Evidence summary: NOVOS Age (NOVOS Labs)

NOVOS Age is an at-home test that uses a few drops of blood from a finger prick, dried on a card and mailed to a laboratory [3]. NOVOS lists four results: a Biological Age, a Pace of Aging based on DunedinPACE, "Organ Ages" for 11 organs and systems, and a telomere length estimate, together with a written report and an app [3][4]. NOVOS says a laboratory partner in a certified US laboratory processes the sample, but the pages we reviewed do not name that laboratory [3]. NOVOS describes the results as informational and not intended to diagnose disease [3].

The laboratory reads DNA methylation, a set of chemical marks on DNA, from the blood sample. NOVOS states that it does not sequence your DNA [3]. NOVOS describes its four results as calculated from this methylation analysis [3]. None is a direct measurement of an organ, of telomeres or of how long someone will live.

• Pace of Aging is attributed to DunedinPACE, a published method described below [3][13].
• Organ Ages are attributed to a "Yale-developed Organ Age algorithm" [3]. The app screen shown on the product page says the results are based on the DunedinPACE clock and a "SymphonyAge algorithm" [4]. SYMPHONYAge is the name TruDiagnostic, another test company, uses for its organ-system scores, which it links to Systems Age, a published Yale method with 11 body-system scores [10][22]. The NOVOS pages we reviewed do not name the published method or the version they use.
• Biological Age: the pages we reviewed do not say which algorithm produces this result [3]. The app screen names only DunedinPACE and SymphonyAge for the results it shows, without saying which result comes from which [4]. We could not confirm which algorithm, if either, produces the Biological Age.
• Telomere length: NOVOS describes this result as follows: "Algorithmic estimate of telomere length based on epigenetics." [3] The pages we reviewed do not name the algorithm. For one published methylation-based estimator, the estimates correlated only moderately with telomere length measured in the laboratory [26]. We could not confirm whether NOVOS uses that estimator.

DunedinPACE was developed and tested in blood drawn from a vein [13]. In a small group of children, vein blood dried on a card gave the same average DunedinPACE value as processed blood samples, though individual values matched less closely (correlation 0.72) [23]. In 20 older adults, capillary blood (the kind a finger prick gives) kept liquid in small tubes gave similar average values to vein blood [24]. A study of finger-prick dried blood spots in Malawi could calculate DunedinPACE but did not compare it with vein blood [25]. In the sources we reviewed, no study compared finger-prick dried blood spots with vein blood for DunedinPACE in the same people.

NOVOS Age is currently rated Promising on Compare [1]. The rating concerns the Pace of Aging result, which uses the published DunedinPACE measure. It does not cover the Biological Age, the Organ Ages or the telomere estimate. This page explains that rating. It does not change it.

• DunedinPACE. This measure was built from 20 years of repeated health measurements in 1,037 people born in Dunedin, New Zealand, in 1972–73 [13]. In other cohorts, faster scores were linked to more disease, disability and death [13]. The validation in the development paper was carried out by the measure's developers, in cohorts that were mostly of white European descent [13]. The developers report that other research teams have since published more than 500 reports using DunedinPACE [15]. We also opened a 2026 study by researchers who did not develop the measure: in about 1,000 older adults in Berlin, with 90 deaths during follow-up, DunedinPACE had the strongest link with death among 14 aging measures, although the combined models' ability to predict who would die was modest [16]. Our methodology lists DunedinPACE as a measure of the pace of aging, linked to mortality in cohort studies [2]. Its inventors licensed it to a commercial entity [13], and the developers name TruDiagnostic as its exclusive commercial licensee [14][21].
• Organ Ages. The published Systems Age method trained 11 system scores on two US cohorts and then tested them in separate cohorts, where the scores were associated with related diseases and aging measures [22]. The developers ran these tests themselves [22]. The authors say "variations" of the scores exist, and two of the inventors have received consulting fees from TruDiagnostic [22]. The authors also note there is no gold-standard way to assign biomarkers to body systems [22]. Because the NOVOS pages we reviewed do not name the published method or version, we could not confirm that its Organ Ages match the published scores [3][22]. NOVOS states that Organ Ages are not a clinical assessment of organ health [3]. A system score is not an organ diagnosis.
• Biological Age and telomere length. Without the name of the Biological Age algorithm, we could not assess the evidence for it [3]. The telomere result is an estimate, as described above [3][26].
• NOVOS Age itself. In the sources we reviewed, we found no published study of NOVOS Age's own accuracy or repeatability: its finger-prick collection, its laboratory processing or its reported scores [3]. NOVOS quotes a repeatability score above 0.96 [3]. DunedinPACE's developers reported scores of 0.96 to 0.97 when the same research blood samples were run twice on the same type of array [13]. That figure describes the method in research data, not a published measurement of NOVOS Age.

