There is no single biological age

Chronological age is time since birth. Biological-age models ask a different question: how does a set of measured characteristics compare with patterns seen across age or health outcomes in a reference population?

Different models can legitimately produce different answers because they use different biomarkers, endpoints, transformations, and reference datasets. A laboratory-based mortality-risk model, a wearable-derived fitness age, and a multi-domain longitudinal estimate are not interchangeable even when all three are expressed in years.

What makes an age estimate interpretable

  • The inputs: which physiological, behavioral, or laboratory variables are included.
  • The reference: the population and age range used to calibrate the model.
  • The endpoint: age itself, mortality risk, morbidity, function, or another target.
  • The timescale: whether the model is intended to move over days, months, or years.
  • The uncertainty: missing data, device error, and model assumptions.

Why two scores may disagree

A disagreement is not automatically evidence that one model is wrong. It can reflect a different scientific question. Levine PhenoAge (clinical), for example, was developed from routine clinical biomarkers and mortality risk; wearable-age products may emphasize fitness, sleep, or cardiovascular features instead.

The practical question is therefore not only “What is my age score?” but “What produced this estimate, and is that model appropriate for the decision I am trying to make?”

How BioMIR approaches the problem

BioMIR treats Adaptive BioAge as a model output expressed in age-equivalent units rather than a literal measurement of aging. The daily estimate is separated into Behavioral, Functional, and Cardiometabolic contributions, while established clinical clocks such as KDM and Levine PhenoAge remain distinct long-view comparisons.

That separation is deliberate: a daily change in higher-frequency physiological signals should not be interpreted as equivalent to a change in an episodic laboratory clock or as evidence that lifespan has changed.

Selected references

  1. Klemera & Doubal (2006), A new approach to the concept and computation of biological age
  2. Liu et al. (2018), A new aging measure captures morbidity and mortality risk across diverse subpopulations
  3. Furrer & Handschin (2025), Biomarkers of aging: from molecules and surrogates to physiology and function