This site and its domain are for sale. Email the owner if you want to talk about it.

bioclocktest.com

The Biological Age Test Buyer's Guide

What to ask before you pay for a biological age test: which clock it uses, its repeat-test noise, what it has been validated against, and how to read your result without overreacting. Twelve questions, one page.

A biological age test takes a sample of you, runs it through a formula, and returns a number that usually differs from the one on your driver’s license. The formula is the product. Different formulas were trained to predict different things. They give different answers on the same blood, and some give different answers on the same blood twice. Most companies selling kits do not say which formula they use, which is itself a kind of answer.

This guide is twelve questions in four tables: the clock behind the number, how noisy it is, what it has been validated against, and how to read what comes back. Each row gives the question, why it matters, and what the published evidence currently says, with the source named. Where the evidence does not settle something, the cell says so, and that happens more often than the marketing would suggest.

What is not settled is fairly basic. As of 2026 the field has no reference standards, so the same sample sent to different validated labs will not reliably produce the same age (The Scientist). No direct-to-consumer clock is approved by the FDA as a diagnostic tool. One AMA Journal of Ethics commentary calls biological age a fiction, while also noting that the same tests are fairly reliable indicators of risk for certain age-related diseases. Both can be true at once, which is roughly where the science sits.

The clock behind the number

Question to askWhy it mattersWhat the evidence says
1. Which clock does the test use?Horvath, PhenoAge, GrimAge and DunedinPACE were trained to predict different things, so their numbers answer different questions.First-generation clocks such as Horvath were trained to predict chronological age by comparing DNA methylation in young and old people (The Scientist). PhenoAge and GrimAge were trained to predict clinical phenotypes, disease incidence and all-cause mortality (CD Genomics). DunedinPACE was trained on differences in multi-organ deterioration between people of the same age (The Scientist). Most kit companies do not detail which clock they use (The Scientist).
2. Does it report one age, a pace, or a set of organ ages?A static age, a rate of aging and a panel of organ scores are different quantities and are not read the same way.DunedinPACE monitors the rate of aging rather than a static biological age (CD Genomics) and was explicitly trained on longitudinal change rather than cross-sectional age prediction (Moffitt and Caspi lab). Its predecessor, DunedinPoAm, was designed as a single-time-point blood measure of pace; the preliminary demonstration was published by Belsky et al. in 2020 (eLife). Index by Elysium Health reports ages for nine body systems; TruDiagnostic’s SymphonyAge measures 11 organ systems. How well organ-level scores are validated: not settled.
3. What does it actually measure?Methylation, blood chemistry and protein clocks are different instruments with different evidence behind them.Consumer tests variously measure metabolic markers, DNA, or blood proteins. Blood-chemistry clocks typically use glucose, LDL cholesterol, triglycerides, sex hormones, GGT and white blood cells. InsideTracker’s InnerAge 2.0 weights glucose above GGT because glucose correlates more strongly with age in the company’s own dataset. Index is based on DNAm PhenoAge. TruDiagnostic’s OMICmAge is a multi-omic-informed methylation clock. A proteomic clock (PAC) built on 53,021 UK Biobank participants and 2,923 proteins showed stronger age-adjusted associations with all-cause mortality and 18 incident diseases than PhenoAge, BioAge or telomere length.
4. What tissue was the clock trained on, and what tissue does the kit collect?A clock applied to a tissue it was not built for is being asked an unfamiliar question.Almost all clocks in commercial tests are trained on blood (The Scientist). Hannum uses 71 CpG sites and performs well specifically in blood; the Horvath pan-tissue clock uses 353 sites with a mean absolute error of about 3.6 years across tissues (CD Genomics). The skin clocks MitraSolo and MitraCluster work from tape-stripped facial samples with an error of about 4 years. Whether a given kit’s sample type matches its clock’s training tissue: for most kits, not public.

