Skip to main content
Evidence Based Longevity (EBL)Check what I take

PhenoAge: a biological-age estimate from nine routine blood tests

PhenoAge, also called Phenotypic Age, was published by Morgan Levine and colleagues in 2018. It combines nine results found on a standard metabolic panel and complete blood count with your age, and gives the age at which an average adult in a large US national survey carried the same estimated 10-year mortality risk. No special test is needed. This page runs the equation exactly as the paper’s supplement prints it, and the arithmetic happens in your browser.

An estimate, not a lifespan prediction.
Biological-age results, from epigenetic clocks A test that estimates biological age from chemical marks on DNA (DNA methylation). to telomere length, are statistical estimates from models built on population data. They do not predict how long any one person will live. Faster epigenetic aging was associated with higher mortality across 13 cohorts, an association in groups of people (Chen 2016). Technical noise alone moved six widely used clocks by up to 9 years between repeat runs of the same sample (Higgins-Chen 2022), and different clocks and companies give different numbers. No trial has yet shown that lowering a clock reading extends life (Waziry 2023).
PMID 27690265 Chen 2016, Aging PMID 36277076 Higgins-Chen 2022, Nat Aging PMID 37118425 Waziry 2023, Nat Aging
Conflict of interest, from the papers' own statements: the Higgins-Chen clock versions are licensed to a company that sells a biological-age test, and DunedinPACE An epigenetic test that estimates how fast a person is aging now, in years of biological aging per calendar year., measured in the CALERIE analysis, is licensed to a commercial company.

The calculator

The nine results are on most annual blood panels: the comprehensive metabolic panel (albumin, creatinine, glucose, alkaline phosphatase), the complete blood count with differential (lymphocyte percent, mean cell volume, red cell distribution width, white cell count) and a C-reactive protein test. The equation was built on fasting glucose.

How it works

The method and its sources

The equation is Phenotypic Age from Levine ME et al., Aging (Albany NY) 2018;10:573-591 (PMID 29676998), Table 1, Supplementary Table S1 and the supplementary methods. The authors started from 42 blood markers and age in 9,926 adults in the third US National Health and Nutrition Examination Survey (NHANES III), followed for deaths for up to 23 years, and kept the nine markers and age that best predicted mortality. A Gompertz survival model turns those ten numbers into a 10-year mortality score, and a second step converts the score into years.

It was then checked in a separate national sample, NHANES IV, in Liu Z et al., PLoS Med 2018;15:e1002718 (PMID 30596641): 11,432 adults aged 20 to 84, followed for about 12.6 years. Each year of PhenoAge above what was typical for a person’s age went with about 9% higher all-cause mortality, and the association held across age groups, race and ethnicity, education, smoking and drinking, and in adults with no reported disease and a normal BMI. These are associations in groups of people.

US units are converted to the equation’s units as follows: albumin g/dL × 10 = g/L; creatinine mg/dL × 88.4 = µmol/L; glucose mg/dL × 0.0555 = mmol/L; CRP mg/L ÷ 10 = mg/dL, of which the equation takes the natural logarithm. The result shows each value both ways, so a unit slip is easy to see.

A note on versions. The equation reprinted in the 2018 follow-up paper differs from the original supplement in three numbers (the glucose and red cell distribution width weights and one constant) and leaves out the survival-model step. Copies of that version circulate online. This page uses the original supplement throughout, and the build checks every number on each commit.

PhenoAge is a different thing from DNAm PhenoAge, the epigenetic clock in the same paper, which is read from DNA methylation and sold as a lab test. This page computes the clinical version only.

What it leaves out: the nine markers were chosen because they predicted death in a national sample, not because changing any one of them is known to change anything. No trial has shown that moving a PhenoAge reading changes how long people live.

What happens to your data

Nothing you type leaves this device. The numbers are worked out in your browser, nothing is sent to EBL or anyone else, and nothing is saved: closing the page clears them.

Build 9b19d78, November 2026