Which Biological Age Calculation Method Works Best?
Summary ID01-BA-RS-00018
Research Summary
Original Publication
No Winners or Losers: Clinical Chemistry-Based Biological Aging Metrics Perform Similarly Across Cohorts and Health Outcomes
Journal
The Journals of Gerontology: Series A, .
- Category
- 01Biological Age
- Evidence
- Cohort study
- Summary Type
- Latest
What Was Studied?
Researchers compared four methods for measuring biological aging using clinical biomarkers: Klemera-Doubal Biological Age (KDM BA), Phenotypic Age, Homeostatic Dysregulation and Pace of Aging.
The analysis included 421,042 participants from five cohort studies in four countries: Italy, the United States, Canada and the United Kingdom.
Rather than allowing each method to use its usual biomarker panel, the researchers calculated the different metrics using consistent sets of biomarkers. The main analysis used nine clinical biomarkers, while a second analysis used a larger set of 15.
They then examined how strongly each measure was associated with six health outcomes: mortality, cardiovascular disease, diabetes, lung disease, self-rated health and number of comorbidities.
The central question was straightforward: does one method consistently capture biological aging better than the others?
Key Findings
No single biological aging metric consistently outperformed the others. Performance varied across health outcomes and populations, without one method emerging as consistently superior.
All four methods were associated with at least one of the health outcomes studied, and the overall strength of the associations was often remarkably similar.
Some methods performed better for particular outcomes. Phenotypic Age showed a stronger association with diabetes, while Homeostatic Dysregulation and Pace of Aging showed stronger associations with lung disease in some analyses. These differences did not produce one overall winner.
The different measures were only weakly or moderately correlated with one another in many comparisons, despite being calculated from the same underlying clinical biomarkers.
This may reflect differences in the aspects of biological aging captured by individual metrics, rather than the methods providing interchangeable estimates of a single underlying biological age.
Results varied between populations. The strength of associations differed substantially across the five cohorts, emphasizing that a method performing well in one population may not perform identically in another.
The larger 15-biomarker sensitivity analysis produced broadly similar conclusions.
Why It Matters for Longevity
Biological age is increasingly presented as a single number: a person may be told that their biological age is younger or older than their chronological age. This study shows why that number needs context.
The methods compared in this study approach biological aging from different perspectives. KDM Biological Age relates a person's physiological profile to age-related population patterns. Phenotypic Age incorporates biomarker information associated with mortality risk. Homeostatic Dysregulation assesses deviation from a reference physiological state, while Pace of Aging focuses on the rate of biological change over time.
Each approach may therefore describe a different dimension of biological aging.
For a Biological Age Calculator, selecting a scientifically established calculation method is essential, but this study provides no evidence that one of these methods should currently be regarded as the universal best choice.
Clinical Perspective
The study reinforces an important principle: a calculated biological age is a model-derived estimate, not a direct measurement of an underlying biological age.
There is currently no directly observable biological age against which these calculations can be definitively calibrated. Biological aging metrics are model-derived measures constructed from biomarkers according to different definitions and dimensions of aging.
A biological age calculation should therefore be evaluated not simply by whether it produces an age-related number, but by how well the resulting measure has been validated against meaningful health outcomes across different populations.
The study also demonstrates the importance of the reference population and clinical context. Associations varied substantially between cohorts despite standardized biomarker sets, indicating that performance observed in one dataset should not automatically be assumed to generalize to every population.
The absence of a clear winner has an important practical implication: choosing a method for a Biological Age Calculator requires understanding what each method was designed to measure and how it has been validated.
For clinical use, this argues against interpreting a Biological Age Calculator as a definitive measurement of how old a person's body really is. The method, biomarkers, reference population and outcome against which the calculation has been validated all influence what the result means.
Key Takeaway
There is currently no single biological age calculation method that consistently performs best. Different methods can use the same clinical biomarkers yet capture different dimensions of biological aging. Understanding the method behind a Biological Age Calculator is therefore essential to interpreting its result.
Reviewed and Summarized by
Source Publication
- Title:
- No Winners or Losers: Clinical Chemistry-Based Biological Aging Metrics Perform Similarly Across Cohorts and Health Outcomes
- Authors:
- Guillaume Provost, Kamaryn Tanner, Véronique Legault, Luigi Ferrucci, Stefania Bandinelli, Linda P. Fried, Daniel W. Belsky, Benoit Laurent, Alan A. Cohen
- Journal:
- The Journals of Gerontology: Series A, November 2025; 80(11):glaf179.
