Is a biological age test worth paying for in Hong Kong? Repeats of one sample 9 years apart, the prices actually published, and the disclaimers the sellers wrote themselves, 2026

TL;DR This is not an article about how to live longer. It is here to help you decide one thing: whether to pay for a "biological age" test. The strongest line on the subject was not written by a critic. It was written by the seller. Bryan Johnson, the most energetic promoter of these tests anywhere, states at the foot of his own protocol site that measurements and claims of this kind are "preliminary", and that "These tests are experimental and intended solely for research purposes. They should not replace or supplement any clinical tests recommended by licensed medical professionals." The most damaging technical problem is repeatability: Higgins-Chen et al. (Nat Aging 2022) found that when the same DNA sample is measured repeatedly on six prominent epigenetic clocks, technical noise alone can produce deviations of up to 9 years. On the telomere side, a blinded ring trial across 10 laboratories using 3 methods (Martin-Ruiz et al., Int J Epidemiol 2015) found that absolute values from different laboratories cannot be compared directly at all, and that reference ranges cannot be shared between laboratories (the same paper also states plainly that the three methods do not differ in accuracy — the problem is not which method, it is cross-laboratory comparison itself). A number that changes when you measure it again is not telling you about your body. On the Hong Kong market: not one of the suppliers this article found names the clock it uses (Horvath, GrimAge, PhenoAge and DunedinPACE appear nowhere). Only two biomarker products were found with a first-hand published price: the telomere length test from MedDx at $ 3800 (analysed by GC Genome in Korea), and ATML Healthspan's "biological age blood test" at $1,850.00 (with no statement of what is measured). Both are written with a bare $ sign and no "HK" prefix. Separately, the itemised contents of OT&P Healthcare's HK$19,800 "Longevity health check" contain no epigenetics, no methylation and no telomeres at all — the list opens with grip strength and VO2 Max. The misreading to guard against most: a result saying your biological age is below your chronological age is not a certificate of health, cannot be used for reassurance, and certainly cannot be used to stop other tests. The reverse holds too: a result saying you are "9 years older" is not a diagnosis. Not one of the tests cited here has been validated to rule out any disease. This article reports only published first-hand material and published research. It neither recommends nor disparages any supplier, and it will not tell you how many suppliers exist in Hong Kong — the reason is below: the list of suppliers found here is not complete. For what actually does have evidence behind it, see 〈What the evidence for longevity science actually shows〉.

#What is "biological age", and why is anyone selling it?

"Biological age" is not something that can be measured directly the way height and weight can. It is an estimate produced by some formula from some set of indicators — change the formula or change the sample, and the number changes.

Start with why anyone would want to measure it. The World Health Organization's "Ageing and health" fact sheet describes ageing this way: at the biological level, it results from the accumulated impact over time of a wide variety of molecular and cellular damage; these changes are neither linear nor consistent, and they are only loosely associated with a person's age in years; there is no typical older person, and some 80-year-olds have physical and mental capacities similar to many 30-year-olds [Note 1].

That is to say: there really is a gap between your age in years and the actual state of your body, and the gap is real. The whole biological age industry is built on the offer to quantify that gap into a single number. The question is what the rest of this article is about: how well that number is measured, whether it repeats, and what it has to do with any decision you might make.

Three words that constantly get run together need separating first, because among the suppliers found here they refer to three different things.

Chronological age — the number of years since you were born.

Biological age, in Chinese either 生物年齡 or 生理年齡 — an estimate derived from biomarkers. ⚠️ Note: there is no single agreed Chinese term on the Hong Kong market. OT&P Healthcare's Chinese page uses the three characters 生理年齡 (the Chinese FAQ heading reads 「生理年齡和實際年齡有什麼區別？」), while the English page puts the same question as "What is the difference between biological and chronological age?" One thing, two Chinese words.

"Body age" or "Bio Age" — in Hong Kong this is also used for a physical fitness assessment involving no blood and no DNA at all. SPORTS LAB HK LIMITED's "Bio Age Test" is $800 for a single session — ⚠️ note that this price sits on that page beneath the words 「限時優惠」, a limited-time offer by the page's own description, not a standing list price; the page also carries no effective or updated date of its own, so whether the figure still holds is something this article has no information on. The page lists seven outputs, in order: body composition analysis, physical fitness performance, core muscle performance, actual body age, quality of life index, joint injury risk, and training recommendations — no blood draw, no DNA. This is a different category of product from the two above and should not be spoken of in the same breath as epigenetic or telomere testing.

⚠️ If you see an advertisement that says "biological age" without saying what is measured: you do not yet know whether it is blood, DNA or press-ups, so you are in no position to judge whether the number is accurate. The first question is always: what are you measuring?

#How do epigenetic clocks work? Three generations measure different things

An "epigenetic clock" takes the degree of methylation at many thousands of positions on the DNA in a tissue or cell as its input, and puts them through a statistical model to produce a number. The key is not the word "methylation" — it is what answer the model was trained against in the first place.

First generation: the training target was chronological age. Horvath (Genome Biol 2013) built the first cross-tissue predictor from 82 methylation array datasets, 8,000 samples in total, covering 51 healthy tissue and cell types, ending with 353 CpG sites. His own conclusion: DNA methylation age measures the cumulative effect of an epigenetic maintenance system [Note 2]. The defining property of a first-generation clock is that it was trained against chronological age — which is to say that a first-generation clock is at its best when it tells you back the number on your identity card.

Second generation: the training target changes to clinical outcomes. DNAm PhenoAge, from Levine et al. (Aging 2018), states plainly that the first generation of epigenetic biomarkers of ageing were developed using chronological age as a surrogate for biological age, and hypothesises that incorporating composite clinical measures of phenotypic age may identify novel CpG sites [Note 3].

DNAm GrimAge, from Lu et al. (Aging 2019), is more particular still: it is a composite of methylation-based surrogates for seven plasma proteins plus a methylation-based estimator of smoking pack-years [Note 4]. In one sentence: a substantial part of what GrimAge measures is the methylation trace that smoking leaves on DNA.

Third generation: DunedinPACE measures a speed, not an age. This gets stated wrongly in advertising and in news coverage all the time. Belsky et al. (eLife 2022) used the New Zealand Dunedin birth cohort born in 1972–73, across four time points spanning two decades and 19 organ-system integrity indicators, to build a "Pace of Aging" first, and then compressed it — using elastic-net regression and a dataset restricted to exclude probes with low test-retest reliability — into an indicator obtainable from a single blood draw [Note 5].

In other words: DunedinPACE tells you "how fast you are ageing now", not "what age you are equivalent to". It is not the same kind of output as a first-generation clock, and two quotations for a "biological age" cannot be set side by side and compared on price as though they were. And its provenance is worth remembering: a single birth cohort, one country, one band of birth years.

⚠️ One point of fairness has to be made, and precisely because it runs the opposite way to the next section it matters more, not less: the first sentence of the results section of that same abstract reads "DunedinPACE showed high test-retest reliability, was associated with morbidity, disability, and mortality". Which is to say that the "up to 9 years" in the next section is about the other six clocks and cannot automatically be treated as DunedinPACE's figure; but since that sentence is the paper's own reported result, and since not one Hong Kong supplier found here states that it sells DunedinPACE, it remains for a Hong Kong buyer something to know rather than something to act on.

⚠️ Belsky 2022's public record carries a competing-interests declaration, and the weight of it is in the second half: DB, AC, DC, KS, RP and TM are listed as inventors on a Duke University and University of Otago invention that has been licensed to a commercial entity; the same declaration records the remaining authors as having no competing interests declared [Note 6].

⚠️ If you see promotional copy saying "we use the newest third-generation clock": the third generation is not "an upgraded first generation with a more accurate number" — what it outputs is a different thing altogether.

#Telomere testing is a separate matter — and the argument has run from 2013 to today

Telomeres are the protective structures at the ends of chromosomes, and they shorten as cells divide. Telomere length is a measurement entirely independent of the epigenetic clocks, and its standing as a marker of ageing has never been settled academically.

The review by Sanders and Newman (Epidemiol Rev 2013) puts it bluntly: conflicting data have generated heated debate about the value of leukocyte telomere length as a biomarker of overall ageing [Note 7]. Their conclusion links shorter telomeres to "older age, male gender, Caucasian race, and possibly atherosclerosis", but the associations with other markers of health are equivocal.

That was the 2013 assessment. In a large 2025 study comparing 14 clocks, DNAm-estimated telomere length was also among those with the smaller effects: smaller effect sizes were noted for PhenoAge (P = 2.9 × 10⁻³) and DNAm telomere length (P < 6.2 × 10⁻⁶) [Note 8].

⚠️ Of the two biomarker products with published prices found in Hong Kong, the more expensive one ($ 3800) measures precisely this — telomere length. The repeatability problems in the next section come in two kinds, one for telomeres and one for the epigenetic clocks, and they are different problems: for the clocks, the same sample measured twice gives different answers; for telomeres, the main problem is that different laboratories cannot be compared at all — but it does not stop there, since there is technical variation within a single laboratory too, and the figures are in the next section.

#Repeatability: measure the same sample twice, do you get the same answer?

This is the pivotal section of the article. A number that changes on remeasurement has told you nothing about your body — and repeatability is exactly the part consumer marketing mentions least.

#Epigenetic clocks: the same DNA, up to 9 years apart

Higgins-Chen et al. (Nat Aging 2022, with Levine ME as corresponding author) state it directly: epigenetic clocks are widely used ageing biomarkers calculated from DNA methylation data, but this data can be surprisingly unreliable; they show that technical noise produces deviations of up to 9 years between replicates for six prominent epigenetic clocks, limiting their utility [Note 9].

The same paper also offers a fix: process the data at CpG level into principal components first, then retrain the six clocks, and agreement between most replicates comes within 1.5 years.

⚠️ That fix carries a competing-interests declaration of its own, and it bears directly on this article's subject. The paper's public declaration states that MEL and AHC have built epigenetic ageing metrics involving the technology described in the manuscript, and that these metrics are licensed by Elysium Health through Yale University; the same passage declares that Elysium Health provided the matched blood and saliva replicate datasets used in the study, and that MEL was previously a scientific adviser to the company and received consulting fees [Note 10].

⚠️ But the half-sentence that follows has to be read with it, because it bounds the relationship: the company otherwise did not fund the study and did not play a role in conceptualisation, design, decision to publish or preparation of the manuscript. And the same passage has two further sentences, without which a reader would assume that no one besides MEL had a commercial connection — the original is explicit: THS was previously an employee of Elysium Health, Inc.; AHC received consulting fees from FOXO Technologies, Inc. for work unrelated to the manuscript [Note 10]. Only the last sentence of that passage says that all other authors report no biomedical financial interests or potential conflicts of interest.

Which is to say: the paper that identified the repeatability problem has more than one author with a financial connection to a commercial body — three authors and two companies in all, not simply the "one person, one company" that appears on the surface. That does not invalidate the "up to 9 years" figure, which is the paper's own measured result, but a buyer is entitled to know it, because the authorship of the "fix" that follows has wider commercial ties than it first appears.

In other words: the problem has a solution, but it only counts if your supplier has used it. And the paper does not state which version any commercial supplier uses — the original or the principal-component retrained one. So the buyer's question is a very specific one: which clock do you use, is it the principal-component version, and what is the test-retest gap in years? A supplier that cannot answer has not told you how repeatable that number is.

(The problem has a history. The reason DunedinPACE excluded low-reliability probes at the modelling stage is that Sugden et al. (Patterns 2020), repeating measurements on 350 blood DNA samples across two Illumina chips, found that the reliability of each of these measurements is not equal, and that unreliable probes were less replicable and generated an unknown volume of false negatives [Note 11].)

