Phenotypic Age Calculator — Biological Age from Blood Tests

Methodology

Phenotypic Age (PhenoAge) is the clinical, blood-based biological-age measure developed by Levine and colleagues (Aging, 2018; originally derived in J Gerontol A, 2013). It combines nine routine blood markers with chronological age into a single estimate of biological age, built to predict differences in mortality risk between people of the same calendar age.

The calculation runs in three stages. First, a weighted linear combination of the markers:

xb = −19.90667 + Σ(coefficient × marker) + 0.08035356 × age

with the published coefficients (each marker in the exact unit its coefficient expects):

Marker	Unit	Coefficient
Albumin	g/L	−0.03359355
Creatinine	µmol/L	+0.009506491
Glucose	mmol/L	+0.1953192
ln(C-reactive protein)	mg/dL	+0.09536762
Lymphocytes	%	−0.01199984
Mean corpuscular volume	fL	+0.02676401
Red cell distribution width	%	+0.3306156
Alkaline phosphatase	U/L	+0.001868778
White blood cells	10⁹/L	+0.05542406
Age	years	+0.08035356

Second, that score is converted to a 10-year mortality estimate under a Gompertz proportional-hazards model. Third, the mortality estimate is back-transformed to the age at which the population-average person carries the same risk — your Phenotypic Age.

We took these coefficients directly from the authors' own published code (the BioAge R package) and locked our implementation to a reproduced worked example, cross-checked two independent ways, before shipping.

A note on CRP units. The formula uses the natural log of CRP in mg/dL. This calculator accepts CRP in mg/L (the common UK unit) or mg/dL and converts internally. A reported value of zero is floored to 0.01 mg/dL so the logarithm stays defined.

Interpreting your result

A Phenotypic Age lower than your chronological age is the favourable direction — at population level it is associated with lower mortality risk. But this is an association across large groups, not a forecast for you: two people with the same PhenoAge can have very different futures.

Treat the number as a conversation-starter with your doctor, not a verdict. It is also a snapshot — a recent infection, injury or flare of inflammation raises CRP and white cell count and will push the estimate up temporarily, then settle. If your result surprises you, the most useful next step is to discuss the underlying blood values with a clinician.

Limitations

• PhenoAge reflects the day's blood panel, not a stable trait — acute illness and inflammation move it.
• It was derived and validated primarily in US (NHANES) data; performance in other populations may differ.
• Nine markers cannot capture everything about ageing; the model is a useful summary, not the whole picture.
• This is a research measure of population risk, not a medical test or a diagnosis. Nothing here should be used to start, stop or change any treatment — that is a decision for you and your clinician.

Sources

Pull the receipts — 3 cited sources

FitTools · sources

1. 01 Levine ME, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging 2018;10:573–591 (clinical Phenotypic Age)
2. 02 Levine ME. Modeling the rate of senescence: can estimated biological age predict mortality? J Gerontol A 2013;68:667–674 (original derivation)
3. 03 Kwon D, Belsky DW. BioAge R package — phenoage_calc.R (authors' reference implementation; source of the exact coefficients used here)

Every formula cited ✓