Skip to content
Homepage
Free · Private · Published formula

Phenotypic Age Calculator (Levine 2018)

Enter nine values from a standard blood count and metabolic panel plus your age, and the calculator returns your phenotypic age and the gap to your calendar age using the exact coefficients Levine and colleagues published in 2018. The formula, its source and its limits are on the page, and nothing you type is sent anywhere.

  • Calculator
  • About 4 min
  • Levine 2018 phenotypic age equation
  • Runs in your browser, nothing is sent
  • Medical review: Medical review pending
Open the calculator

What this calculator works out

Phenotypic age is a way of reading nine ordinary blood markers together with your age as a single number on the age scale. It was built to track mortality risk in a US population survey, not to describe how you feel, and it is one of the few biological age measures with a fully published formula.

  1. 1

    Phenotypic age in years

    The age at which your combination of blood markers would be average in the reference population. A 50-year-old with a phenotypic age of 44 has a blood profile that looked typical for 44-year-olds in the NHANES III survey the formula was built on.

  2. 2

    Age acceleration

    Phenotypic age minus chronological age, in years. A negative number means the blood profile looks younger than the calendar; a positive number means older. This gap is the part most studies follow over time, and it moves when the underlying markers move.

  3. 3

    The published equation, in the open

    Every coefficient, the Gompertz mortality step and the unit conversions are printed on the page and come from the 2018 paper and its supplement. Most online PhenoAge calculators hide the arithmetic, use US units without saying so, or ask for an email before they show the number.

How this compares with a typical online quiz

FeatureTypical free quizRegenerated.com
Published equation namedSometimesLevine, Lu, Quach et al. 2018, with the coefficient table
Units handledOne unit set, often unstatedSI inputs with the US conversion on every field
CRP floor explainedRarelyYes: values under 0.01 mg/dL are floored so the logarithm stays defined
Limits of the model statedRarelyYes: a mortality model from a US survey from the late 1980s to 1994, not a diagnosis
Email required to see the resultOftenNever
Answers sent to a serverOftenNever; scored in your browser, counts-only events

Which labs you need and how to read them off a US report

All nine markers come from two routine panels: a complete blood count with differential (white cell count, lymphocyte percentage, mean corpuscular volume and red cell distribution width) and a comprehensive metabolic panel (albumin, creatinine, glucose and alkaline phosphatase), plus a C-reactive protein test. Many clinics order a high-sensitivity CRP, which is the same protein measured with a more sensitive assay; it works here as long as you convert the unit.

US reports usually print albumin in g/dL (multiply by 10 for g/L), creatinine in mg/dL (multiply by 88.4 for umol/L), glucose in mg/dL (divide by 18 for mmol/L), CRP in mg/L (divide by 10 for mg/dL) and white cells in K/uL or 10^3/uL, which is the same number as 10^9/L. MCV in fL, RDW in percent and lymphocytes in percent are the same on both sides of the Atlantic. If your differential gives lymphocytes as an absolute count, divide it by the total white count and multiply by 100.

Use a fasting sample if you can, because glucose is one of the heavier-weighted inputs, and avoid testing in the week after an infection, a vaccine or a hard training block, which push CRP and white cells up and can add years to the result that have nothing to do with aging.

What moves phenotypic age, and what does not

In the published coefficients, three markers carry most of the weight for a given change: red cell distribution width (0.3306 per percentage point), glucose (0.1953 per mmol/L) and the natural logarithm of CRP (0.0954 per log unit). Albumin and lymphocyte percentage pull the age down as they rise; creatinine, MCV, alkaline phosphatase and white cells push it up. Because the model is additive in these terms, a single abnormal value can shift the result by several years, which is why two draws a few months apart are worth more than one.

The measure does not include blood pressure, lipids, weight, fitness or any genetic information, and it was never trained to respond to a specific intervention. Studies that report phenotypic age falling after a program of exercise, weight loss or dietary change are reporting movement in these nine markers; whether that translates into longer life has not been shown in a randomized trial. Treat the number as a summary of your current blood work, not as a prediction about you personally.

Methodology and sources

The calculator implements the phenotypic age measure published by Levine, Lu, Quach and colleagues in Aging in 2018. The authors took 42 clinical markers from NHANES III, the third US National Health and Nutrition Examination Survey, run from the late 1980s to 1994, and used a penalized Cox regression on 9,926 adults aged 20 and over with up to 23 years of mortality follow-up to select nine markers plus chronological age. The linear combination is converted into a ten-year mortality score through a Gompertz model and then mapped back onto the age scale, so that phenotypic age is the age at which that mortality score would be typical in the training population. The coefficients were fixed in the paper's supplement and are not re-estimated here. Because the model was derived in a US adult sample, with US laboratory methods of the early 1990s, it is less certain in people under 20, in pregnancy, in acute illness and in laboratories using different assays.