#Telomeres: even cross-laboratory comparison does not work

The corresponding work on telomeres is blunter still. Martin-Ruiz et al. (Int J Epidemiol 2015) organised an international blinded ring trial: 10 laboratories, 3 methods (Southern blotting, STELA, qPCR), 10 human DNA samples per round, two rounds. The full original is at [Note 12]: absolute results from different laboratories differed widely and could not be compared directly, though rankings of relative telomere length were highly correlated (correlation coefficients of 0.63 to 0.99); inter-laboratory coefficients of variation averaged about 10% for Southern blotting and STELA and more than 20% for qPCR; this difference was compensated for by a higher dynamic range for the qPCR method; gel-based and PCR-based techniques were not different in accuracy; and intra- and inter-laboratory technical variation severely limits the usefulness of data pooling and excludes sharing of reference ranges between laboratories.

⚠️ Those clauses have to be read together — the "more than 20%" cannot be lifted out on its own. The paper says immediately afterwards that the higher coefficient of variation for qPCR is offset by its larger dynamic range, and its conclusion is that gel-based and PCR-based techniques do not differ in accuracy. So this paper is not saying qPCR is especially poor. What it says is that absolute values and reference ranges cannot be shared across laboratories — and that holds for all three methods alike.

In one sentence: laboratory A saying your telomeres are 8.2 kb and laboratory B saying 7.1 kb does not mean they shortened in between — it may simply mean you changed laboratory. And a "compared with people your age" reference value, on this paper's conclusion, cannot be shared between laboratories at all.

⚠️ The same paper also measured the "stay with one laboratory" case, and it is not zero. Measured as the median of intra- and inter-batch coefficients of variation, technical variation per laboratory ranged from 1.4% to 9.5%, with differences between laboratories only marginally significant (P = 0.06) [Note 12]. That figure has to be read alongside the conclusion above: the conclusion sentence says intra- and inter-laboratory — both count. So the inference that "sticking to the same laboratory makes it fine" is not one this paper supports.

⚠️ Two things must travel with this quotation: (i) the paper carries an erratum of its own (Int J Epidemiol 2015;44(5):1749-54) and drew four commentaries in the same issue, so it is contested literature; and (ii) this was a research-use ring trial in 2015, not performance data on commercial testing in 2026. Whether consumer telomere testing in 2026 does better or worse is something this article has no information on.

#What do the suppliers themselves say? The same thing

This is not an outsider's criticism. MedDx's own sample report for its telomere test says it (page 5, "Limitation", in English — the whole 8-page PDF is English only). That box lists four items, three of which are: telomere length may vary slightly depending on the test method and test conditions; the bias of the result is due to the value of the telomere length of the reference gene used in each test; and the bio-age result of this test is calculated using internally constructed data of the age-specific population group, with the accumulated data regularly updated [Note 13]. The fourth item — that a telomere test cannot be used to diagnose disease — is below.

The peer-reviewed round-robin and the seller's limitations page are saying the same thing: this number is tied to one particular laboratory and one particular set of internal reference data. That is the most concrete point in this article.

#Do the arithmetic: how many years is that ± sign on the report worth?

First, what the ± sign is — no guessing required, because page 1 of that report says so. The first item of the "Interpretation" box on page 1 starts with a sentence that has to be read alongside it: telomeres naturally shorten with age, but the initial length differs among individuals, and the rate of shortening may vary depending on lifestyle, stress and disease [Note 14]. The item immediately after is the ± sign itself: the telomere length is 8.2 kb (margin of error: ±0.3kb), which is longer than the average length of the same age group (±1 year) [Note 14]. Page 2 then converts that same ± into an interval and states its source as the error of the test method itself: considering the error of the test method, the telomere length is from 7.9 kb to 8.5 kb [Note 14].

Which is to say that the 7.9 to 8.5 kb interval is not calculated here — the document prints it itself, and attributes that 0.6 kb width to the error of the test method.

Two further published figures from the same report convert that width into years. Page 4 of the same report says: the telomere length of adults is known to decrease on average by 0.04 to 0.06 kb per year; this represents the average decline within the general population, while the rate of decline for an individual may differ depending on lifestyle, stress and disease [Note 15].

Divide 0.3 kb by the annual rate of shortening: 0.3 divided by 0.06 is 5 years; 0.3 divided by 0.04 is 7.5 years. In other words, the one-sided 0.3 kb margin of error printed on the sample report converts, at that same document's own average shortening rate, into 5 to 7.5 years of average telomere ageing; and the full 7.9–8.5 kb width the document itself prints converts into 10 to 15 years.

⚠️ Two other published figures on the same page point the opposite way, and both have to be laid out too. Page 1 prints, besides 8.2±0.3 kb, an "Average telomere length of same age group 6.8kb" and a "Your Position Compared to the same age group top 6.2%". Which is to say that in this example the distance between the individual value and the same-age average is 1.4 kb, more than four times that ±0.3 kb margin of error. In other words: the margin of error converts into a large number of years, but it does not necessarily swamp a result that sits a long way from the average. Both sides are set out here; the side that suits the argument is not the only one shown.

The same page carries a more direct comparison still. The sample report lists two published measurements: "1st / 24.04.21 / 8.5±0.3kb / slow" and "2nd(Present) / 25.06.04 / 8.2±0.3kb / slow" [Note 16]. The two differ by 0.3 kb — exactly the margin of error the document itself states — and both are printed with an aging rate of "slow". The same page also states that the telomere test is recommended at intervals of no less than one year.

⚠️ The limits of this arithmetic have to be stated and not overread: all of the figures on pages 1, 2 and 4 are the sample report's example outputs, not any real customer's readings. The division converting kb into years is arithmetic performed here on figures the document has published, not a statement made by the document. So this calculation shows one thing only, and a narrow one: when a result expressed in years differs from your own previous reading by less than or about the same as the measurement error converted into years, that difference cannot support a reading like "I got two years younger this year".

⚠️ If you are planning to test again in a few months to see whether anything has "improved": that is precisely the use repeatability is most fatal to. The change you see may be the same order of magnitude as the technical noise.

#What do these clocks actually predict? The largest head-to-head comparison

Second- and third-generation clocks really are associated with disease risk, and clearly more strongly than the first generation; but even in this, the most favourable study, only a minority of models reached statistical significance.

A 2025 paper in Nature Communications (Mavrommatis et al., Nat Commun 2025;16(1):11164) used the Scottish Generation Scotland cohort (n = 18,859) — one of the largest DNA methylation datasets in the world — to compare 14 commonly used clocks against 174 incident diseases plus all-cause mortality over ten years of follow-up. The concluding sentence of the abstract: second- and third-generation clocks significantly outperform first-generation clocks, which have limited applications in disease settings [Note 17].

And the other concluding sentence of the same abstract has to be given alongside it, because it points a different way: second- and third-generation epigenetic clocks show promise for disease risk prediction, particularly in relation to respiratory and liver-based conditions [Note 17]. In other words: this is not a paper that dismisses the clocks. It is a paper saying they show promise, and that the promise is concentrated in certain classes of disease.

The body of the paper states: 174 models were run for each clock, giving 2,436 models in total; the Bonferroni threshold was P < 0.05/174 = 2.9 × 10⁻⁴.

Do the arithmetic (on the published figures above; not a statement of the paper's own): the abstract records 176 associations passing Bonferroni correction. 176 divided by 2,436 is 7.2%. Which is to say that even in this large analysis, the one most favourable to the clocks, about one in fourteen of the clock-and-disease models reached significance.

Two further figures belong with it, both of them the paper's own statements: for 27 diseases (primary lung cancer and diabetes among them) the clock hazard ratio exceeded that same clock's association with all-cause mortality; and 32 of the 176 findings improved classification accuracy by more than 1% when the clock was added to a model containing only traditional risk factors — 32 divided by 176 is 18.2% (calculated here).

⚠️ This paper's own figures do not add up, and that is reported here as found, not patched. The abstract says 176 significant associations; the body says first-generation clocks accounted for 9 of them and second- and third-generation clocks for 162. 9 plus 162 is 171, not 176. At the same time the body describes those 9 as "representing ~5% of all significant findings" — about 5% of all the significant findings — and 9 divided by 176 is 5.1%, which is consistent with 176 instead. Where the remaining 5 are is not known here, and no guess will be made to balance the sum. So the following sentence stays as the original puts it: first-generation clocks recorded only 9 disease associations passing correction, which the authors themselves describe as "limited applications in disease settings". As for the "9 against 176" ratio itself — since the component figures do not sum to the total, it should not be quoted as a clean calculation.

⚠️ Two further limits: (i) the paper's own stated limitations include that they related DNAm profiling from whole-blood samples against multi-tissue diseases and non-blood tissue specific diseases in a Scottish-based cohort, and that they acknowledge not having adjusted exhaustively for covariates such as family history, polygenic risk scores, medication, social factors and blood pressure [Note 18]; and (ii) Generation Scotland is a Scottish family cohort, the paper offers no validation in any Chinese or East Asian population, and it cannot be treated as evidence about Hong Kong people.

⚠️ If you want a number that tells you whether you are "high risk": what this paper supports is an association at population level between second- and third-generation clocks and the incidence of certain diseases. It does not support the proposition that this one reading of yours can guide your medical decisions. Between a population association and an individual decision lie two barriers — repeatability, and clinical validation.

#What did the most energetic seller write on his own website?

This is the shortest section here and the weightiest: the loudest promoter of the category describes his own tests, on his own website, as preliminary, experimental and for research purposes only.

The disclaimer at the foot of Bryan Johnson's protocol site (protocol.bryanjohnson.com, whose sitemap lists a single page — the whole site is one document; retrieved 3 August 2026) is headed "Biological age measurements and claims", and is quoted in full at [Note 19]: any claims regarding the measurement or reversal of biological age are preliminary and can be influenced by various biostatistical errors; ongoing, formal peer-reviewed studies are essential for validating these biological age tests for FDA approval and confirming any clinically relevant changes; please note: these tests are experimental and intended solely for research purposes; they should not replace or supplement any clinical tests recommended by licensed medical professionals.

⚠️ The middle sentence cannot be cut out, because it runs the opposite way to the two around it: he describes his own data as rigorous, and describes validation studies as under way. This article has no information with which to check what those studies are, how far along they are, or whether anything has been published — but it is part of the same passage, and cutting it would be taking only the two sentences that suit the argument.

Another passage on the same page describes the nature of the whole protocol: it encompasses a mix of on-label, off-label and unlicensed therapies, as well as research-use-only tests; this protocol represents an experimental clinical research project [Note 19].

In other words: even the person most invested in this does not tell you the number is a clinical indicator you can act on medically. His own words are preliminary, experimental, research purposes only, should not replace or supplement clinical tests. What a buyer is worth asking before paying: has the supplier in front of you written anything equivalent?

⚠️ A limit: this is the disclaimer on a personal protocol website, not the position of a regulator or a professional body. Its force comes from who wrote it, not from any standing it has.

#What do regulators and professional bodies say?

Not one regulator found here has issued any guidance, statement or enforcement action on consumer "biological age" or epigenetic age testing. That is a blank, not an endorsement — and in the American system the blank has a very specific explanation.

#The United States: these tests are simply not subject to premarket review

The US Food and Drug Administration's "Direct-to-Consumer Tests" page states: in general, direct-to-consumer tests for non-medical, general wellness or low risk medical purposes are not reviewed by the FDA before they are offered; those for moderate to high risk medical purposes are generally reviewed; and, as a matter of policy, the FDA generally does not review laboratory developed tests (LDTs), which are created and performed in a single laboratory and offered to patients only when prescribed by a health care provider — such tests typically do not have the FDA's independent assurance of analytical validity, clinical validity or clear communication of test results [Note 20].

In other words: the line is drawn by risk, not by a blanket presence or absence of review. A biological age test sold as "general wellness" falls on the not-reviewed side — and that is the source of the blank described below.

