Biological age is often presented as a simple number — for example, 48 years instead of a chronological age of 55.
But where does that number come from?
Biological age cannot be measured directly. Researchers select biological markers associated with aging and use mathematical models to combine them into an estimate of biological age.
Different models do this in different ways.
Some traditional methods examine how biomarkers change with chronological age and use these relationships to estimate biological age.
Other approaches, such as the Klemera–Doubal method (KDM), were developed specifically to combine multiple biomarkers into an estimate of biological age. Newer approaches use machine learning to analyze larger and more complex sets of biological information.
So which model is best?
At present, there is no single answer.
Studies comparing biological-age methods have found that their performance varies. One method may perform better in predicting mortality or another health outcome in one study, while another may perform better in a different population or for a different outcome.
The result also depends on which biomarkers are included. Changing the biological measurements — or the method used to combine them — can change the estimated biological age.
This means that two biological-age calculations do not necessarily produce the same result, even when they are trying to answer the same general question.
Research has not yet established one mathematical model as the universally accepted best method for estimating biological age.
