategies, but candid about what remains unproven, willing to test rather than assume, and, importantly, not selling the supplements they recommend. Certification through a body such as the Institute for Functional Medicine or the American Board of Lifestyle Medicine is a reasonable starting filter. A clinician who says “we don’t know yet” about most of what follows is demonstrating competence, not timidity.

Compounds With the Strongest Human Data

A handful of interventions have moved past cell culture and mouse studies into randomized human trials with real endpoints. Urolithin A, a metabolite produced from pomegranate ellagitannins, has improved muscle endurance and mitochondrial biomarkers in randomized trials in older adults, mapping directly onto mitochondrial dysfunction and macroautophagy. Omega-3 fatty acids, vitamin D, and creatine each have substantial trial literature touching inflammation, immune function, and muscle maintenance. Fiber and fermented-food interventions have measurably shifted microbiome diversity and inflammatory markers, addressing dysbiosis.

Compounds That Are Promising but Unsettled

NAD+ precursors such as nicotinamide riboside and NMN raise NAD+ levels reliably in humans; whether that translates to clinical benefit is still unresolved. Spermidine, which induces autophagy, has encouraging observational and small-trial data. Senolytics, including the dasatinib-plus-quercetin combination and fisetin, remain among the most exciting ideas in the field, with early trials in osteoporosis, cognitive impairment, and diabetic kidney disease. But no regimen is established for general use, and dasatinib is a chemotherapeutic with a real toxicity profile.

Repurposed Medications

Metformin’s geroprotective reputation rests largely on observational data; the TAME trial designed to test it properly has been discussed for over a decade and still lacks full funding. Low-dose intermittent rapamycin, the most direct available lever on mTOR, was studied in the 48-week randomized PEARL trial: it proved reasonably safe, missed its primary endpoint, and produced modest secondary signals in lean mass and self-reported wellbeing. SGLT2 inhibitors and GLP-1 receptor agonists are increasingly discussed in geroscience terms for their metabolic and anti-inflammatory effects. None of these is approved for aging. Prescribing them for that indication is off-label, and warrants a documented conversation about what is known, what is not, and what will be monitored.

What Good Practice Looks Like

Establish a baseline first: metabolic panel, inflammatory markers, body composition, and, if you want a hallmark-level readout, an epigenetic age estimate (interpreted cautiously, since these clocks are noisy at the individual level). Change one variable at a time. Retest on a defined schedule. Discontinue anything that isn’t demonstrably doing something. And treat supplements as what they are: a category subject to far less regulatory oversight than pharmaceuticals, where product quality varies enormously and third-party testing is not optional.

The framework’s own logic argues for humility here. Nothing in a bottle currently touches as many hallmarks, as reliably, as a hard training session and a full night’s sleep. The additions are additions, not substitutes.

References

ADDENDUM: A Gap in the Framework

August 5, 2026

Does the hallmarks framework account for cell-free chromatin particles? I went looking. It does not, and the way it doesn’t turns out to be more interesting than a simple omission.

First, What These Particles Are

Every day, somewhere between fifty and seventy billion of your cells die. This is normal and necessary. What is less appreciated is what they leave behind.

When a cell dies, its nucleus ruptures and spills its contents. What escapes is chromatin, the cell’s DNA still wound around its histone proteins, broken into fragments. Those fragments enter the circulation as cell-free chromatin particles.

They are not inert debris. Healthy cells absorb them. Once inside, the fragments can migrate to the nucleus and fuse with the host genome, producing double-strand breaks, the most severe form of DNA damage. They also trip the ancient sensors our cells use to detect foreign DNA, because a fragment of a dying neighbor’s chromatin looks a great deal like a virus. That sets off inflammation. And they injure mitochondria, which respond by leaking reactive oxygen species.

The most useful way to think about them is as metabolic waste, in the same category as advanced glycation end products or oxidized LDL. The body generates them continuously and must clear them continuously. When clearance falters, through age, illness, or a heavy burden of dying cells, they accumulate and their damage compounds.

Why This Should Matter to Anyone Trying to Slow Aging

Three reasons.

The first is leverage. If the model holds, these particles are not a symptom of aging but a driver of it, and they produce DNA damage, mitochondrial dysfunction, and inflammation at the same time. That is three hallmarks from one source, spanning all three tiers. The original article argued that upstream interventions beat downstream ones. This would be further upstream than any of the twelve.

The second is that waste can be cleared. Telomere attrition is hard to reverse. Epigenetic drift is hard to reverse. But a circulating toxin is a different kind of problem, one where the goal is not to boost a pathway but to neutralize something and carry it out. That is a more tractable engineering problem than most of what the longevity field is attempting.

The third is conceptual. We have always described aging damage as accumulating. Accumulation implies that something is physically accumulating. Cell-free chromatin particles are a candidate for what that something actually is.

How Close the Framework Comes

The 2023 paper never mentions them. The terms “cell-free chromatin,” “cell-free DNA,” and “extracellular chromatin” appear nowhere in it, nor in the 2013 original. Neither bibliography, 297 references in the first paper and 300 in the second, cites Indraneel Mittra, whose group at Tata Memorial Centre has produced most of the work in this area.

What makes the absence notable is how near the paper comes before stopping.

Foreign-DNA sensing runs all through it. The cGAS-STING pathway, one of the cell’s alarm systems for stray DNA, appears under genomic instability, epigenetic alterations, cellular senescence, and chronic inflammation. But in every case the DNA setting off the alarm is traced to one of two sources: dormant viral-like sequences in our own genome that reawaken with age, or mitochondria leaking their own DNA. The chromatin always originates inside the cell it goes on to damage.

The single closest passage is a clause in the senescence section describing HMGB1, a protein that normally sits in the nucleus alongside DNA, escaping the cell and acting as an inflammatory signal in the tissue around it. That is chromatin material leaving a cell and causing trouble elsewhere. But it is the protein only. No DNA, no uptake by a recipient cell, no integration into another genome.

There is even an obvious place to put the idea. The hallmark of altered intercellular communication already covers blood-borne effects, including parabiosis experiments and what the authors call “contagious aging” through extracellular vesicles. It is exactly where circulating chromatin would belong. At present it contains no nucleic acid of any kind.

The Oversight Runs Both Directions

Fairness requires noting that Mittra’s group has not pressed its case inside the hallmarks literature either. Their major papers propose an explicit unifying model of aging without the word “hallmark” appearing anywhere. Even the 2022 mouse study framed around downregulating “biological hallmarks of ageing” cites López-Otín a single time, as background, and measures eight markers of which only five are hallmarks in that framework.

So this is not really a case of one side ignoring the other. It is two bodies of work making overlapping claims about the same upstream biology and largely not reading each other.

Where This Leaves Us

This remains a hypothesis. The human data is early; most of it comes from a single research group, and the mechanism needs independent replication before anyone redraws the diagram. Nothing here changes the advice in the original article, where exercise, sleep, diet, and caloric modulation still carry more evidence than anything in a bottle.

But an idea with this much reach should at least be argued over. At the moment it is not being argued over at all, which is a stranger outcome than being wrong.

References

Dr. Daniel Thomas, DO, MS