Open Physiological Age · draft 0.1
Biological age, measured in the open.
An open standard for estimating biological age from wearables and routine clinical tests. Every rule, coefficient, citation and validation result is public, and scientists, not vendors, decide how it changes.
wearable "age" scores we reviewed publish their coefficients or code.
How weakly a clock's accuracy against calendar age tracks its ability to predict death, across 39 clocks benchmarked by Biolearn.
in the draft, each with its hazard ratio, evidence grade and citation, from VO₂max to sleep regularity.
The problem
Every "age" on the market is a black box
Same word, different science. A watch, a ring and a blood test can give one person three ages, and nobody outside the company can check how any of them was made.
| Vendor | Metric | Built from | What is public | Openness |
|---|
Compiled on 1 November 2026 from public vendor pages, white papers and press releases; it describes what is publicly disclosed, not product quality. Academic clocks such as PhenoAge, KDM and DunedinPACE are published, but none covers wearables and routine labs together. Corrections are welcome at bruno@healthspancommons.org.
Reference implementation, running in your browser
Try OPA 0.1
Pick an example person or change any value. Untick what you have not measured: the estimate stays, and the range gets wider. Nothing you enter leaves this page.
Open Physiological Age
What moves the number (years, after overlap)
The standard
A data file anyone can read, cite and implement
OPA is a machine-readable specification. Each signal is scored as the log hazard ratio against an average person of the same age and sex. Overlapping signals are discounted, and the total is converted to years with a Gompertz base of 1.09 (about 9% more mortality risk per year of age).
| Signal | Area | Rule | Evidence | Source |
|---|
Governance (proposed charter)
Scientists decide. Everyone else proposes, advises or funds.
The model will change only through public proposals, evidence review and a vote of an independent Scientific Steering Committee. The founding committee is being formed now. Device makers, the founder's company included, take part on the same terms as anyone else.
Proposal
An OPA Enhancement Proposal: the rule, the evidence, the effect on test persons.
Public comment
Open discussion in the repository.
Evidence review
Graded strong, moderate or limited with a published rubric.
Vote
Two thirds needed to change numbers. Members with a conflict sit out.
New version
Spec, conformance tests and change log, all versioned.
Licences
Open content, protected name
Code
Apache-2.0
Reference implementations and tests, with an explicit patent grant.
Specification
CC BY 4.0
Free to implement and cite, including in commercial products.
Test data
CDLA-Permissive-2.0
Conformance vectors and benchmark tables. Cohort data stays under its own terms.
Only implementations that pass the conformance suite may say "OPA conformant". The full code repository opens with the founding group; the spec and test vectors are available now.
Get involved
Help set the reference
We are inviting a founding group of scientists in aging biomarkers, exercise and sleep epidemiology, biostatistics and data standards. Questions go to bruno@healthspancommons.org .
Scientists
Join the steering committee, lead an evidence working group, or co-author the spec paper.
Cohorts and data holders
Run the validation harness on your data. Results are published; your data never leaves your servers.
Implementers
Build OPA into a device or app, pass the conformance suite, and give feedback through the Implementers Forum.