StemCellAtlas

Biological-age tests compared

“Biological age” can mean different measurements. Clinical Phenotypic Age uses blood biomarkers and your age; DNAm PhenoAge and the original Horvath clock use DNA methylation. DunedinPACE estimates a rate of aging, not an age in years. Check the named algorithm before comparing two numbers.

DunedinPACE development study · Original Horvath clock study · Phenotypic Age and DNAm PhenoAge study

Four methods, different inputs and targets

Swipe or scroll sideways to compare all columns.

Four methods, different inputs and targets
MethodInputsWhat the model estimates
Clinical Phenotypic AgeNine clinical biomarkers plus chronological ageA mortality-risk-based age estimate, in years
DNAm PhenoAgeDNA methylation at 513 sitesA methylation estimate of Phenotypic Age
Original Horvath clockDNA methylation at 353 sitesAge from a model trained on chronological age
DunedinPACEBlood DNA methylation at 173 sitesPace of aging, not age in years
DunedinPACE development study · Original Horvath clock study · Phenotypic Age and DNAm PhenoAge study

PhenoAge does not always mean a DNA test

The original paper has two steps. First it combines nine clinical biomarkers with chronological age to estimate Phenotypic Age. It then trains a DNA-methylation model to predict that measure. The second model is DNAm PhenoAge. A report that omits “DNAm” may be referring to the clinical calculation, so ask which inputs were actually used.

Phenotypic Age and DNAm PhenoAge study

Why a single blood sample can represent a rate

The DunedinPACE developers first modelled change in 19 organ-system indicators measured at ages 26, 32, 38 and 45. They then trained a DNA-methylation algorithm to estimate that longitudinal pattern from blood. The rate interpretation comes from its training target, not from taking several blood samples every time the test is used.

DunedinPACE development study

What to look for on a test report

Identify the named algorithm, the sample used and the units of the output before comparing two reports. The paper makes clear that DunedinPACE complements other methylation clocks rather than being the same measurement under another brand. A lower number across two different algorithms is not automatically an improvement.

DunedinPACE development study
A sample tube in a metal rack beside a cotton swab and sleeve
AI-generated sample-collection illustration; not a branded test kit.

Use the table and its sources

Download table CSV · Download table JSON · Dataset notes

Related reading

Source-linked editorial synthesis, prepared with AI assistance. This is not a clinician review or a personalised treatment recommendation. Corrections: editorial standards.