Noise and repeatability

Question to askWhy it mattersWhat the evidence says
5. How much does a repeat test move on the same sample?Some clocks give the same sample results years apart. A change smaller than the noise is not a change.Technical replicates of the published Horvath, Hannum and PhenoAge clocks deviated by three to nine years, and one sample processed repeatedly can vary by up to nine years (Elysium Health). Elysium reports that its Index agrees across technical replicates within zero to 1.5 years; that figure comes from the company. Replicate figures for most other kits: not public.
6. How much does it move between samples taken hours apart?Lab precision is not the same as stability in a living person who has just eaten lunch.In an evaluation of 18 clocks on the TranslAGE platform, biological reliability across samples taken hours apart, before and after meals and under acute stress was markedly lower than technical reproducibility for most clocks, and most showed only moderate stability. PCGrimAge was the only clock with good biological reliability (ICC above 0.75). Technical robustness did not predict biological reliability.
7. Would a different lab give the same number?Without reference standards, a result belongs to the lab and algorithm that produced it, not to the person.As of 2026 there are no reference standards, so the same sample sent to different validated labs does not yield the same readout (The Scientist). Cross-lab agreement for any named consumer kit: not public.

Validation

Question to askWhy it mattersWhat the evidence says
8. What has it been checked against?A clock that predicts disease and death tells you more than one that only predicts your birthday.Clocks trained purely on chronological age are better at capturing normal aging than aging tied to disease risk or mortality (CD Genomics). The AMA Journal of Ethics commentary says epigenetic clocks are fairly reliable indicators of risk for certain age-related diseases. In the Add Health study of younger U.S. adults, social and lifestyle factors were more often associated with PhenoAge, GrimAge and DunedinPACE than with Horvath.
9. How close does it get to calendar age, and is that the point?Low error against chronological age is how first-generation clocks are scored, and it is not the same as predicting health.Reported mean absolute errors include about 3.6 years for Horvath across tissues, 2.98 years for a deep learning NHANES model, 3.58 years for an ensemble NHANES model, and 3.149 years for a sequence-graph method on 3,707 blood samples. A proteomic clock correlated with chronological age at r = 0.92 in 3,977 Chinese and r = 0.94 in 1,990 Finnish biobank participants. Researchers treat lifespan and healthspan as separate outcomes; which one a given kit was tested against: often not public.
10. Was it validated in a population like the one being tested?Accuracy can shift with age, health status and blood cell composition.In a Pakistani cohort, prediction accuracy fell with increasing age. In people living with HIV, first-generation clocks show roughly 5 to 7 years of extra biological age, which may be confounded by HIV-related changes in immune-cell composition. The PhenoAge validation sample included 185 oldest-old adults top-coded at age 85. Validation across other groups for most consumer kits: not settled.

Reading the result

Question to askWhy it mattersWhat the evidence says
11. What is known to move it?A number that changes should change for a documented reason, and different clocks respond to different things.Starting antiretroviral therapy partly reversed epigenetic age acceleration on Horvath, Hannum, GrimAge and PhenoAge over 2 years, though PhenoAge and GrimAge stayed elevated compared with HIV-negative controls (NEAT001/ANRS143). Long-term caloric restriction in the CALERIE trial did not significantly change PhenoAge or GrimAge. Prenatal and childhood lead exposure was associated with accelerated aging in males and decelerated aging in females on several clocks, measured at a mean age of 14.5. The cortisol/DHEAS ratio correlated with Hannum, Horvath Skin and Blood and PhenoAge, but not with Horvath or GrimAge.
12. What can the number not tell you?The result is a model’s estimate, not a diagnosis, and the market around it is lightly policed.No direct-to-consumer clock is FDA approved as a diagnostic; they are sold as wellness or informational products. The AMA Journal of Ethics commentary argues the tests cannot say how old someone really is, calls on companies to be more transparent about reliability and limitations, and warns of psychological harm and of results making age discrimination look justifiable. Tony Wyss-Coray of Stanford described the market as “a bit like the wild West,” noting tests coupled with products never tested in a human clinical trial.