⚠️ But the "general wellness" label is not something the FDA assigned to biological age tests. Section III of the FDA's General Wellness: Policy for Low Risk Devices (reissued 6 January 2026, 14 pages) itemises the categories of general wellness claim it means: weight management, physical fitness (including products intended for recreational use), relaxation or stress management, mental acuity, self-esteem, sleep management, or sexual function [Note 21]. Ageing, longevity and biological age are not one of them — this FDA document neither lists them as permitted nor excludes them expressly. So all that can be said here is that a product positioned by its seller as general wellness will not undergo premarket review; this article has no information showing that the FDA has ever taken a position on whether a biological age test is "general wellness".

And the rule that was going to close that gap has been vacated. The FDA's "Laboratory Developed Tests" page (content current as of 19 September 2025) states: on 6 May 2024 the FDA issued a final rule amending the definition of "in vitro diagnostic products"; on 31 March 2025 a federal district court vacated that rule; on 19 September 2025 the FDA issued a final rule reverting to the text of the regulation as it existed before the effective date of the May 2024 final rule [Note 22].

In one sentence: an epigenetic age test operating as an LDT is, under FDA policy, generally not subject to premarket review — the original wording is "in general" and "some types", not "in every case".

⚠️ That sentence has to be read together with the FDA's own condition: the passage quoted above defines an LDT as created and performed in a single laboratory and "offered to patients only when prescribed by a health care provider". Whether a test ordered online by a consumer with no prescription meets that condition is not something the page says and not something this article has information on — so it will not be claimed here that all consumer biological age tests travel the LDT route.

This matches the FDA's public device databases. Across the FDA's three public device databases — 510(k), PMA and device classification (the databases report themselves as updated on 20 July 2026) — not one device record contains the words "telomere", "epigenetic" or "biological age".

⚠️ That sentence has to be read within its limits and cannot be inflated into "the FDA has never approved any related device". It is a fact about device names and record text, not a judgment about device purpose.

#The nearest thing to an enforcement action was against a supplement, not a test

On 18 April 2018 the US Federal Trade Commission issued a complaint (21 pages) over the TA-65 products against Telomerase Activation Sciences, Inc. and Noel Thomas Patton. The complaint sets out six counts in this order: false or unsubstantiated efficacy claims; false establishment claims; deceptive format (a paid television segment presented as an independent programme); failure to disclose a material connection with consumer endorsers; false "independent user" claims; and providing the means and instrumentalities to commercial customers to engage in deception. Only the first two, the ones closest to this article's subject, are quoted below; the original is at [Note 23].

Paragraph 13 of the complaint sets out three price tiers, and states expressly that they are taken from earlier versions of the website: $600 for a three-month supply at a low dose level (one 250-unit capsule daily); $1,200 at a mid dose (two capsules daily); and $2,200 at a high dose (four capsules daily) [Note 24]. The same paragraph also records the website price at the time of the complaint as $600 for a bottle of 90 capsules and $100 for a bottle of 30.

⚠️ A limit on the paragraph numbers: what is quoted here is paragraph 43 and paragraph 45; paragraph 44 in between is not held here. By the structure of the complaint, paragraph 45 answers the representations alleged in the paragraph immediately preceding — that is, paragraph 44. The original text of paragraph 44 has not been obtained here, so nothing will be said on its behalf; a reader coming to paragraph 45 should know it is a response, not a free-standing scientific ruling.

⚠️ The limit has to be stated plainly: TA-65 is a telomerase activation supplement, not a biological age test. It is the closest regulatory action to this subject that was found, and what it is must be said straight. In the FTC's press releases and case library the result count for telomere is 0; epigenetic returns one press release, and that one is a company name — "The Epigenetics Healing Center, LLC" on a list of coronavirus-related warning letters dated 7 May 2020, concerning intravenous and vitamin C therapy claims, and unrelated to epigenetic age testing.

#The United Kingdom: a caution about commercial health checks in general

The UK National Screening Committee's NHS and commercial health screening tests: important considerations, published 21 November 2023, states: companies may offer screening that is not evidence-based, meaning there are no reliable studies confirming that it leads to improved health outcomes for those screened, and companies often promote screening as simply offering peace of mind; often there is no quality assurance, meaning the tests might not be the same everywhere; many people, including healthcare professionals, overestimate the benefits of screening and underestimate both the harms and the difficulty of delivering screening effectively; and if the test only claims to offer peace of mind, then a fitness check and good advice about nutrition and physical activity might be a better option [Note 25].

Note how tightly the second of those clauses fits the repeatability section: "often there is no quality assurance, meaning the tests might not be the same everywhere" — that is exactly what the telomere ring trial demonstrated with data. One is a regulator's general caution and the other a peer-reviewed measurement, and they point the same way.

⚠️ This document is about commercial screening in general; nowhere in it does it mention biological age, epigenetic clocks or telomeres, and none of the committee's 146 publications on gov.uk mentions them either. It cannot be written up as "the UK authorities have commented on biological age tests".

Pointing the same way is the letter of 11 September 2023 to the Secretary of State for Health signed jointly by four UK professional bodies (the Association for Laboratory Medicine, the Royal College of Pathologists, the Institute of Biomedical Science and the Royal College of General Practitioners): diagnostic tests are being marketed to the public in ways that create patient safety issues; capacity issues within the NHS are enabling the exploitation of the "worried well" — selling anxious people data without providing them with any professional support, and often no interpretation of their data [Note 26]. That letter likewise does not name biological age testing.

#Hong Kong: no specific provision, but one general provision governing promotion

Clauses 5.2.1.1 and 5.2.1.2 of the Medical Council of Hong Kong's Code of Professional Conduct for the Guidance of Registered Medical Practitioners (revised October 2022) are quoted in full at [Note 27]: any information provided by a doctor to the public or his patients must be accurate, factual, objectively verifiable and presented in a balanced manner; and such information must not be exaggerated or misleading, must not be comparative with or claim superiority over other doctors, must not claim uniqueness without proper justification, must not aim to solicit or canvass for patients, must not be used for commercial promotion of medical and health related products and services, must not be sensational or unduly persuasive, must not arouse unjustified public concern or distress, must not generate unrealistic expectations, and must not disparage other doctors.

Of those nine limbs, the one bearing most directly on this subject is the fifth — and it is precisely the limb where the Chinese and English versions differ.

⚠️ The difference is the words "and services". The Chinese version reads 「為相關醫療及健康產品作出商業宣傳」, while the English reads "commercial promotion of medical and health related products and services" — the English has the extra "and services". A biological age test is a service, so on the governing version (the English) it falls squarely within this limb; reading the Chinese alone, you might not see that. The difference is itself something a reader is entitled to know.

⚠️ But the more important point is this: the Code in full (85 pages in English, 74 in Chinese) contains none of anti-ag*, rejuvenat*, telomere, epigenetic, biological age, longevity, stem cell, or the Chinese 抗衰老, 逆齡, 生物年齡, 端粒, 幹細胞. Which is to say: the Code has no provision specific to anti-ageing or biological age, only the general provision above; and the general provision governs doctors, not retail testing services that involve no registered medical practitioner.

At the Consumer Council the position is equally blank. On its Chinese search interface, the result counts for 生物年齡, 端粒 and 表觀遺傳 are all 0 (on the same interface 抗衰老 returns 14 items, 基因檢測 11 and 體檢 105). There is no article on the Consumer Council website on the subject of biological age testing. It does have one press release on an adjacent subject (14 July 2021, CHOICE issue 537) about genetic (DNA) testing in general, which nowhere mentions biological age, epigenetic age, DNA methylation age or telomere testing [Note 29].

That release reports that the Council sent staff posing as ordinary customers to enquire about these companies' services and found that the reliability of many of the genetic tests offered still awaits confirmation by clinical research, so the results are open to doubt; it reminds consumers that predictive tests may be affected by multiple genes or external factors and may even produce false positives or false negatives, and that before undergoing such testing one must seek the advice of a healthcare professional (one's own doctor, for instance) before deciding whether to proceed (our translation from the Chinese original).

Chinese original:

「消費者委員會派員以一般顧客身分，向這類公司查詢服務詳情，發現不少公司提供的基因測試，至今仍有待臨床研究結果證明其可靠性，檢測結果存疑。消委會提醒消費者，預測性的檢測，有機會受多個基因或外在因素影響，甚至出現假陽性或假陰性等問題。故此，消費者在進行這類檢測前，必須尋求專業醫護人員（例如主診醫生）的意見，再決定是否接受檢測，以免繳付大筆檢測費之餘，出現誤診，或增添不必要的焦慮。」

That press release also lists three points to watch, quoted in full at [Note 30]; the second of them connects directly to the finding in the last section: be wary of genetic testing suppliers using the results as an opportunity to sell nutritional supplements or treatment plans, which may be nutritionally unnecessary and, if taken to excess or misused, may harm health. Of the two biomarker suppliers with published prices found here, one runs a website that is itself an NMN supplement shop, and the other (which publishes no price) describes the deliverable of its test as a supplement plan tailored to you. ⚠️ This press release is about genetic testing in general, not biological age testing; it is cited here because the combination it describes is visible in the market found here.

#The earliest and heaviest academic statement dates from 2002

The Position statement on human aging signed by 52 ageing researchers (Olshansky, Hayflick, Carnes, J Gerontol A Biol Sci Med Sci 2002) states: a large number of products are currently being sold by antiaging entrepreneurs who claim that it is now possible to slow, stop or reverse human ageing; the products being sold have no scientifically demonstrated efficacy, in some cases they may be harmful, and those selling them often misrepresent the science upon which they are based [Note 28].

⚠️ That statement is from 2002, earlier than every clock mentioned here. Its target is anti-ageing products, not tests. It is cited to show how long this tension has run, not as a verdict on the clocks of 2026.

⚠️ If you see phrases like "FDA recognised" or "internationally accredited": the FDA's three public device databases contain no device record with the words "telomere", "epigenetic" or "biological age" in them; and on the supplier pages found in Hong Kong, the only laboratory accreditation claim is CAP, and it concerns GC Genome in Korea, not a Hong Kong laboratory.

#What is actually on sale in Hong Kong, and at what price?

The limits of this section have to be stated first, because they bear directly on how these figures may be used.

⚠️ Every sentence below reports only what each supplier has published on its own website, and this list is not complete. "Not on the list" does not mean "does not exist" — so no conclusion is drawn here about the overall size of the Hong Kong market or its price range. Among the things outside the list, on which nothing is stated either way: whether the health check packages of Hong Kong's private hospitals include biological age, telomere or epigenetic items, and whether Gleneagles, Sun Medical Laboratory, Diagcor or TruDiagnostic supply Hong Kong.

⚠️ A second limit, equally important: not one of the suppliers found here states which clock it uses. The names Horvath, GrimAge, PhenoAge, DunedinPACE and OMICmAge appear on no product page of any supplier found (BMS Clinic's Chinese and English pages both cite Horvath and Raj 2021 in the reference list at the foot of the page, but neither page says anywhere that it sells such a test — citing a paper is not the same as selling that paper's clock).

#Products with a published price

⚠️ The currency notation is reproduced as found and nothing is added to it. The MedDx and ATML pages carry only a bare $ sign with no "HK" prefix (other tests on the same MedDx page are written the same way, ($5800), ($7900)). They are not rewritten here as "HK$3,800". The OT&P pages do carry an explicit HK$ prefix.

ATML's product does not say what it measures, but its page prints a statutory notice of its own, and that is the passage a buyer should read most closely [Note 31]:

This product is not registered under the Pharmacy and Poisons Ordinance or the Chinese Medicine Ordinance; any claim made for this product has not been subject to any assessment in Hong Kong for the purposes of such registration; this product is not intended to diagnose, treat or prevent any disease; and results vary between individuals and circumstances (our translation from the Chinese original).