The arithmetic runs in three steps. First, the linear term: xb = -19.907 - 0.0336 x albumin (g/L) + 0.0095 x creatinine (umol/L) + 0.1953 x glucose (mmol/L) + 0.0954 x ln(CRP in mg/dL) - 0.0120 x lymphocytes (%) + 0.0268 x MCV (fL) + 0.3306 x RDW (%) + 0.00188 x alkaline phosphatase (U/L) + 0.0554 x white cells (10^9/L) + 0.0804 x age. Second, the mortality score: M = 1 - exp(-exp(xb) x (exp(120 x 0.0076927) - 1) / 0.0076927). Third, phenotypic age = 141.50225 + ln(-0.00553 x ln(1 - M)) / 0.090165. Worked example: a 40-year-old with albumin 45 g/L, creatinine 80 umol/L, glucose 5.0 mmol/L, CRP 0.1 mg/dL, lymphocytes 30 percent, MCV 90 fL, RDW 13 percent, ALP 70 U/L and white cells 6 x 10^9/L gets a phenotypic age of 33.0 years and an age difference of -7.0 years. CRP is floored at 0.01 mg/dL because the logarithm of zero is undefined.

The bands on this page are educational, not published thresholds. In the NHANES IV validation by Liu and colleagues (2018), each one-year increase in phenotypic age above chronological age was associated with roughly a 9 percent higher all-cause mortality hazard, and people whose phenotypic age ran five or more years ahead of their calendar age carried clearly higher risk of death, disease count and physical limitation than those running five or more years behind. We use that five-year gap, in both directions, to split the result into orientation bands and treat a gap over ten years as a reason to review the individual markers with a clinician. Reference ranges for every input differ between laboratories.

Medical review: Medical review pending. Last updated .

References

  1. Levine ME, Lu AT, Quach A, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018;10(4):573-591. doi:10.18632/aging.101414
  2. Liu Z, Kuo PL, Horvath S, Crimmins E, Ferrucci L, Levine M. A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: a cohort study. PLoS Med. 2018;15(12):e1002718. doi:10.1371/journal.pmed.1002718
  3. Levine ME. Modeling the rate of senescence: can estimated biological age predict mortality more accurately than chronological age? J Gerontol A Biol Sci Med Sci. 2013;68(6):667-674. doi:10.1093/gerona/gls233

Phenotypic Age Calculator (Levine 2018) FAQ

It is one of several measures that get called biological age. Phenotypic age is built from nine blood markers; epigenetic clocks such as DNAm PhenoAge and GrimAge are built from DNA methylation and need a specialized test. The Levine paper trained its epigenetic clock on this blood-based phenotypic age, which is why the two share a name. Different measures can give different numbers for the same person, and none of them has been shown in a randomized trial to improve outcomes when used to guide treatment.

Not on its own. The result is a weighted sum of nine markers, so one high value, such as a CRP raised by a recent cold or a glucose drawn after breakfast, can add years. Look at which inputs sit outside your laboratory's reference range, repeat the draw under better conditions, and bring both reports to a clinician. A persistent gap of more than five years is the point at which the research cohorts showed clearly higher risk, and it is worth a proper review.

The formula computes a ten-year mortality probability as an intermediate step, and some calculators print it. We do not, because it is a population average calibrated on a US survey from the late 1980s to 1994 and does not take your blood pressure, lipids, smoking, weight, fitness or history into account. Shown as a personal forecast it would be misleading. The phenotypic age and the age difference carry the same information in a form the research literature actually uses.

Yes. High-sensitivity CRP measures the same protein with a lower detection limit. Convert from mg/L to mg/dL by dividing by 10, so an hs-CRP of 1.2 mg/L is entered as 0.12 mg/dL. A result reported as below the detection limit can be entered as 0.01, which is the floor the calculator applies anyway.

The markers change slowly outside illness, so every six to twelve months is plenty, ideally on the same laboratory with the same preparation. Two readings a few months apart tell you far more than one, because they show whether a gap is stable or was a one-off. Keep the earlier report so you can compare individual inputs, not only the final number.

Yes. Your answers and the numbers you enter are scored in your browser and are never sent to us. They stay in this browser tab so you can return to your result, and closing the tab clears them. The only thing the page sends us is a counts-only event: which test was started or completed, the result band and the headline number, with no answers, no entered values and no account. There is nothing to sign up for.

Talk to a longevity clinic that measures before and after

Browse US clinics offering NAD+ IV therapy and longevity programs. Look for physician oversight, a baseline panel repeated at the same laboratory, and honest copy about what biological age measures can and cannot show. A listing on Regenerated.com is not an endorsement.

Browse longevity and NAD+ clinics

A listing is not a treatment endorsement. Read each clinic's published checks and talk to your clinician before starting any treatment.

Important: This calculator is educational arithmetic on laboratory values you enter, using a published population model. It is not a diagnosis, not a prediction of your lifespan and not a recommendation for or against any treatment. Only a clinician who knows your history and sees the full reports can interpret these markers. Reference ranges differ between laboratories, and two readings beat one. If your symptoms are severe or sudden, seek urgent medical care.