For a field obsessed with complexity, longevity science has an unusual attachment to one simple idea: aging can be organized into a finite set of recurring biological failures. That idea did not begin the modern longevity industry, but it gave the field something almost as valuable — a common language. Since the first “Hallmarks of Aging” paper was published in 2013, the framework has served as a guidepost for aging biology research and as a benchmark for anti-aging interventions across different contexts.
That matters because aging research is messy by definition. It spans DNA repair, metabolism, inflammation, stem cells, organ decline, frailty, dementia, cancer risk, immune drift, and the still unresolved question of what exactly counts as “biological age.” A framework that can hold all of that without collapsing into slogan or pseudoscience is rare. The hallmarks model became influential precisely because it did not claim to explain everything. It offered a disciplined way to sort cause from consequence, damage from adaptation, and mechanism from phenotype.
The 2013 paper gave the field its map
The original 2013 Cell review proposed nine hallmarks of aging: genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, deregulated nutrient-sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, and altered intercellular communication. The authors argued that a true hallmark should meet three tests: it should appear during normal aging, become worse when experimentally accentuated, and improve aging when attenuated. They also grouped the nine into three layers — primary hallmarks that drive damage, antagonistic hallmarks that initially compensate but later become harmful, and integrative hallmarks that help produce the organism-level decline people actually experience.
That architecture was more than a tidy diagram. It gave researchers a way to say that aging was not one thing, but also not an incoherent pile of things. DNA damage, proteostasis failure, and epigenetic drift could all be real at once. So could senescence, mitochondrial dysfunction, and stem-cell decline. The framework did not erase disagreement. It made disagreement legible.
Why the framework had to expand
A decade later, the same authors returned to the model with a blunt explanation: too much had happened to leave the 2013 version untouched. In the 2023 update, they wrote that close to 300,000 articles on aging had been published since the first edition, roughly as many as in the preceding century, and argued that the evidence base in mammals had become much stronger than it had been in 2013. They also noted that human age-related diseases were statistically more likely to co-occur and share genomic features when linked to the same hallmark rather than to different hallmarks — a sign that the framework had clinical as well as conceptual value.
The result was a new total: 12 hallmarks of aging. The updated list kept the original core but added disabled macroautophagy, chronic inflammation, and dysbiosis. Those additions were not arbitrary. Macroautophagy had previously sat inside the broader idea of proteostasis, but the authors argued it had become too important — and too biologically distinct — to remain a subcategory. Meanwhile, chronic inflammation and dysbiosis were extracted from the much broader bucket of altered intercellular communication, which in hindsight had become too vague to do all the conceptual work being asked of it.
That change is easy to underestimate. The shift from nine to 12 was not just an expanded checklist. It marked a deeper change in how the field thinks about causality. In 2013, the hallmarks framework mostly organized the molecular and cellular suspects behind aging. By 2023, it had become a more explicit theory of interacting systems — a network in which autophagy, inflammation, mitochondrial function, stem-cell performance, microbiome ecology, and cellular signaling all push and pull on one another. The 2023 paper says this plainly: the distinctions among hallmarks are “intrinsically diffuse,” their classification is “inevitably arbitrary,” and their interdependence is part of the point, not a flaw.
What really changed was the field’s mindset
The most important shift was not the extra three hallmarks. It was the field’s growing insistence that aging biology had to become actionable.
The updated framework leaned harder into intervention logic. It was no longer enough to name mechanisms. Researchers increasingly wanted to know which pathways could be measured, modified, combined, and eventually tested in humans. The 2023 hallmarks paper explicitly framed each hallmark as a potential point of entry for future anti-aging medicines. It also emphasized that objective quantification — through physiology, functional testing, and increasingly sophisticated omics — was essential if biological aging was going to become something more than metaphor.
This is where the hallmarks framework began to shape not just papers, but the entire structure of the modern longevity field. Once aging is organized into mechanistic domains, every part of the translational chain changes with it. Biomarker companies can ask which hallmarks they are actually measuring. Drug developers can ask whether an intervention is acting on one hallmark or several. Trial designers can ask whether they are tracking a pathway, a clinical outcome, or a surrogate that merely shadows chronological age. Clinics can ask — or should ask — whether the tests and protocols they market map onto biology that has any serious evidence behind it. That is an inference from the hallmarks, biomarker, and validation literature, but it is a strong one.