Chinese original:

「此產品沒有根據《藥劑業及毒藥條例》或《中醫藥條例》註冊。此產品作出的任何聲稱亦沒有為進行該等註冊而在香港接受評核。此產品並不供作診斷、治療或預防任何疾病之用。產品效果因人、不同情況而異。」

⚠️ Two points travel with that passage: (i) the English and Chinese halves of it are printed on the Chinese page on the same line and within the same paragraph, while the English page carries only the English half. A buyer reading the English page alone will see the English of that passage but not the Chinese. (ii) It is a statutory notice in a standard form applied to the products on that site, not a statement of the limits of this particular test's performance — but what it says (not for diagnosing, treating or preventing any disease; results vary between individuals) points the same way as everything below.

⚠️ Both biomarker suppliers with a price also sell supplements or deliver a supplement plan: ATML Healthspan's website is an NMN/NAD+ supplement shop — its own navigation menu categories are 「抗衰老醫學」, 「NMN活健齡」, 「抗衰老咖啡」, 「美白淡斑」, 「皮膚抗老」 and 「網上商店」, while the biological age blood test itself is filed on that site under "Medical , Science"; and the "related products" panel at the foot of the same page (whose contents change over time; on 3 August 2026 it showed 「活健齡® NMN15000 (250mg x 60粒裝)」 and 「皮膚抗老醫學評估 $ 490.00」, among others) is likewise entirely supplements or beauty products. And the test from Central Wellness (below, no price published) describes its own deliverable as a supplement plan tailored to you. This is factual description only; no supplier is being evaluated.

#One product, two of its own documents, saying opposite things

This is not an assessment of MedDx. It is a statement of fact about the contents of two of its documents: its product page and its own sample report say opposite things about what a biological age below chronological age means.

The product page (Chinese version) says: ageing speed refers to the difference between biological age and actual age; if the subject's biological age is close to the actual age, the ageing speed is normal; if the biological age is lower than the actual age, the ageing speed is considered too fast [Note 32]. (The English version of the same passage is consistent with it: "If the biological age is lower than the actual age, the ageing speed is considered too fast.")

And the sample report PDF from the same company, reached from the product page's own 「查看樣本報告」 link, says the opposite on page 3: "Because the bio-age is [younger than] the actual age, the aging rate is [slow]" [Note 32]. The example output on page 1 of the report is likewise "Estimated biological age 45yrs" paired with "Aging rate slow". Page 3 of the same report also prints its definition of the term: the aging rate is the value obtained by dividing the bio-age by the actual age.

Which is to say: the sales page and the delivered document of the same product say opposite things about what its own central number means. That contradiction is reported here as found, neither patched nor adjudicated — because the contradiction is itself the thing a buyer needs to know: if the seller's own two documents disagree about whether a low number is good or bad, how to read the report you are holding is not self-evident.

#A more interesting finding: the most expensive package does not sell a clock

The itemised contents of OT&P Healthcare's HK$19,800 "Longevity health check" contain no epigenetic clock, no DNA methylation and no telomeres. The list opens with the "Longevity physiological function test" group, and the first two items of that group are grip strength and VO2 Max.

The Chinese page itemises the package's contents; the complete list is at [Note 33]. What matters is that from beginning to end not one item on that list is an epigenetic clock, a methylation assay or a telomere test — that is the point of this section.

The page does use the term 生理年齡: its Chinese text says these tests give a comprehensive picture of your physiological age, DNA, general health indicators, inflammation levels, hormonal health and cardiovascular condition (including arterial age and so on) — and yet the itemised list contains no DNA test of any kind. That gap is itself this section's finding.

⚠️ The Chinese and English lists for the same package are not identical. The English list carries "Apolipoprotein B" between "Fasting Insulin" and "eGFR"; the Chinese list at the same position goes straight from 「空腹胰島素」 to 「估算腎小球過濾率（eGFR）」, with no apolipoprotein B item. The remaining items on the two lists correspond broadly. A buyer who cares whether a particular item is included should go by the contract or quotation actually received, and ask directly.

As for grip strength and maximal oxygen uptake: both have published associations with mortality in prospective cohort studies. The PURE study (Leong et al., Lancet 2015) followed 139,691 participants in 17 countries, with a median follow-up of 4.0 years during which 3,379 (2%) died; each 5 kg reduction in grip strength carried an all-cause mortality hazard ratio of 1.16 (95% CI 1.13–1.20). ⚠️ But the paper's own conclusion states plainly that the causal question is unanswered: further research is needed to identify determinants of muscular strength and to test whether improvement in strength reduces mortality and cardiovascular disease [Note 34]. Kokkinos et al. (J Am Coll Cardiol 2022), using 750,302 US veterans with a median follow-up of 10.2 years and 174,807 deaths, found that the least fit fifth of the population had roughly four times the mortality risk of the extremely fit (HR 4.09; 95% CI 3.90–4.20); that cohort consisted of US veterans referred for exercise stress testing for clinical reasons, and the paper itself makes no statement about applicability to Hong Kong. Neither of these is this article's subject; for the detailed discussion see 〈What the evidence for longevity science actually shows〉.

⚠️ If you are about to pay on the basis that "this package will measure my biological age": ask for the itemised list, then look on it for the words "methylation", "epigenetic" and "telomere". The HK$19,800 package above has none of them — and the package contents are published, so anyone can check.

#Suppliers that promote such testing but publish no price

Each of the following promotes testing of this kind on its own web pages, and the relevant published pages carry no price.

Central Wellness (Central & Stanley Wellness) — promotes a "Bio Age" / "Bioyon" test. None of the three relevant pages carries a price; the site has no online shop and no fee page; and the "Bio Age" link in its own services menu returns a 404. The call to action is a WhatsApp enquiry. The pages name no clock and no laboratory method; the brand name is "Bioyon". ⚠️ The above is the page content as at 2 August 2026; nothing is stated here about the position since.

BMS Clinic Limited — its Chinese page lists 生物年齡 as one item under "advanced biomarker testing", but names no test product, analyte or clock, and publishes no price; the call to action is to book a first consultation. ⚠️ The disclaimer on the English version of that page is itself cut off mid-sentence on the page, while the Chinese version is complete — neither version is used here to fill in the other.

In addition, the published product catalogues of the following organisations contain no biological age or telomere product: Prenetics (corporate site, no diagnostic products), CircleDNA (per the archived copy of 12 March 2026; ⚠️ that is an archived copy, not the live state on 2 August 2026), Cordlife (Hong Kong), BGI Hong Kong, Genetrack Hong Kong and re:HEALTH. S-CELL's site has a product category page headed 「生物年齡測試」, but per the archived copy of 11 October 2025 that page is empty; the live state could not be determined.

#What a result can do, and what it cannot

This is the most important section here, because it is not about whether the number is accurate. It is about what you can do once you are holding it — and the answer is: not in either direction.

#"Younger" is not a certificate of health

A result showing your biological age below your chronological age cannot be used for reassurance, and certainly cannot be used to postpone or cancel any other test. Not one of the tests cited here has been validated to rule out any disease. And it is not only this article saying so — it is the products' own documents [Note 35]:

Bryan Johnson's site says "They should not replace or supplement any clinical tests recommended by licensed medical professionals." Page 8 of the MedDx sample report says "In this test, the clinical significance of the test results has not been established, and there is still insufficient objective validity that the health-related behaviors that follow are useful", and "This test was developed and its performance characteristics determined by GC Genome. It has not been cleared or approved by the Korea Ministry of Food and Drug Safety (MFDS)." Page 5 of the same report says "Telomere test cannot be used to diagnose diseases or to determine actions related to the treatment of a disease." And the MedDx website terms (Chinese version) and the ATML Healthspan product page say this:

The testing is a health check only. Although the results may appear normal on their face, some hidden disease may still become apparent at a later time. Where the result is abnormal, a doctor should be consulted for any advice on diagnosis or treatment, and the doctor's fee is not included in the amount already paid to MedDx and its designated laboratory 「金域檢驗」. / This product is not intended to diagnose, treat or prevent any disease. Results vary between individuals and circumstances. (our translation from the Chinese original)

Chinese original:

「檢測只屬 身體檢查 ，儘管檢查結果表面上屬正常，還有可能有某些隱藏的疾病在稍後時間才會顯然。如結果異常，應求診醫生給予任何診斷或治療的建議，而醫生費用並不包括在已附給 MedDx及其指定實驗室"金域檢驗"之金額當中。」 「此產品並不供作診斷、治療或預防任何疾病之用。產品效果因人、不同情況而異。」

In one sentence: the sellers themselves state in writing that this result rules nothing out. So the inference "my biological age is 5 years lower, I can skip the colonoscopy, skip the blood pressure check, skip the doctor" collapses in front of these documents. Whether and when to have screening or a health check is decided by your doctor on your clinical situation and risk, and has nothing to do with this number.

#"Older" is not a diagnosis either

The reverse has to be said plainly too. A commentary in the December 2025 issue of the AMA Journal of Ethics (by two philosophy researchers at the University of Liverpool, Hauskeller M and Shore L; ⚠️ this is a commentary published in that journal, not a position statement of the American Medical Association) writes, of a case of a 56-year-old who receives a result of "biologically 65-years-old": if, like patient AA, we are 56 years old and told that our biological age is actually 65, then all that can possibly mean is that whatever parts of our bodies have been tested are in a worse state than is to be expected for someone our age, and that this state is more commonly found in someone who is 9 years older — which is, of course, still a good reason to be concerned about the state of one's health and possibly to make some changes in one's life [Note 36].

The same article is equally explicit about what the results do not entail: they do not entail that he will develop morbidity or die earlier than would be expected on the basis of his chronological age; rather, they merely indicate that certain aspects of his physical condition make it more likely that particular health hazards lie in store for him in the future if not addressed [Note 36]. The authors also raise the psychological dimension: to be told that one is actually much older than one thought one was might well have a similar effect to being diagnosed as suffering from a terminal illness.

⚠️ There is a tension inside that article, and both halves are the authors' own sentences, so both must be quoted. On one hand they write that these tests "have been shown to be fairly reliable indicators of a person's risk of developing certain age-related diseases"; on the other that "there is no such thing as a particular person's biological age", and in their conclusion that "Biological age is a fiction that should, at the very least, be clearly identified as such by clinicians". And in that same concluding passage, a third sentence of their own: "Epigenetic testing can, of course, reveal potential or existing health problems and provide valuable information that can be used to improve a patient's health and lifespan." [Note 37] The authors are not saying these tests have nothing to do with risk. They are saying that "risk association" and "how old you really are" are two different things — the first may exist; the second is a constructed number.

The same article also explains why your doctor may not be able to help you read that report: for this discussion to be possible, the doctor would have to know exactly how the test results were generated, which is harder to determine than it should be because of the proprietary nature of current epigenetic clocks and the economic incentives that accompany them [Note 37].

⚠️ If you are planning to take the report to your family doctor for interpretation: if the supplier does not disclose which clock, which version and where the reference values come from, your doctor can see no further into where that number came from than you can. That is a question to ask before buying, not to discover afterwards.

#A concrete example: in one real trial, it registered nothing at all

The PEARL trial (a 48-week, double-blind, placebo-controlled safety trial of low-dose rapamycin, completed by 114 people) sent a subset of its samples (n = 24, 9 women and 15 men) for TruDiagnostic TruAge epigenetic ageing analysis, and the result is stated simply: within the epigenetic testing results, they saw no meaningful significant changes between groups [Note 38].

⚠️ A limit: n = 24 is very small, and the trial's primary outcome (visceral fat) was itself negative, so this result cannot be used to prove that "epigenetic testing measures nothing". It is a concrete reference point: a consumer-grade epigenetic platform, in a small sample within an intervention trial running nearly a year, registered no between-group difference.

