Biological Age Calculator: Levine PhenoAge From a Blood Test

Where the formula comes from#

In 2018 Morgan Levine and colleagues took data from NHANES III — around 10,000 American adults followed for mortality — and asked which routine clinical measurements, combined how, best predicted who would die. A Cox regression narrowed 42 candidate markers down to nine, each with a weight, plus chronological age.

The result is a weighted sum that gets converted through a Gompertz mortality model into something with an intuitive unit: years. Your PhenoAge is the age at which your calculated ten-year mortality risk would be average.

This matters because it makes the number honest about what it is. PhenoAge is not measuring your cells. It is a risk score wearing the costume of an age.

The nine markers, and what each one is doing#

None of the nine is there because it is a marker of ageing in any mechanistic sense. They are there because they predicted death in the training data.

Marker	What it tracks	Direction
Albumin	Liver synthesis, nutrition, chronic inflammation	Lower is worse
Creatinine	Kidney filtration	Higher is worse
Fasting glucose	Glucose regulation	Higher is worse
CRP	Systemic inflammation	Higher is worse
Lymphocyte %	Immune competence	Lower is worse
MCV	Red cell size; B12, folate, alcohol	Higher is worse
RDW	Variability in red cell size	Higher is worse
Alkaline phosphatase	Liver and bone turnover	Higher is worse
White blood cells	Inflammation, infection	Higher is worse

Two of these carry more weight than people expect. RDW — the variability in the size of your red blood cells — has the largest coefficient of the nine, and it is a marker almost nobody looks at. CRP enters the equation as a logarithm, which means the jump from 0.5 to 5 mg/L matters far more than the jump from 5 to 10.

Reading your result properly#

The calculator shows you the gap between your PhenoAge and your chronological age, then breaks down how much each marker contributes. That breakdown is the part worth your attention.

A number five years above your age is not a verdict on your ageing. It is the sum of your inputs, and in practice it is nearly always dominated by one or two of them. If your gap is +6 years and the breakdown says CRP is adding 5, you do not have an ageing problem — you have an inflammation finding, and inflammation has causes worth identifying.

The reverse is also true and less comfortable. A PhenoAge below your chronological age means your markers are better than average for your age. It does not mean the interventions you have been doing are working, because you have no before-and-after. Single measurements describe where you are, not which way you are moving.

Читайте также: Biological Age Tests: What They Measure, Are They Worth It

Why not just buy a DNA age test#

Because the reliability problem is severe and vendors do not disclose it. Higgins-Chen and colleagues reported in Nature Aging in 2022 that technical noise alone produced deviations of up to nine years between replicate runs of the same sample across six prominent epigenetic clocks. A test whose answer can move by nine years without your body changing at all cannot tell you whether your body changed.

The blood panel has the opposite profile. Its assays are decades old, standardised, run in every hospital, and reproduce within a few percent. It is measuring cruder things — but it is measuring them properly, and every one of the nine values is independently interpretable by a doctor.

Читайте также: The Horvath Clock: What It Is and What the Test Can Tell You

A note on the lifestyle mode#

The second tab estimates an adjustment to your age from eight behavioural and clinical factors, using effect sizes from large cohort studies — smoking from Jha 2013, activity from Moore 2012, BMI from the Prospective Studies Collaboration, alcohol from Wood 2018, social connection from Holt-Lunstad 2010.

It exists because most people asking about their biological age do not have a blood panel in front of them. But it is a different kind of object, and the interface says so: adding independent effect sizes assumes the factors do not overlap, and they overlap heavily. Someone who smokes, drinks heavily and never exercises does not lose the arithmetic sum of the three penalties.

Read the breakdown rather than the total. The rows are ranked by weight, and the top one is the single highest-value thing you could change.

What none of this measures#

Neither mode captures anything about your organs individually, your cognition, your musculoskeletal reserve or your risk of the specific diseases in your family history. For those, the boring measurements still win: blood pressure, lipids, HbA1c, waist circumference, VO2max, grip strength, gait speed, and the age-appropriate screening tests. Each is cheap, reproducible, validated against hard outcomes, and — the part clocks still cannot claim — attached to a treatment that is known to work.