How to read this

Read the tables in order. Questions 1 to 4 establish what the number is: which clock, which kind of output, which molecules, which tissue. Questions 5 to 7 establish how much the number would wobble if nothing about the person changed, which sets the minimum gap worth noticing between two results. Questions 8 to 10 establish whether the clock predicts anything beyond the calendar, and for whom. Questions 11 and 12 cover what has actually been shown to shift these numbers and what they cannot say. A result from a clock whose identity, noise and validation are all not public is a number with very little attached to it. Where a figure comes from the company selling the test, the row says so.

What would change it

The biggest single change would be reference standards, so that one sample gives one age regardless of lab. After that: kit companies routinely naming their clock and publishing both technical and biological repeatability; more studies like the 18-clock TranslAGE evaluation, extended to consumer products; validation against disease and mortality in populations beyond the original training cohorts; any FDA decision treating a clock as a diagnostic; and further trials, like CALERIE and NEAT001/ANRS143, that test whether an intervention moves a clock and whether that movement tracks real health outcomes. Several rows above would then move from “not public” to a number.

Sources

  1. The Sociodemographic and Lifestyle Correlates of Epigenetic Aging in a Nationally Representative U.S. Study of Younger Adults
  2. Epigenetic Clocks Power Biological Age Data, But Why Do Results Differ Across Tests? | The Scientist
  3. What Are the Most Ethically Salient Implications of Epigenetic Age Testing? – Dec 2025 | Journal of Ethics | American Medical Association
  4. Biological age versus chronological age: What the science says
  5. The Science Behind Index | Elysium Health
  6. Four Aspects to Interpret Epigenetic Clock Results - CD Genomics
  7. Choosing Epigenetic Clock Models: Horvath, Hannum, PhenoAge, GrimAge and Beyond - CD Genomics
  8. Epigenetic age predictors for non-invasive assessment of human skin
  9. Why You Need a Biological Age Blood Test -...
  10. Epigenetic age acceleration changes 2 years after antiretroviral therapy initiation in adults with HIV: a substudy of the NEAT001/ANRS143 randomised trial.
  11. Elysium Health Launches At-Home Test to Measure Rate of Aging, Biological Age
  12. Prenatal and childhood lead exposure is prospectively associated with biological markers of aging in adolescence
  13. About TruAge
  14. [2605.10541v1] Bridging Sequence and Graph Structure for Epigenetic Age Prediction
  15. Effect of long-term caloric restriction on DNA methylation measures of biological aging in healthy adults from the CALERIE trial
  16. Cortisol, DHEAS, and the cortisol/DHEAS ratio as predictors of epigenetic age acceleration
  17. DunedinPACE, a DNA methylation biomarker of the pace of aging | eLife
  18. Frontiers | Development and validation of deep learning- and ensemble learning-based biological ages in the NHANES study
  19. Epigenetic clocks and accelerated biological aging in people living with HIV: Emerging mechanisms and clinical implications
  20. Are Epigenetic Clocks Ready to Determine Your Biological Age?
  21. Biological versus Technical Reliability of Epigenetic Clocks and Implications for Disease Prognosis and Intervention Response
  22. Proteomic aging clock (PAC) predicts age-related outcomes in middle-aged and older adults
  23. DunedinPACE | Moffitt & Caspi: Genes, Environment, Health, Behavior
  24. Biological versus Technical Reliability of Epigenetic Clocks and Implications for Disease Prognosis and Intervention Response
  25. Quantification of the pace of biological aging in humans through a blood test, the DunedinPoAm DNA methylation algorithm | eLife
  26. TruDiagnostic™ Biological Age Test
  27. Chronological age estimation using CpG methylation signatures in Pakistani population
  28. A new aging measure captures morbidity and mortality risk across diverse subpopulations from NHANES IV: A cohort study | PLOS Medicine
  29. Proteomic aging clock predicts mortality and risk of common age-related diseases in diverse populations

Bio Clock Test explains published research. It is not medical advice and it does not diagnose anything. We never recommend a test, a supplement or a treatment. Talk to a clinician about your health.