⚠️ The interests must be stated with it, and stated to the scope of the original: the trial's declaration of interests names seven authors as employees and shareholders of AgelessRx [Note 38]. ⚠️ That sentence names seven authors, not all of them, and it will not be written up any larger here. AgelessRx is a commercial body selling rapamycin and other longevity-related prescription services (its website price is "Starting at $65 / mo"). And two things on the clinical trial registration (NCT04488601) belong together: the lead sponsor is AgelessRx (class INDUSTRY), and the collaborator field lists the University of California, Los Angeles. The direction of that relationship is worth noticing: a body that sells longevity interventions published a negative epigenetic result that does it no favours.

#What to do next

This article will not answer "do it" or "don't do it" for you, but it can narrow that decision down to a few questions you can answer yourself.

Before paying, settle one thing: what do you plan to do with the number? If the answer is "use it to decide whether to have some test or take some drug" — every seller document above states in writing that it cannot do that, including Bryan Johnson's own passage. If the answer is "use it as a baseline and test again in a few months to see whether it improves" — that runs straight into the repeatability wall: on the six commonly used clocks as published, technical noise alone reaches 9 years. If the answer is "for peace of mind" — the UK NSC's caution on commercial screening addresses exactly that: if a test only claims to offer peace of mind, then a fitness check and good advice about nutrition and physical activity might be a better option. And, as it happens, the first two items of the most expensive Hong Kong package found here are grip strength and VO2 Max.

Then put these six questions to the supplier, and get answers:

1. What are you measuring? Telomere length, DNA methylation, or a physical fitness assessment? All three go by "biological age" or "body age" in Hong Kong, and they are three different things.
2. Which clock, and which version? If it is an epigenetic clock: is it first generation (trained against chronological age), second generation (trained against clinical outcomes), or an indicator like DunedinPACE that measures a speed rather than an age? Is it the original or the principal-component retrained version?
3. What is the test-retest gap in years? Measure the same sample twice — how far apart does your platform land? There is a published reference point for this question: up to 9 years across six commonly used clocks, narrowing to within about 1.5 years for the principal-component retrained versions.
4. Where do the reference values come from, and which laboratory does the work? Because on the published international ring trial, telomere reference ranges cannot be shared between laboratories; change the laboratory and the number is not directly comparable. And even if you do not, the technical variation within a single laboratory measured by that ring trial ranged from 1.4% to 9.5%.
5. Have you written down what this result cannot be used for? Every seller document cited above has. One that has not is missing a document.
6. Does that price include interpretation, and interpretation by whom? The MedDx page, for instance, says the price includes an explanation by a dietitian, while seeing a specialist 「如有呈陽性結果」 costs a further $ 500. Do the arithmetic: $ 3800 plus $500 is $4300 (adding the two figures that page itself publishes; that page writes only a $ sign, with no HK prefix).

As for "what actually does have evidence behind it" — that is another article's subject; see 〈What the evidence for longevity science actually shows〉.

#Frequently asked questions

Which Hong Kong supplier gives the best value?

This article will not answer that, and has no material with which to answer it. The suppliers found here are not a complete list (see the limits above), and not one of them names the clock it uses — without knowing what is being measured or how far apart repeat measurements land, "value" cannot be calculated at all. Nor does this article recommend or disparage any supplier.

Can you get GrimAge or DunedinPACE in Hong Kong?

Not one of the suppliers found here names on its product page which clock it sells. That is "not among the suppliers listed here", not "not in Hong Kong" — the list of suppliers found here is not complete, and private hospital health check packages were not examined.

My report says my biological age is 5 years lower. Does that mean I am healthy?

It cannot be read that way. Not one of the tests cited here has been validated to rule out any disease; MedDx's own sample report states that "the clinical significance of the test results has not been established", and its website terms state that although the results may appear normal on their face, some hidden disease may still become apparent at a later time.

And if it says I am 9 years older — should I be frightened?

The AMA Journal of Ethics commentary puts it clearly: the result does not entail that you will fall ill or die early; it merely indicates that certain tested aspects of your body are in a worse state than would be expected for your age. It is a reason to attend to your health, not a diagnosis. Whether you have a disease is settled by clinical assessment and the corresponding tests, not by this number.

I have changed my habits — can I test again in six months to see the effect?

This is the use most people want and the one repeatability hits most directly. Higgins-Chen et al. (Nat Aging 2022) showed that technical noise alone can put replicates on six commonly used clocks up to 9 years apart; the principal-component retrained versions narrow that to within about 1.5 years, but not one Hong Kong supplier found here states which version it uses. To make that kind of before-and-after comparison you would at minimum need an answer to question 3 above.

Are telomere tests and epigenetic clocks the same thing?

No, they are two independent measurements. And in the 2025 study comparing 14 clocks, DNAm-estimated telomere length was among those with the smaller effects. Separately, the value of telomere length as an overall marker of ageing was already described in a 2013 review as the subject of "heated debate", with its associations with other markers of health "equivocal".

Is DunedinPACE the newest and most accurate one?

It is third generation, but what it outputs is not an age in years — it is a pace of ageing, so it is not a more accurate version of the first generation but a different kind of indicator. It excluded probes with low test-retest reliability at the modelling stage, and its abstract reports that it "showed high test-retest reliability" — so the "up to 9 years" in the repeatability section is about the other six clocks and cannot be transferred to DunedinPACE; but it is built on a single New Zealand birth cohort (born 1972–73). The paper's abstract says only that the indicator was evaluated in five further datasets without listing which five, so nothing is stated here about whether it has been validated in any Chinese or East Asian population.

Has any regulator said these tests are no good?

No. Not one regulator found here has issued any guidance, statement or enforcement action on consumer biological age or epigenetic age testing. But note the direction of that blank: it is not "reviewed and found safe", it is "never reviewed at all" — the FDA's own page states that direct-to-consumer tests for general wellness or low risk purposes are not reviewed by the FDA before they are offered, and the regulation covering laboratory developed tests was vacated by a court in 2025, with the FDA reverting to the earlier text. ⚠️ One qualification: the "general wellness" label is a result of a product's own positioning, not something the FDA assigned to biological age tests — the FDA's own general wellness policy document itemises the categories it means, and ageing, longevity and biological age are not among them.

#Notes: the original texts

[Note 1] World Health Organization, "Ageing and health" fact sheet: "At the biological level, ageing results from the impact of the accumulation of a wide variety of molecular and cellular damage over time. … These changes are neither linear nor consistent, and they are only loosely associated with a person's age in years." and "There is no typical older person. Some 80-year-olds have physical and mental capacities similar to many 30-year-olds."

[Note 2] Horvath S, Genome Biol 2013;14(10):R115: "I propose that DNA methylation age measures the cumulative effect of an epigenetic maintenance system."

[Note 3] Levine ME et al., Aging 2018;10(4):573-591: "While the first generation of epigenetic biomarkers of aging were developed using chronological age as a surrogate for biological age, we hypothesized that incorporation of composite clinical measures of phenotypic age … may identify novel CpGs"

[Note 4] Lu AT et al., Aging 2019;11(2):303-327: "The resulting predictor of lifespan, DNAm GrimAge (in units of years), is a composite biomarker based on the seven DNAm surrogates and a DNAm-based estimator of smoking pack-years."

[Note 5] Belsky DW et al., eLife 2022;11:e73420: "We distilled this two-decade Pace of Aging into a single-time-point DNA-methylation blood-test using elastic-net regression and a DNA-methylation dataset restricted to exclude probes with low test-retest reliability." First sentence of the results section of the abstract: "DunedinPACE showed high test-retest reliability, was associated with morbidity, disability, and mortality"

[Note 6] The same, competing interests: "DB, AC, DC, KS, RP, TM is listed as an inventor on a Duke University and University of Otago invention that was licensed to a commercial entity"; the remaining authors are recorded as "No competing interests declared".

[Note 7] Sanders JL, Newman AB, Epidemiol Rev 2013;35(1):112-31: "Conflicting data have generated heated debate about the value of LTL as a biomarker of overall aging." The conclusion also links shorter telomeres to "older age, male gender, Caucasian race, and possibly atherosclerosis", while "associations with other markers of health are equivocal".

[Note 8] Mavrommatis C et al., Nat Commun 2025;16(1):11164: "Smaller effect sizes were also noted for PhenoAge (P = 2.9 × 10−3) and DNAm telomere length (P < 6.2 × 10−6)."

[Note 9] Higgins-Chen AT et al., Nat Aging 2022;2(7):644-661: "Epigenetic clocks are widely used aging biomarkers calculated from DNA methylation data, but this data can be surprisingly unreliable. Here we show technical noise produces deviations up to 9 years between replicates for six prominent epigenetic clocks, limiting their utility." And, after the fix, "agreement between most replicates within 1.5 years".

[Note 10] The same, competing interests: "MEL and AHC have built epigenetic aging metrics involving the technology described in the present manuscript, and these metrics are licensed by Elysium Health through Yale University." "otherwise did not fund the study and did not play a role in conceptualization, design, decision to publish, or preparation of the manuscript" "THS was previously an employee of Elysium Health, Inc. AHC received consulting fees from FOXO Technologies, Inc. for work unrelated to the present manuscript." "All other authors report no biomedical financial interests or potential conflicts of interest"

[Note 11] Sugden K et al., Patterns 2020;1(2):100014: "the reliability of each of these measurements is not equal"; unreliable probes "were less replicable and generated an unknown volume of false negatives".

[Note 12] Martin-Ruiz CM et al., Int J Epidemiol 2015;44(5):1673-83: "Absolute results from different laboratories differed widely and could thus not be compared directly, but rankings of relative telomere lengths were highly correlated (correlation coefficients of 0.63-0.99). … However, inter-laboratory coefficients of variation (CVs) averaged about 10% for Southern blotting and STELA and more than 20% for qPCR. This difference was compensated for by a higher dynamic range for the qPCR method as shown by equal variance after z-scoring. … Gel-based and PCR-based techniques were not different in accuracy." "Intra- and inter-laboratory technical variation severely limits the usefulness of data pooling and excludes sharing of reference ranges between laboratories." And "Technical variation per laboratory, measured as median of intra- and inter-batch CVs, ranged from 1.4% to 9.5%, with differences between laboratories only marginally significant (P = 0.06)."

[Note 13] MedDx telomere test sample report, page 5, "Limitation" box (three of the four items): "* Telomere length may vary slightly depending on the test method and test conditions." "* The bias of the result is due to the value of the telomere length of the reference gene used in each test." "* The bio-age result of this test is calculated using internally constructed data of the age-specific population group, and the accumulated data is regularly updated."

[Note 14] The same, page 1, "Interpretation": "Telomeres naturally shorten with age, but the initial length differs among individuals, and the rate of shortening may vary depending on lifestyle, stress, and disease." and "The telomere length is [8.2 kb (margin of error: ±0.3kb)], which is [longer than the average length] of the same age group (±1 year)." Page 2: "Considering the error of the test method, the telomere length is from [7.9]kb to [8.5]kb." Page 1 also carries: "Average telomere length of same age group 6.8kb" "Your Position Compared to the same age group top 6.2%"

[Note 15] The same, page 4: "The telomere length of adults is known to decrease on average by 0.04–0.06 kb per year." "This represents the average decline within the general population, while the rate of decline for an individual may differ depending on lifestyle, stress, and disease."

[Note 16] The same: "1st / 24.04.21 / 8.5±0.3kb / slow" "2nd(Present) / 25.06.04 / 8.2±0.3kb / slow" and "The telomere test is recommended at intervals of no less than one year."

[Note 17] Mavrommatis C et al., Nat Commun 2025;16(1):11164, abstract: "Second- and third-generation clocks significantly outperform first-generation clocks, which have limited applications in disease settings." and "Second- and third-generation epigenetic clocks show promise for disease risk prediction, particularly in relation to respiratory and liver-based conditions." Body: "A total of 174 models were run for each of the 14 clocks, giving 2436 models in total."

