What the hallmarks of aging actually claim

The hallmarks of aging are the most cited organising idea in the field, and the most commonly borrowed. Supplement pages invoke them, clinics list them, and the phrase “targets a hallmark of aging” carries an implication of established science that the framework itself does not supply.

What the framework actually is: a list of twelve biological processes proposed as the drivers of aging, and — more usefully — a set of conditions each one has to satisfy.

The three conditions

In their 2023 paper in Cell, López-Otín, Blasco, Partridge, Serrano and Kroemer state the premises explicitly. A hallmark must manifest with age. Experimentally accentuating it must accelerate aging. And intervening on it therapeutically must offer the opportunity to decelerate, stop or reverse aging.

The three conditions are the valuable part, because they are a test rather than a description. A process that merely correlates with growing older does not qualify. Making it worse has to make aging worse, and making it better has to make aging better, and both have to be shown rather than argued.

The twelve

The 2023 paper proposes: 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.

Three of those are new. The original 2013 version listed nine; disabled macroautophagy, chronic inflammation and dysbiosis were added in the expansion. That the list grew is not a weakness — it is what a research framework does when evidence accumulates — but it does indicate how far the field is from a closed account.

The authors also stress that the hallmarks are interconnected, both with each other and with a parallel set of proposed hallmarks of health. Nothing in the framework presents them as twelve independent dials.

What the framework does not settle

A 2026 review in the International Journal of Molecular Sciences went through the mechanistic and translational evidence for all twelve. Its finding is worth quoting in substance: experimental studies in model organisms support causal roles for several hallmarks, while their causal priority and necessity remain open questions.

That is the honest state of play. Several of these processes have been shown, in animals, to do something to aging when manipulated. Which of them matter most, which are causes and which are consequences, and how much any of it transfers to people are all unresolved.

The same review examined two further candidate processes — defects in RNA processing and remodelling of the extracellular matrix — without classifying either as an additional hallmark. The framework is still being argued over by the people who built it.

Why the vocabulary travels further than the evidence

A framework is useful to marketing precisely because it is organised, named and cited. A product that “supports mitochondrial function” or “targets cellular senescence” is pointing at a real entry on a real list in a real journal.

What that phrasing does not establish is any of the things a buyer assumes. That the compound reaches the process in a human body at the dose sold. That moving the process in that direction is beneficial. That a benefit to the process becomes a benefit to the person. Each of those is a separate claim requiring separate evidence, and the geroscience programme exists precisely because none of them can be assumed.

The National Institute on Aging’s testing programme is the sharpest illustration. Compounds proposed on strong mechanistic grounds, tested in mice at three sites under identical protocols: 30 agents over eleven years, with significant longevity effects published for six. A plausible mechanism is a reason to test something. It is not a result.