Biological Age Testing Services Compared: A 2026 Buyer’s Framework

Question One: Which Clock, and Which Generation?

This is the single most informative question and a surprising number of services will not answer it plainly.

First generation clocks such as Horvath and Hannum were trained to predict chronological age. They are accurate at that and comparatively weak at predicting health outcomes.

Second generation clocks such as PhenoAge and GrimAge were trained on clinical markers and mortality. They predict outcomes considerably better and are the ones most worth reporting.

Third generation, principally DunedinPACE, estimates rate of ageing rather than accumulated age, which is the most decision-relevant of the three.

Proprietary clocks are the category to scrutinise. A service that reports its own unpublished model is asking you to accept validation you cannot inspect. Some are built on sound methodology. You have no way of telling which.

A good answer names the published clock and cites the paper. A poor answer is "our proprietary algorithm based on the latest research".

Question Two: Was It a Principal-Component Version?

Standard methylation clocks read individual CpG sites, and individual site measurement is noisy. Test-retest on the same blood sample can differ by two to three years, which is larger than most changes anyone hopes to detect.

Principal-component versions of the major clocks were developed to address exactly this, aggregating across many correlated sites to reduce measurement error substantially. They are the current methodological standard for anything intended to be repeated.

If a service does not use them and does not mention reliability at all, the number it returns to you carries an uncertainty band it is not showing you. Ask what the test-retest reliability is. A service that has measured it will tell you.

Question Three: What Is Being Sampled, and How?

Cell composition is the recurring confounder. A blood sample is a mixture whose proportions shift with recent infection, stress and time of day. Good services adjust statistically for estimated cell counts and say so.

Consistency matters more than the choice itself. Comparing a saliva result to a blood result, or one laboratory's output to another's, produces a difference driven by method rather than by biology.

Questions Four Through Seven

Four: is the laboratory accredited, and named?

A CLIA-certified or equivalently accredited laboratory should be named. Services that decline to say where the assay runs are asking for trust without offering the basis for it.

Five: does the report show uncertainty?

A single number with no confidence interval implies a precision the method does not have. A report that says "biological age 42, plus or minus 3" is being more honest than one that says 42.

Six: what does the recommendation engine sell?

If the test result routes to the same company's supplement line, the incentive is structural. It does not make the test invalid, and it does mean the interpretation layer is not independent of the sales layer. This is worth knowing before rather than after.

Seven: what happens to the data?

Methylation data are among the most identifying and most inferentially rich biological data a consumer can generate, carrying information well beyond an age estimate. Read the retention terms, the research consent, the third-party sharing clauses and the deletion path. In most jurisdictions this sits outside the protections that apply to clinical records.

What a Test Cannot Tell You

Consumer biological age tests are not diagnostic, and none is validated to indicate whether a specific intervention is working for a specific person. The clocks are validated as population-scale associations, and that is a different claim from individual guidance.

Short-interval retesting is the most common way people waste money on this category. Measurement noise exceeds real biological change over a few months, so a three-month follow-up largely measures the assay. Annual is the shortest interval likely to be informative, and even then a single change should be held loosely.

Cross-service comparison is the second most common error. Different clocks, laboratories, sample types and normalisation pipelines produce different numbers for the same person on the same day. A difference between two services is a difference between two methods.

A Reasonable Approach

If you are curious. One test from a service that names a published second- or third-generation clock, uses a principal-component version, runs in a named accredited laboratory and shows uncertainty. Read the result as a baseline and nothing more.

If you intend to track. Commit to one service, one sample type and one annual timing, and change none of them. Consistency is worth more than choosing the best available option and then switching.

If the budget is limited. A standard metabolic and inflammatory panel, a cardiorespiratory fitness assessment and honest sleep and activity tracking will inform decisions more than any epigenetic test currently sold, and cost less. This is not a fashionable answer and it is the correct one for most people.

The market will keep producing new entrants. The questions above will keep separating them.