[Note 18] The same, limitations: "we related DNAm profiling from whole-blood samples against multi-tissue diseases and non-blood tissue specific diseases in a Scottish-based cohort"

[Note 19] Bryan Johnson protocol site, "Biological age measurements and claims": "Any claims regarding the measurement or reversal of biological age are preliminary and can be influenced by various biostatistical errors, including statistical variation, reference range relevancy, and clinical outcome significance. Ongoing, formal peer-reviewed studies are essential for validating these biological age tests for FDA approval (or its international equivalent) and confirming any clinically relevant changes in biological age. The data presented reflect our team's rigorous efforts to adhere to current scientific and biostatistical standards, while formal peer-reviewed validation studies are underway. Please note: These tests are experimental and intended solely for research purposes. They should not replace or supplement any clinical tests recommended by licensed medical professionals." The same page also carries: "The protocol encompasses a mix of on-label, off-label, and unlicensed therapies, as well as research-use-only tests. … This protocol represents an experimental clinical research project."

[Note 20] FDA "Direct-to-Consumer Tests": "Some direct-to-consumer tests are reviewed by the FDA while others are not. In general, direct-to-consumer tests for non-medical, general wellness, or low risk medical purposes are not reviewed by the FDA before they are offered. Direct-to-consumer tests for moderate to high risk medical purposes, which may have a higher impact on medical care, are generally reviewed by the FDA to determine the validity of test claims." "As a matter of policy, the FDA generally does not review some types of tests, called laboratory developed tests (LDTs), that are created and performed in a single laboratory, if they are offered to patients only when prescribed by a health care provider. These tests typically do not have the FDA's independent assurance of the analytical validity, clinical validity, or clear communication of test results."

[Note 21] FDA General Wellness: Policy for Low Risk Devices (reissued 6 January 2026), section III: "weight management, • physical fitness, including products intended for recreational use, • relaxation or stress management, • mental acuity, • self-esteem …, • sleep management, or • sexual function"

[Note 22] FDA "Laboratory Developed Tests" (content current as of 19 September 2025): "On May 6, 2024, the FDA issued a final rule amending the definition of "in vitro diagnostic products" … On March 31, 2025, a federal district court vacated that rule. On September 19, 2025, the FDA issued a final rule reverting to the text of the regulation as it existed prior to the effective date of the May 2024 final rule."

[Note 23] FTC complaint (18 April 2018): "Respondents claim that the TA-65 products activate telomerase, lengthen short telomeres, and, thereby, extend the cellular lifespan of normal cells." "Count I — False or Unsubstantiated Efficacy Claims … Respondents have represented … that: a. TA-65 products reverse aging; … 43. The representations set forth in Paragraph 42 are false or misleading, or were not substantiated at the time the representations were made." "Count II — False Establishment Claims … 45. In fact, TA-65MD is not clinically or scientifically proven to reverse aging; prevent and repair DNA damage; restore aging immune systems; and increase bone density."

[Note 24] The same, paragraph 13: "According to earlier versions of the website, TA-65MD capsules retailed for the following approximate amounts: $600 for a three-month supply at a low dose level (one 250-unit capsule daily); $1,200 for a three-month supply at a mid-dose level (two 250-unit capsules daily); and $2,200 for a three-month supply at a high-dose level (four 250-unit capsules daily)."

[Note 25] UK National Screening Committee, NHS and commercial health screening tests: important considerations (21 November 2023): "Companies may offer screening that is not evidence-based. This means there are no reliable studies confirming that it leads to improved health outcomes for those screened. Companies often promote screening as simply offering peace of mind." "Often there is no quality assurance, meaning the tests might not be the same everywhere." "Many people, including healthcare professionals, overestimate the benefits of screening and underestimate both the harms and the difficulty of delivering screening effectively." "What would be the benefits of having the test? Is there any evidence that the test can lead to avoidance of an important health problem some time in the future? If the test only claims to offer peace of mind, then a fitness check and good advice about nutrition and physical activity might be a better option."

[Note 26] Letter of 11 September 2023 signed by four UK professional bodies: "Diagnostic tests are being marketed to the public in ways that create patient safety issues." "Capacity issues within the NHS are enabling the exploitation of the 'worried well' – selling anxious people data without providing them with any professional support, and often no interpretation of their data."

[Note 27] Medical Council of Hong Kong, Code of Professional Conduct for the Guidance of Registered Medical Practitioners (revised October 2022), clauses 5.2.1.1 and 5.2.1.2, English version: "Any information provided by a doctor to the public or his patients must be:- (a) accurate; (b) factual; (c) objectively verifiable; (d) presented in a balanced manner…" / "Such information must not:- (a) be exaggerated or misleading; (b) be comparative with or claim superiority over other doctors; (c) claim uniqueness without proper justifications for such claim; (d) aim to solicit or canvass for patients; (e) be used for commercial promotion of medical and health related products and services (for the avoidance of doubt, recommendations in clinical consultations are not regarded as commercial promotion of products and services); (f) be sensational or unduly persuasive; (g) arouse unjustified public concern or distress; (h) generate unrealistic expectations; (i) disparage other doctors (fair comments excepted)." Chinese version: 「醫生向公眾或病人提供的任何資料必須 (a) 準確；(b) 有根據；(c) 經客觀確認；(d) 以持平的方式表達(提及某種治療方法的成效時，必須列明該療法的利與弊)。」「該等資料不得：(a) 誇大或帶有誤導成分；(b) 與其他醫生作比較或聲稱優於其他醫生；(c) 聲稱服務獨一無二，但沒有適當理由支持該項聲稱；(d) 以招徠病人或兜攬生意為目的；(e) 為相關醫療及健康產品作出商業宣傳(為免生疑問，臨牀諮詢時提出的建議不屬於產品及服務的商業宣傳)；(f) 煽情或帶有過度勸誘成分；(g) 引起不適當的公眾關注或不安；(h) 造成不切實際的期望；(i) 貶低其他醫生(公允評論除外)。」

[Note 28] Olshansky SJ, Hayflick L, Carnes BA, J Gerontol A Biol Sci Med Sci 2002: "A large number of products are currently being sold by antiaging entrepreneurs who claim that it is now possible to slow, stop, or reverse human aging. … The products being sold have no scientifically demonstrated efficacy, in some cases they may be harmful, and those selling them often misrepresent the science upon which they are based. In the position statement that follows, 52 researchers in the field of aging have collaborated to inform the public of the distinction between the pseudoscientific antiaging industry, and the genuine science of aging"

[Note 29] Consumer Council press release (14 July 2021, CHOICE issue 537):

The Council reports that it sent staff posing as ordinary customers to enquire about these companies' services, and found that the reliability of many of the genetic tests offered still awaits confirmation by clinical research, so the results are open to doubt; that it reminds consumers that predictive testing may be affected by multiple genes or by external factors, and may even produce false positives or false negatives; and that consumers must therefore, before undergoing testing of this kind, seek the advice of a healthcare professional (one's own doctor, for instance) before deciding whether to proceed, so as to avoid paying a large testing fee only to be misdiagnosed or to acquire unnecessary anxiety (our translation from the Chinese original).

Chinese original:

「消費者委員會派員以一般顧客身分，向這類公司查詢服務詳情，發現不少公司提供的基因測試，至今仍有待臨床研究結果證明其可靠性，檢測結果存疑。消委會提醒消費者，預測性的檢測，有機會受多個基因或外在因素影響，甚至出現假陽性或假陰性等問題。故此，消費者在進行這類檢測前，必須尋求專業醫護人員（例如主診醫生）的意見，再決定是否接受檢測，以免繳付大筆檢測費之餘，出現誤診，或增添不必要的焦慮。」

[Note 30] The same, the three points to watch in full (the release makes no claim that these three exhaust everything that ought to be considered):

Before undergoing genetic testing one should seek the advice of a healthcare professional and consider the actual need and the risks; be wary of genetic testing suppliers using the results as an opportunity to sell nutritional supplements or treatment plans, which may be nutritionally unnecessary and which, if taken to excess or misused, may harm health; and before undergoing genetic testing one should consider the question of personal privacy, and should establish who is entitled to know the results, what procedures exist to protect the result data, and how the sample is dealt with after testing is complete (our translation from the Chinese original).

Chinese original:

「進行基因測試前應尋求專業醫護人員的意見，考慮實際需要和風險；慎防基因測試供應商利用測試結果，趁機推銷營養補充劑或治療方案，這些產品在營養學上可能沒有必要，若過量或濫用更有可能損害健康；進行基因測試前需要考慮個人私隱問題，應要知道誰人有權知悉其基因測試結果，有甚麼程序去保護結果資料，測試完成後樣本的處理方法等。」

[Note 31] ATML Healthspan product page (Chinese version): 「此產品沒有根據《藥劑業及毒藥條例》或《中醫藥條例》註冊。此產品作出的任何聲稱亦沒有為進行該等註冊而在香港接受評核。此產品並不供作診斷、治療或預防任何疾病之用。產品效果因人、不同情況而異。」

[Note 32] MedDx product page (Chinese version): 「衰老速度是指生物年齡與實際年齡之間的差異。如果受檢者的生物年齡與實際年齡相近，則衰老速度正常。如果生物年齡小於實際年齡，則老化速度被認為過快。」 English version: "If the biological age is lower than the actual age, the ageing speed is considered too fast." The same company's sample report PDF, page 3: "Because the bio-age is [younger than] the actual age, the aging rate is [slow]"; page 1 example output "Estimated biological age 45yrs" paired with "Aging rate slow"; page 3 definition: "The aging rate: The value obtained by dividing the bio-age by the actual age"

[Note 33] OT&P Healthcare "Longevity health check", the complete list on the Chinese page:

The Longevity physiological function test (including every item in the Longevity physiological function test) / grip strength testing / VO2 Max / strength and agility testing / body composition and bone density scanning (DEXA) / resting metabolic rate (RMR) testing / CBC (complete blood count) / Chemfile 20+ / glycated haemoglobin (HBA1C) / fasting insulin / estimated glomerular filtration rate (eGFR) / lipoprotein (a) / lipid panel / CRP us (high-sensitivity C-reactive protein) / CA19-9 / vitamin D, B12, folate / homocysteine / assessment of oxidative stress and antioxidant status / DHEA-S (dehydroepiandrosterone sulfate) / hormone testing / thyroid profile (TSH, FT3, FT4) / tumour markers (differing for men and women) / upper abdominal ultrasound scanning (USS) / faecal occult blood testing / microscopic urine examination / DEXA body composition, hip and spine scanning (BNS) / doctor's consultation and report (our translation from the Chinese original)

Chinese original:

「Longevity生理機能測試（包括所有 Longevity 生理機能測試項目）／手握力測試／VO2 Max 最大攝氧量／力量和敏捷測驗／身體組成和骨質密度檢查（DEXA）／休息代謝率(RMR)測試／CBC（全血細胞計數）／Chemfile 20+／糖化血色素測試（HBA1C）／空腹胰島素／估算腎小球過濾率（eGFR）／脂蛋白（a）／血脂檢查／CRP us（超敏 C 反應蛋白）／CA19-9／維他命 D、B12、葉酸／同半胱胺酸（Homocysteine）／氧化壓力和抗氧化狀態評估／DHEA-S（硫酸脫氫表雄酮）／荷爾蒙測試／甲狀腺概況（TSH、FT3、FT4）／癌症標誌物(男女不同)／上腹部超聲波掃描（USS）／糞便隱血測試／顯微鏡尿液檢查／DEXA 身體組成，髖部和脊柱掃描（BNS）／醫生諮詢及報告」

The same page also carries 「這些測試可以全面了解你的生理年齡、DNA、一般健康指標、發炎水平、荷爾蒙健康和心血管狀況（包括動脈年齡等）。」 — these tests give a comprehensive picture of your physiological age, DNA, general health indicators, inflammation levels, hormonal health and cardiovascular condition, arterial age among them (our translation from the Chinese original). The English list carries, between "Fasting Insulin" and "eGFR": "Apolipoprotein B — Helps measure the number of low-density lipoprotein (LDL) particles, indicating cardiovascular risk and helping assess cholesterol-related conditions."