Biomarkers turned the framework from theory into infrastructure
If the hallmarks paper gave the field its map, biomarker research tried to build the measuring tools. In 2023, a large consortium paper in Cell argued that biomarkers of aging were “critically important tools” for identifying and evaluating longevity interventions within realistic human timeframes. The authors proposed a framework for classifying biomarkers, defining use cases, and setting validation steps — precisely because the field still lacked consensus on standards and methods.
The key tension is easy to grasp. A biomarker that tracks chronological age perfectly is not automatically useful for intervention science. What matters is whether it captures something biologically meaningful, predicts age-related outcomes, and ideally changes in response to interventions. The biomarker papers are notably careful here: there may never be one perfect aging biomarker, and even sophisticated algorithmic clocks can mix causal biology with mere correlation, population quirks, or technical noise. That caution is not a footnote. It is central to how the field now thinks about translation.
This is one reason the hallmarks framework has remained so influential. It does not solve the biomarker problem, but it gives biomarker developers a disciplined target. A useful aging measure should not just generate a biological-age number. It should, where possible, connect back to interpretable processes — inflammation, proteostasis, mitochondrial stress, cellular senescence, stem-cell decline, and the rest. In other words, the hallmarks model has become a kind of conceptual infrastructure for deciding what a biomarker is even supposed to mean.
The framework is already moving again
Even before the field had fully absorbed the 9-to-12 shift, the conceptual perimeter moved again. A 2025 Cell paper on precision geromedicine argued that aging science was entering a new phase centered on actionable hallmarks, gerogenes, and gerosuppressors — genes and pathways that accelerate or slow biological aging. The authors wrote that new molecular profiling technologies were enabling these pathways to function as biomarkers and opening the door to more individualized gerotherapeutic strategies.
And, crucially, that 2025 work did not simply rest with 12 hallmarks. It explicitly pointed to two more: extracellular matrix changes and psychosocial isolation. A later 2025 perspective by López-Otín and Kroemer then described the framework in terms of 14 biological hallmarks and argued for tighter integration of biological and social determinants of aging. In that piece, they wrote that extracellular matrix changes had been added because evidence increasingly showed a body-wide reduction in matrix viscoelasticity with age, while psychosocial isolation reflected the growing recognition that social and behavioral stressors are not external to aging biology; they shape it.
That matters for a simple reason: the hallmarks model is no longer only a map of what goes wrong inside the cell. It is becoming a framework for understanding how cells, tissues, behavior, exposures, and social life converge to shape aging trajectories. That does not mean every longevity company that starts talking about “psychosocial” aging suddenly becomes credible. It means the frontier is widening — and the serious end of the field is widening with it.
The case for the hallmarks — and the case against overconfidence
The hallmarks framework has survived because it is useful, but it is not beyond criticism. Even sympathetic reviewers note that the evidence base remains heavily concentrated in humans and a few model systems, and that not all hallmarks are equally easy to assess across diverse species. A 2025 review for comparative biologists made the point directly: the hallmarks have become guideposts for aging biology, yet many of them are still backed by restricted datasets and uneven measurement tools.
That limitation is not trivial. Any framework that becomes too influential risks becoming self-reinforcing. Researchers look where the map tells them to look. Companies market whatever fits the map. Biomarkers are optimized around what the map can currently describe. The danger is not that the hallmarks are wrong. It is that they are treated as finished. The authors themselves have repeatedly warned against that reading by stressing that the categories are diffuse, interdependent, and subject to revision.
Still, it is difficult to name another framework in longevity science that has done more with so little conceptual clutter. It gave the field a way to sort mechanisms without pretending aging is reducible to one magic pathway. It helped shift the conversation from lifespan theater toward healthspan, resilience, and measurable decline. And as biomarkers, clinical trials, and preventive-care models keep moving closer to the center of the longevity economy, the hallmarks framework is likely to remain the default grammar in which those arguments are made. That conclusion follows from the way the framework papers, biomarker work, and validation literature now fit together.
The real lesson
The point of the hallmarks was never that aging could be reduced to a single elegant poster. The point was that aging could be studied as a structured, testable, interconnected biological process. That was the breakthrough in 2013. The 2023 expansion showed the field had learned enough to redraw the map. The 2025 precision-geromedicine turn suggests the map is now being used to build diagnostics, interventions, and — eventually — more personalized forms of care.
So yes, the shift from nine to 12 hallmarks was real. But the bigger story is that the model has become the organizing logic of modern longevity science — and it is still evolving. That is exactly why it matters.