An epigenetic clock is not enough: what can the brain, walking and PAF tell us about ageing?

Biological age can provide a useful clue. It is not a diagnosis of the whole body, or proof that ageing has been reversed. A new paper in Aging Cell offers a reason to examine this distinction — and to ask harder questions of the Polycentric Aging Framework.

A marker is not yet a mechanism

Consider a warning light on a dashboard. It may reliably signal trouble, but switching it off does not necessarily repair the engine. Ageing biomarkers require similar care: a measurement can predict future health without itself causing health to deteriorate.

Epigenetic clocks use DNA methylation patterns — chemical marks associated with the regulation of gene activity. A “clock” is a computational model, rather than a single biological mechanism. A change in its output must be distinguished from a change in the processes that actually limit function and lifespan.

What does the new study show?

During 16-year follow-up in a Scottish cohort, GrimAge2 was strongly associated with mortality. In a joint analysis of 21 biomarkers among 460 participants, four measures retained independent significance: brain volume, MRI white matter changes, cognitive ability and walking time. This observational study used baseline measurements in relatively healthy, similarly aged people. It does not establish causality. [1]

The interesting distinction is between two questions: “is this measurement associated with future risk?” and “does it add information beyond the other measurements?”. Neither yet answers: “what should we change to help people live longer in good health?”.

If several markers describe overlapping aspects of the same problem, some can lose significance when entered into one model. This does not automatically make the underlying biology irrelevant. The strength of a statistical association also does not settle a test’s usefulness in clinical practice. Nor should this analysis be turned into a universal recommendation for a particular testing panel.

Where does PAF fit?

In my published Polycentric Aging Framework, ageing results from partially independent, interacting centres. A “centre” means a set of processes capable of limiting biological maintenance and renewal — not simply an organ or an individual laboratory result.

The framework proposes a special, dominant but non-exclusive role for epigenetic regulation in limiting human lifespan. This remains a hypothesis to test. It does not imply that one methylation clock measures all ageing, or that every epigenetic change is harmful. Some changes may respond to damage or help compensate for it.

PAF also considers constraints arising from genetic and physical damage, and from impaired quality control and removal of worn cellular components. Restoring regulation may not remove all these burdens. In this view, an intervention’s effectiveness also depends on what actually limits regeneration in a particular tissue and at a particular stage.

The new results are consistent with the need for a multidimensional assessment of the body. They do not establish the hierarchy of centres proposed by PAF. A proteomic organ-age estimate does not, by itself, identify a causal centre. Likewise, a strong clock–mortality association does not establish epigenetic dominance.

A fair test of the framework

PAF should generate testable predictions. If epigenetic regulation is a major constraint, targeted modification should produce sustained improvements in relevant functions, while accounting for other damage and constraints. Assessment should include safety, function, morbidity and, where feasible, survival.

Imagine, hypothetically, that an intervention lowers a clock estimate by five years, but the participant does not walk better, cognition does not improve and disease risk remains unchanged. That result supports a claim about a biomarker change. Calling it “rejuvenation of the body” would require further evidence.

The framework must also allow findings that challenge its own hierarchy. If well-designed studies repeatedly improved function and survival through other mechanisms, without corresponding changes in epigenetic regulation, the scope of the dominance hypothesis would need reassessment. An unchanged result from one clock, however, would not be enough to establish that conclusion.

The central question is therefore: have we changed a measurement, or improved the body’s capacity to live in good health? Answering requires molecular measurements to be connected with function and intervention studies. That is the level at which these new data should inform discussion of PAF.

Reference and context

Reference update: 11 October 2026. The paper describing PAF was published on 9 October 2026. Michał Bartoszewicz created the model; Michał Bartoszewicz and Mariusz Niemczyk authored the publication.

Bartoszewicz M, Niemczyk M. The polycentric aging framework: epigenetic regulation as a dominant lifespan-limiting center in human aging. Aging. 2026;18:1365–1423. DOI: 10.18632/aging.206426.

[1] Pyrgioti M. et al. Aging Cell, 2 October 2026. DOI: 10.1111/acel.70747 — open-access full text.

PAF: assumptions, authorship and source publication.

A literature commentary by Michał Bartoszewicz. The interpretation in relation to PAF is the author’s analysis, rather than a conclusion attributed to the study’s authors.

Editorial preparation and translation assisted by AI tools. PAF and the interpretative direction: Michał Bartoszewicz.

Przeczytaj wersję polską