[Note 34] Leong DP et al., Lancet 2015: "Further research is needed to identify determinants of muscular strength and to test whether improvement in strength reduces mortality and cardiovascular disease."

[Note 35] Five passages from seller documents: Bryan Johnson's site, "They should not replace or supplement any clinical tests recommended by licensed medical professionals."; MedDx sample report page 8, "※ In this test, the clinical significance of the test results has not been established, and there is still insufficient objective validity that the health-related behaviors that follow are useful." and "※ This test was developed and its performance characteristics determined by GC Genome. It has not been cleared or approved by the Korea Ministry of Food and Drug Safety (MFDS)."; page 5 of the same report, "* Telomere test cannot be used to diagnose diseases or to determine actions related to the treatment of a disease."; the MedDx website terms (Chinese version), 「檢測只屬 身體檢查 ，儘管檢查結果表面上屬正常，還有可能有某些隱藏的疾病在稍後時間才會顯然。如結果異常，應求診醫生給予任何診斷或治療的建議，而醫生費用並不包括在已附給 MedDx及其指定實驗室"金域檢驗"之金額當中。」; and the ATML Healthspan product page (Chinese version), 「此產品並不供作診斷、治療或預防任何疾病之用。產品效果因人、不同情況而異。」

[Note 36] Hauskeller M, Shore L, AMA Journal of Ethics, December 2025 issue: "So, if, like patient AA, we are 56 years old and told that our biological age is actually 65, then all that can possibly mean is that whatever parts of our bodies have been tested are in a worse state than is to be expected for someone our age and that this state is more commonly found in someone who is 9 years older, which is, of course, still a good reason for AA to be concerned about the state of his health and possibly make some changes in his life." "[the results] do not entail that he will develop morbidity or die earlier than would be expected on the basis of his chronological age. Rather, they merely indicate that certain aspects of his physical condition make it more likely that particular health hazards lie in store for him in the future if not addressed." "To be told that one is actually much older than one thought one was might well have a similar effect to being diagnosed as suffering from a terminal illness."

[Note 37] The same: "While these tests have been shown to be fairly reliable indicators of a person's risk of developing certain age-related diseases, we should be very clear that they cannot tell us how old we "really" are" "there is no such thing as a particular person's biological age" "Biological age is a fiction that should, at the very least, be clearly identified as such by clinicians" "Epigenetic testing can, of course, reveal potential or existing health problems and provide valuable information that can be used to improve a patient's health and lifespan." "for this discussion to be possible, Dr B would have to know exactly how the test results were generated, which is harder to determine than it should be because of the proprietary nature of current epigenetic clocks and the economic incentives that accompany them."

[Note 38] PEARL trial: "Within the epigenetic testing results, we saw no meaningful significant changes between groups." Declaration of interests: "GH, VL, AN, MM, SM, AI, and SZ are employees and shareholders of AgelessRx"

#What this article does not state

There is a set of things this article will not say on this subject, because it does not hold enough first-hand material for them. Each is set out below:

• This article will not say how many suppliers exist in Hong Kong, nor what the market price range is. Every supplier listed here comes from what that supplier has published on its own website, and that list is not complete. One whole category sits outside it altogether: whether the health check packages of Hong Kong's private hospitals include biological age, telomere or epigenetic items.
• This article will not say whether epigenetic clock testing is available in Hong Kong. All it can say is that among the suppliers listed here, not one names on its product page which clock it uses.
• This article will not say whether TruDiagnostic supplies Hong Kong. It holds no first-hand material either way, so it states nothing in either direction.
• This article holds no test-retest figure for any named consumer product. The "up to 9 years" in Higgins-Chen 2022 concerns six "prominent" clocks; the paper names no commercial supplier, and says nothing about which version any supplier ships. This is the figure most wanted on this subject and the one that is missing.
• This article holds no empirical material on what consumers actually do after receiving a result. The only relevant material on this subject is a case constructed for teaching purposes (patient AA in the AMA Journal of Ethics), which is a constructed case and not empirical research, so it is not cited here to describe consumer behaviour.
• The telomere reproducibility data cited here come from a research-use international ring trial in 2015, not from performance data on commercial testing in 2026. Whether consumer telomere testing in 2026 does better or worse is something this article has no information on.
• This article holds no authoritative Chinese-language source on biological age or epigenetic clocks. Every scientific paper and every foreign regulatory document on this subject exists in English only, which is why they are cited here from the English originals. (This article does quote Chinese originals, but all of them come from three other kinds of document — the Chinese version of the Medical Council code, the Consumer Council's Chinese press release, and suppliers' own Chinese pages — and none of those three is an authoritative source on biological age testing.)
• On the position of three bodies this article has no material, so it states nothing in either direction: the UK National Institute for Health and Care Excellence (NICE), the Gerontological Society of America, and the work of the UK Medicines and Healthcare products Regulatory Agency (MHRA) on direct-to-consumer testing. "This article has no material" is not the same as "they have published nothing".
• Two professional body documents are known here only to exist, and their contents are not quoted: the American College of Medical Genetics and Genomics (ACMG) revised statement on direct-to-consumer testing of 2016 (the full text is paywalled; its PubMed abstract carries only boilerplate disclaimer text, and substituting an abstract for a position is a way of getting it wrong), and the American College of Physicians position paper Ethical Considerations in Precision Medicine and Genetic Testing in Internal Medicine Practice.
• A press release from the CUHK Faculty of Medicine, linked out to from MedDx's commercial telomere page, is not quoted here. From the link's own title text it appears to concern telomere length and cardiovascular risk in people with diabetes — that is, a clinical patient population, which is a different thing from a consumer product — and this article goes no further than saying so.
• On the SPORTS LAB HK page, this article gives factual description only (product category, published price, listed outputs), because that page has no language version switch and its language version status could not be determined.
• This article does not evaluate, compare or recommend any supplier. Naming a supplier, its published price and its published contents is factual description; pointing out that two of one product's own documents contradict each other is a statement of fact about those two documents.

