Blood-cell study sharpens interpretation of biological-age clocks

A study published October 2 in Genome Medicine adds an important distinction to the interpretation of biological-age tests: blood-cell composition relates differently to a clock’s age estimate and to its measure of age acceleration. The findings help clarify what researchers are measuring when they use these tools to study healthier aging.

Across six epigenetic clocks, cell composition explained up to 53% of variation in DNA methylation age, compared with up to 21% in age acceleration. However, accounting for that composition only modestly weakened associations between age acceleration and subsequent health outcomes.

Two measurements with different meanings

Epigenetic clocks calculate aging-related scores from chemical marks on DNA. An age estimate and an age-acceleration score answer different questions. The latter measures how far a person’s estimate departs from what would be expected for their chronological age within the statistical model.

The distinction was already central to a May 2025 preprint from the research team. That earlier work analyzed 4,058 whole-blood samples from six Dutch cohorts, spanning ages 18–87, and estimated the proportions of 12 blood-cell types from methylation data.

Studying those proportions presents a mathematical problem: they must add up to 100%. If one rises, at least one other must fall. The researchers used principal component analysis to summarize patterns of variation without treating every cell fraction as independent. They also tested their findings using computationally constructed mixtures of cell methylation profiles.

The October publication extends the investigation to associations with 176 incident health outcomes in 18,859 Generation Scotland participants. Its endpoints concern biomarker variation and disease associations; this was not a treatment trial.

What the cell mixture contributes

The balance between naïve and memory T cells contributed strongly to DNA methylation age. For newer clocks, associations between cell composition and age acceleration involved neutrophils more prominently. These findings suggest that the two measures capture partly different biological processes.

Independent research has also shown why cell identity matters. A separate study of 428 samples representing 12 blood-cell types found substantial differences in biological-age indicators between cell types. T cells generally produced younger estimates and monocytes older estimates, with differences depending on the particular measure used.

That earlier study cautioned that cell composition needs consideration when clocks are used to evaluate aging interventions. It supplies useful context for the new paper, but does not independently replicate the new analysis of subsequent health outcomes.

Implications for healthspan research

One implication is that a change in an aging score needs a biological explanation before it can support a claim about improved health. Researchers must consider which clock changed, whether the blood-cell mixture changed, and whether clinically meaningful outcomes improved alongside it.

The findings also resist an overly simple dismissal of aging clocks. Blood-cell composition did not account for most of the observed relationship between age acceleration and later illness or mortality. Understanding one component of a biomarker does not establish that the remaining signal is either clinically useful or meaningless.

The evidence remains observational and methodological. Computational mixtures help investigate measurement behavior, but cannot show that deliberately changing immune-cell proportions prolongs life. The paper does not establish a treatment, validate a consumer test for individual decisions, or demonstrate that lowering a clock score prevents disease. Its practical contribution is a clearer framework for interpreting aging measurements and designing studies that connect them to health outcomes.