Our methodology's Strong tier asks for independent replication across populations and laboratories, open methods and demonstrated value beyond standard risk factors [2]. It also rates a commercial test on the evidence for that product's own output, not for the method behind it [2]. In the sources we reviewed, the development and main validation of DunedinPACE and Systems Age were carried out by their developers [13][22]. Other research teams use DunedinPACE widely [15], and we opened one such study of mortality [16], but we found no published validation of NOVOS Age's own results. Methods are only partly open: code for the published DunedinPACE and Systems Age methods is public [13][22], DunedinPACE's code page limits it to research use [14], and we could not confirm which algorithms and versions the report uses [3][22]. Of the four results, the Pace of Aging has the most published evidence behind it; we could not assess the Biological Age, and we could not confirm that the Organ Ages and the telomere estimate match published methods. Our proposed reading is that these gaps explain why the test sits below Strong under the current criteria. This is our editorial interpretation, not the original rating rationale.

In a two-year randomised trial, 220 adults without obesity were assigned to calorie restriction or their usual diet [17]. A later analysis of stored blood samples found that DunedinPACE slowed slightly on average in the restriction group compared with controls [17]. The effect was small, and individual results overlapped substantially between groups [17]. Two other clocks, PhenoAge and GrimAge, did not change significantly [17]. The authors state that whether this translates into long-term clinical benefit is unknown [17]. This trial did not test NOVOS Age or using it to guide care [17]. This summary is not a recommendation to restrict calories.

NOVOS Labs also sells dietary supplements. Its test page features customer stories that link lower Pace of Aging scores to its products, often alongside exercise and diet, and its website describes a small company pilot that measured Pace of Aging before and after six months of its supplements, with no comparison group [3][5][6]. Without a comparison group, a study like this cannot show what caused a change in scores. We do not assess supplements on this page.

A 2026 study found that many clocks give near-identical results when the same sample is run twice [20]. When small groups had blood drawn several times within hours or days, around events such as a meal, a stress test or travel to altitude, scores varied far more than between repeat runs of one sample [20]. Without control groups, the authors could not tell whether the events or simply the timing of the draws caused this [20]. In this analysis DunedinPACE was technically precise but less stable between draws [20]. Adjusting scores for estimated immune-cell proportions generally made them less stable between draws, and made no clear difference for DunedinPACE [20]. NOVOS says its test uses "published and patented" methods to account for 12 immune cell types, but the pages we reviewed do not cite them or explain how they enter the reported results [3]. The study was small, mostly involved adults aged 18–27, and used older array versions [20]. Two of its authors are Systems Age co-inventors who have received consulting fees from TruDiagnostic [20].

NOVOS suggests retesting after 6 to 12 months under comparable conditions [3]. None of the sources we reviewed sets a threshold for a meaningful change in one person.

DunedinPACE's developers describe its scale against an average of one year of biological aging per calendar year in the cohort it was built from [13][14]. The NOVOS app restates a score as months of aging per calendar year [4]. The score is an estimate: in the development cohort it tracked the measured pace of aging closely but not exactly (correlation 0.78) [13], and it varies between blood draws [20]. Reading a single score as an exact personal rate goes beyond what the sources we reviewed show. It is not a lifespan forecast.

These scores show population-level associations with health outcomes [13][16][22]. Experts on aging biomarkers separate whether a measurement is technically reliable, whether it predicts outcomes, and whether using it improves health [21]. The reviewed studies do not show whether using this test to guide decisions improves health. We do not treat this test as a diagnostic tool.

For current price and availability, see Compare. If you spot an error, please tell us.