#Sources

• Ageing "only loosely associated with a person's age in years" and "There is no typical older person": World Health Organization, "Ageing and health" fact sheet (quoted here from the English original; the page also has a simplified Chinese version, which is not the traditional written Chinese this estate uses, so its Chinese is not quoted) (retrieved 2 August 2026)
• First-generation clock (8,000 samples, 82 datasets, 51 tissues, 353 CpG sites; training target chronological age): Horvath S. DNA methylation age of human tissues and cell types. Genome Biol. 2013;14(10):R115 (DOI 10.1186/gb-2013-14-10-r115; PMID 24138928; erratum: Genome Biol. 2015;16:96, DOI 10.1186/s13059-015-0649-6)
• Second-generation PhenoAge (training target: a phenotypic age composed of clinical measures): Levine ME, et al. An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018;10(4):573-591 (DOI 10.18632/aging.101414; PMID 29676998)
• Second-generation GrimAge (methylation surrogates for seven plasma proteins plus a methylation-based estimator of smoking pack-years): Lu AT, et al. DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging (Albany NY). 2019;11(2):303-327 (DOI 10.18632/aging.101684; PMID 30669119)
• DunedinPACE measures pace of ageing rather than age; probes with low test-retest reliability excluded at the modelling stage; the results section of the abstract reports "DunedinPACE showed high test-retest reliability"; a single New Zealand birth cohort; authors' declaration of interests: Belsky DW, et al. DunedinPACE, a DNA methylation biomarker of the pace of aging. eLife. 2022;11:e73420 (DOI 10.7554/eLife.73420; PMID 35029144)
• Repeated measurement of the same DNA sample on six prominent clocks differing by up to 9 years; the principal-component retrained versions narrowing to within about 1.5 years; authors' declaration of interests (the relevant metrics licensed to Elysium Health through Yale University): Higgins-Chen AT, et al. A computational solution for bolstering reliability of epigenetic clocks. Nat Aging. 2022;2(7):644-661 (DOI 10.1038/s43587-022-00248-2; PMID 36277076)
• Reliability differs between methylation probes, and unreliable probes generate an unknown volume of false negatives (350 blood DNA samples): Sugden K, et al. Patterns of Reliability: Assessing the Reproducibility and Integrity of DNA Methylation Measurement. Patterns (N Y). 2020;1(2):100014 (DOI 10.1016/j.patter.2020.100014; PMID 32885222)
• Comparison of 14 clocks against 174 diseases (n = 18,859, Generation Scotland, 10 years of follow-up); 2,436 models; 176 associations passing Bonferroni correction; 9 for the first generation and 162 for the second and third (the two sum to 171, which does not match the abstract's 176; this article reports that as found); clock hazard ratios exceeding the same clock's all-cause mortality association for 27 diseases; 32 of the 176 improving classification accuracy by more than 1%; smaller effects for DNAm telomere length; the paper's own stated limitations: Mavrommatis C, et al. An unbiased comparison of 14 epigenetic clocks in relation to 174 incident disease outcomes. Nat Commun. 2025;16(1):11164 (DOI 10.1038/s41467-025-66106-y; published online 16 December 2025), https://www.nature.com/articles/s41467-025-66106-y (retrieved 3 August 2026)
• Telomere length as a biomarker of overall ageing generating heated debate, with associations with other markers of health equivocal: Sanders JL, Newman AB. Telomere length in epidemiology: a biomarker of aging, age-related disease, both, or neither? Epidemiol Rev. 2013;35(1):112-31 (DOI 10.1093/epirev/mxs008; PMID 23302541)
• Blinded telomere ring trial across 10 laboratories and 3 methods: absolute results not directly comparable, reference ranges not shareable between laboratories; inter-laboratory coefficients of variation averaging more than 20% for qPCR, though the paper states immediately afterwards that the difference is compensated for by a higher dynamic range, and concludes that gel-based and PCR-based techniques do not differ in accuracy; technical variation per laboratory (median of intra- and inter-batch CVs) ranging from 1.4% to 9.5%, with differences between laboratories only marginally significant (P = 0.06): Martin-Ruiz CM, et al. Reproducibility of telomere length assessment: an international collaborative study. Int J Epidemiol. 2015;44(5):1673-83 (DOI 10.1093/ije/dyu191; PMID 25239152; erratum Int J Epidemiol. 2015;44(5):1749-54, DOI 10.1093/ije/dyv171; four commentaries in the same issue)
• Biological age measurements and claims described as "preliminary"; the same passage describing the data as reflecting "our team's rigorous efforts…while formal peer-reviewed validation studies are underway"; the tests described as "experimental and intended solely for research purposes" and as something that "should not replace or supplement any clinical tests recommended by licensed medical professionals"; the whole protocol described as "an experimental clinical research project": Bryan Johnson, "protocol", https://protocol.bryanjohnson.com/ (the page carries no date of its own; its sitemap lists a single page and the whole site is one document; retrieved 3 August 2026; the site is English only)
• Direct-to-consumer tests divided by risk: non-medical, general wellness and low risk uses are generally not subject to FDA premarket review, while moderate to high risk uses generally are; laboratory developed tests (LDTs) typically carry no independent FDA assurance of analytical or clinical validity: U.S. Food and Drug Administration, "Direct-to-Consumer Tests", https://www.fda.gov/medical-devices/in-vitro-diagnostics/direct-consumer-tests (the page contains none of the words biological age, epigenetic, telomere, aging, longevity or methylation anywhere in its text; retrieved 3 August 2026)
• The FDA's itemised classification of "general wellness" claims (ageing, longevity and biological age are not on that list): U.S. Food and Drug Administration, General Wellness: Policy for Low Risk Devices (reissued 6 January 2026, 14 pages) §III, https://www.fda.gov/media/90652/download
• The LDT rule vacated by a federal district court on 31 March 2025 and the FDA reverting to the earlier text on 19 September 2025 (page content current as of 19 September 2025): U.S. Food and Drug Administration, "Laboratory Developed Tests", https://www.fda.gov/medical-devices/in-vitro-diagnostics/laboratory-developed-tests; independently checked against Federal Register document 2025-18239 (published and effective 19 September 2025)
• No record in any of the FDA's three public device databases (510(k), PMA, device classification) contains telomere, epigenetic or biological age: https://api.fda.gov/device/{510k,pma,classification}.json (the databases report themselves as updated on 20 July 2026; retrieved 3 August 2026)
• The content of the TA-65 complaint, the names of the six counts, the content of the first two counts, and the three price tiers in paragraph 13 with the qualification that they come from earlier versions of the website (the original is quoted in the body): U.S. Federal Trade Commission, In the Matter of Telomerase Activation Sciences, Inc., and Noel Thomas Patton (file numbers 142 3103 / C-4644), complaint issued 18 April 2018, https://www.ftc.gov/legal-library/browse/cases-proceedings/142-3103-c-4644-telomerase-activation-sciences-inc-noel-thomas-patton-matter, complaint PDF https://www.ftc.gov/system/files/documents/cases/142_3103_-_telomerase_complaint_final.pdf (21 pages; retrieved 3 August 2026)
• Commercial screening not necessarily evidence-based; the public and healthcare professionals overestimating the benefits of screening; the alternative suggested where a test claims only to offer peace of mind: UK National Screening Committee, "NHS and commercial health screening tests: important considerations" (published 21 November 2023), https://www.gov.uk/government/publications/uk-nsc-commercial-screening-test-considerations/nhs-and-commercial-health-screening-tests-important-considerations (the page contains none of the words biological age, epigenetic or telomere anywhere in its text; retrieved 3 August 2026)
• Diagnostic tests marketed to the public in ways that create patient safety issues, and the exploitation of the worried well: letter to the Secretary of State for Health signed jointly by the Association for Laboratory Medicine, the Royal College of Pathologists, the Institute of Biomedical Science and the Royal College of General Practitioners, 11 September 2023, https://labmed.org.uk/our-resources/news/marketing-of-laboratory-tests-direct-to-the-public.html (retrieved 3 August 2026)
• Information provided by a doctor must be accurate, factual, objectively verifiable and balanced; must not be exaggerated or misleading; the nine limbs of clause 5.2.1.2 in full and the difference between the Chinese and English versions of limb (e) (the English carries the extra "and services"); clause 5.2.4.6; the introduction to clause 24.4 and the five limbs (a)–(e) in full, with the two differences between the Chinese and English versions; the Chinese version stating that the English text prevails; neither version containing any of the words anti-ageing, biological age or telomere: Medical Council of Hong Kong, Code of Professional Conduct for the Guidance of Registered Medical Practitioners (revised October 2022), Chinese version https://www.mchk.org.hk/tc_chi/code/files/Code_of_Professional_Conduct_(Chinese_Version)_2022.pdf; English version https://www.mchk.org.hk/english/code/files/Code_of_Professional_Conduct_(English_Version)_(Revised_in_October_2022).pdf (retrieved 3 August 2026)
• The reliability of genetic testing being open to doubt, false positives and false negatives, the need to consult a healthcare professional first, and three points to watch (the text nowhere mentions biological age, epigenetic age or telomere testing): Consumer Council, CHOICE issue 537 press release, 14 July 2021, Chinese version https://www.consumer.org.hk/tc/press-release/537-dna; English version https://www.consumer.org.hk/en/press-release/537-dna (retrieved 2 August 2026)
• Searches of the Consumer Council website returning 0 results for 生物年齡, 端粒 and 表觀遺傳 (on the same interface 抗衰老 returns 14, 基因檢測 11 and 體檢 105): https://www.consumer.org.hk/tc/search (retrieved 3 August 2026)
• The position of 52 ageing researchers on the anti-ageing industry: Olshansky SJ, Hayflick L, Carnes BA. Position statement on human aging. J Gerontol A Biol Sci Med Sci. 2002;57(8):B292-7 (DOI 10.1093/gerona/57.8.b292; PMID 12145354)
• What a biological age result means; that it does not entail illness or early death; the psychological effect; the proprietary nature of the clocks making interpretation hard for doctors; "there is no such thing as a particular person's biological age" and "Biological age is a fiction": Hauskeller M, Shore L. What Are the Most Ethically Salient Implications of Epigenetic Age Testing? AMA J Ethics. 2025;27(12):E828-833 (DOI 10.1001/amajethics.2025.828; PMID 41411396). ⚠️ The authors are philosophy researchers at the University of Liverpool in the United Kingdom, and this is a commentary published in that journal, not a position statement of the American Medical Association. Cited here from the archived version of 26 December 2025 (full text re-read 7 August 2026): https://web.archive.org/web/20251226051140/https://journalofethics.ama-assn.org/article/what-are-most-ethically-salient-implications-epigenetic-age-testing/2025-12
• Epigenetic testing sub-study (n = 24, 9 women and 15 men) showing no meaningful significant change between groups; declaration of interests naming seven authors (GH, VL, AN, MM, SM, AI, SZ) as employees and shareholders of AgelessRx: Moel M, et al. Influence of rapamycin on safety and healthspan metrics after one year: PEARL trial results. Aging (Albany NY). 2025;17(4):908-936 (DOI 10.18632/aging.206235; PMID 40188830; full text https://pmc.ncbi.nlm.nih.gov/articles/PMC12074816/)
• The trial's lead sponsor being AgelessRx (INDUSTRY) with UCLA as collaborator: ClinicalTrials.gov record NCT04488601; AgelessRx's list price for rapamycin, "As low as $65 / month": https://agelessrx.com/rapamycin/
• MedDx: the $ 3800 price, the further $ 500 for a specialist consultation, 10 ml of blood, a report in about 4 weeks, performed by GC Genome, the CAP accreditation claim (concerning GC Genome's laboratory in Korea), 「如果生物年齡小於實際年齡，則老化速度被認為過快」, and the website terms (naming the designated laboratory 「金域檢驗」, KingMed Diagnostics (Hong Kong) Limited): Chinese page https://meddx.com.hk/telomere-length-testing.html; English page https://meddx.com.hk/telomere-length-testing-english.html (neither page carries an effective or updated date of its own; retrieved 3 August 2026)
• MedDx sample report: page 1 example output "Telomere length 8.2±0.3kb", "Estimated biological age 45yrs", "Aging rate slow", "Average telomere length of same age group 6.8kb", "Your Position Compared to the same age group top 6.2%" and the Interpretation box "margin of error: ±0.3kb"; page 2 "Considering the error of the test method, the telomere length is from [7.9]kb to [8.5]kb"; page 3 "Because the bio-age is [younger than] the actual age, the aging rate is [slow]"; page 4 the average annual shortening of 0.04–0.06 kb and the population-average qualification immediately following it; page 4 two example measurements (24.04.21 8.5±0.3kb / 25.06.04 8.2±0.3kb) and "recommended at intervals of no less than one year"; page 5 four Limitation items; page 8 two statements: https://meddx.com.hk/img/gcg_telorisk_2026.pdf (all 8 pages are in English only; retrieved 3 August 2026, all 8 pages re-read 7 August 2026)
• ATML Healthspan Limited (site brand 「NMN活健齡」): $ 1,850.00, item no. T2001, a product page that says only 「血液檢測」 without naming analyte, laboratory or method, the statutory notice (printed in Chinese and English within the same paragraph on the Chinese page, with the English page carrying only the English half), and the other products on the same page: Chinese page https://www.antiaging.com.hk/shop/biological-age-blood-test/; English page https://www.antiaging.com.hk/en/shop/biological-age-blood-test/; the statutory name appears on its terms page https://www.antiaging.com.hk/en/terms-and-conditions/ (the pages carry no date of their own; retrieved 3 August 2026; ⚠️ the "related products" panel at the foot of that product page changes over time)
• SPORTS LAB HK LIMITED: $800 for a single session and three bundle prices (the page prints them beneath a 「限時優惠」 heading), seven outputs, 45–60 minutes: https://sportslabhk.net/bio-age-test (the page carries no date of its own; retrieved 3 August 2026)
• OT&P Healthcare: the three prices HK$2,200, HK$6,120 and HK$19,800; the itemised contents of the HK$19,800 package (containing no epigenetic, methylation or telomere item); the Chinese page using 生理年齡 where the English page uses "biological age"; the Chinese and English lists differing over Apolipoprotein B: Chinese page https://www.otandp.com/zh-hk/longevity-services; English page https://www.otandp.com/longevity-services (the pages carry no date of their own; the itemised list for the HK$19,800 package is transcribed from the Chinese page; neither version contains any of the words epigenetic, methylation, telomere, 表觀遺傳, 甲基化 or 端粒; retrieved 3 August 2026, the Chinese itemised list re-read 7 August 2026)
• Central Wellness (Central & Stanley Wellness): the "Bio Age" / "Bioyon" promotional content, no price published, and the /service/bio-age/ link in the services menu returning a 404: https://centralwellness.com.hk/service/longevity-medicine-advanced-biotech-testing/ and https://centralwellness.com.hk/blogs/program/bioyon-test-discover-your-biological-age-and-start-reversing-it/ (retrieved 2 August 2026)
• BMS Clinic Limited: 「從先進生物標記檢測開始，包括：氧化壓力 / 激素平衡 / 微量營養素 / 生物年齡」, no test product or clock named, no price published, and an English disclaimer cut off mid-sentence on the page where the Chinese version is complete: Chinese page https://www.bmsclinic.com.hk/medical-anti-aging/; English page https://www.bmsclinic.com.hk/en/longevity-anti-aging-medicine/ (Horvath and Raj 2021 appear in the reference list at the foot of both versions; retrieved 3 August 2026)
• No biological age or telomere product in the published catalogue: Prenetics https://www.prenetics.com/; Cordlife (Hong Kong) https://biotech.cordlife.com.hk/; BGI Hong Kong https://www.hkbgi.com/; Genetrack Hong Kong https://www.genetrackhk.com/; re:HEALTH (all retrieved 2 August 2026). The CircleDNA report catalogue is taken from the archived version of 12 March 2026 https://web.archive.org/web/20260312170414/https://circledna.com/pages/reports; S-CELL's category page headed 「生物年齡測試」 is empty in the archived version of 11 October 2025 https://web.archive.org/web/20251011030945/https://s-cell.hk/collections/biological-age-testing (this article states only what the archived versions show)
• An all-cause mortality hazard ratio of 1.16 per 5 kg reduction in grip strength (139,691 participants, median follow-up 4.0 years, 3,379 deaths) and the authors' own statement that the causal question is unresolved: Leong DP, et al. Prognostic value of grip strength: findings from the PURE study. Lancet. 2015;386(9990):266-73 (DOI 10.1016/S0140-6736(14)62000-6; PMID 25982160)
• The least fit fifth of the population carrying roughly four times the mortality risk of the extremely fit (750,302 US veterans, median follow-up 10.2 years, 174,807 deaths): Kokkinos P, et al. Cardiorespiratory Fitness and Mortality Risk Across the Spectra of Age, Race, and Sex. J Am Coll Cardiol. 2022;80(6):598-609 (DOI 10.1016/j.jacc.2022.05.031; PMID 35926933)

This article was compiled by the editorial team from official, academic and supplier first-hand sources, and every clinical statement carries its source; it is health information, not medical advice.