Biological age

Biological age is an estimate of how worn a body is, produced by a statistical model rather than measured directly, and it is not a fact about a person in the way a birth date is. Two people born on the same day can be in visibly different condition, and the idea is to put a number on that difference.

The number always comes from a model, and the model was trained to predict something. That target matters and is rarely mentioned when a result is sold. Some models were trained to guess chronological age from a biological sample, so a person whose “biological age” reads low is a person the model found hard to place. Others were trained to predict time to death, or the rate at which health measures are declining. These are different quantities with the same label.

The models have been genuinely useful at the level of populations, where they have shown how behaviour, environment and social circumstances shape the risk of disease. Their usefulness for one individual is a separate question and a contested one. A 2025 review in Epigenomics argued that technical and biological properties of these algorithms currently prohibit their use at the individual level, and that they fail to meet the standards applied to established clinical biomarkers.

The practical consequence is that a biological age result is a research output, not a diagnosis. It supports a sentence about a group. It does not support a decision about a person.