Evidence summary: TruAge Biological Age Test (TruMe Labs)

You collect saliva at home and post it to TruMe's laboratory [3]. The product page lists "9 aging methylation sites" and results "within 1 to 2 weeks after lab delivery" [3]. TruMe says results appear as a report in your online account [4]. Its science page describes a "Comprehensive 10-Page Report" [6]. A sample report and a sample questionnaire-based assessment that TruMe sent us in October 2026 run to 9 and 7 pages.

The report gives one biological age estimate, a percentile against people of the same age in TruMe's reference data, and a sample-quality check [1][3][9]. It also gives general lifestyle guidance [5]. A "Rate of Aging" appears only once you have taken a second test, which means buying another kit. TruMe told us in October 2026 that the second test must be at least 90 days after the first. TruMe also told us that customers who complete an optional health questionnaire get fitness, sleep and body-composition estimates and supplement suggestions. No public TruMe page we reviewed mentions this questionnaire.

The laboratory measures DNA methylation, a chemical mark on DNA, at nine sites in three genes, using bisulfite conversion and Sanger sequencing or pyrosequencing [8][10]. That is the only laboratory measurement.

Everything else is calculated. The biological age comes from a formula: a constant plus a weighted sum of the nine methylation values [8]. TruMe has not published which nine sites it uses or their weights; the developers say these "remain proprietary" [11]. The formula was trained to predict calendar age, and the 2026 study's authors call it a "first‐generation epigenetic clock" for that reason [8]. The percentile and the Rate of Aging are both worked out from this age estimate. TruMe told us, and the sample assessment it sent us shows, that the fitness, sleep and body-composition figures come from questionnaire answers, not from the methylation test.

The product specifications list "Blood or Saliva" as the sample type [3], but every published study of the test that we found used saliva [8][10][11], and we found no blood version described. The 2026 study says saliva has methylation patterns similar to blood, citing a 2025 study [8]. That 2025 study, which did not test TruMe's nine sites, found that first-generation clocks gave broadly similar ages in saliva and blood from the same people, mainly because both track calendar age, but agreed poorly on whether someone read older or younger than that age [13].

TruAge Biological Age Test is currently rated Emerging on Compare [1]. The rating concerns how well its biological-age result estimates calendar age from saliva. It does not cover the Rate of Aging, the percentile, the questionnaire-based figures, disease risk or diagnosis. This page explains that rating. It does not change it.

We rate the TruAge Biological Age Test Emerging for its own biological-age result. There is published, peer-reviewed data on that result, but all of it that we found comes from studies with a TruMe co-founder among the authors, and the formula itself is not public.

• Accuracy in customers. In a 2026 study of 4,260 people who had bought or been given the test between 2020 and 2025, the estimate differed from calendar age by about 5.4 years on average [8]. Older people were "consistently judged to be younger" than their calendar age [8].
• Who did the work. University researchers in Singapore analysed the 2026 data. The data were stored at TruMe, and a TruMe co-founder was an author who maintained them [8][11]. The paper says TruMe "had no role in the interpretation of the data" [8], but with a co-founder among the authors we count the study as developer-involved, not independent. In the sources we reviewed, every published study of this test has a TruMe co-founder among its authors [8][10][11].
• Possible overlap. The developers refined the formula on data from about 1,000 customers [8]. The paper does not say whether those customers are among the 4,260, so the 5.4-year figure may partly describe people the formula was tuned on.
• Health links. In the 2026 study, people who smoked, slept poorly or rated their own health as poor had results that were higher for their calendar age, measured at the same time [8]. These links became much weaker once the authors allowed more fully for age, and the link with poor sleep was no longer statistically significant [8]. Eleven factors, including smoking, sleep, exercise and health status, together explained only 2 to 3% of the differences between people's age-adjusted results [8]. In the sources we reviewed, no study has linked this test's result to later disease or survival.

Our Promising tier asks for a published method and at least one independent replication or large-cohort validation [2]. TruMe's formula is unpublished, and the 4,260-person study, though large, was developer-involved; in the sources we reviewed, no study without a TruMe co-founder has checked the result. So the test does not reach Promising. Our Insufficient tier starts with "Vendor-only claims without published validation" [2], which does not fit, because peer-reviewed accuracy data on this product's own result exist [8]. Two of its other criteria do fit: "Proprietary models without methodological transparency" and "Marketing language exceeds published evidence" [2]. So the test sits close to that line. Our confidence in the Emerging rating is low to moderate. On its own, the evidence for the Rate of Aging would fall in our Insufficient tier.

• "850,000" sites. TruMe's how-it-works page says its platform "analyzes over 850,000 methylation sites" [5], and its science page makes a similar claim [6]. Its own product page lists nine sites [3], and the published studies describe nine [8][10][11].
• "99.9% Accuracy". The home and science pages show this figure [6][7]. We found no source that explains what it measures. The published error against calendar age is about 5.4 years on average [8].
• "Peer-reviewed scientific methodology". The science page uses this phrase [6]. The studies were peer-reviewed, but the formula itself has not been published [8][11].
• How fast you are aging. The product page says the test measures "how fast your body is aging" [3], and the subscription page offers to "track how fast you're aging" [9]. The biological age is a formula trained to estimate calendar age, and the Rate of Aging is the change in that estimate between your earliest and latest test, divided by the years between them [1]. We found no published study that validates that rate as a measure of how fast a person ages.
• Reductions in biological age. The home page promises reductions in biological age from lifestyle changes [7]. We found no published study showing that lifestyle changes lower TruAge results. The studies that reported lower TruAge results tested supplements and had no comparison group (see below) [10][11].

Two small studies, each with TruMe co-founders among the authors, measured TruAge results before and after several months of taking a commercial supplement, a different one in each study, in 42 and 45 people [10][11]. Each study was linked to the maker of the supplement it tested [10][11]. Both reported lower average results afterwards. Neither had a comparison group. The 2021 authors write that their study "is not placebo controlled" [10]. The 2025 authors say placebo effects, regression to the mean (unusual first results tend to be followed by more ordinary ones) and selection bias "cannot be excluded" [11]. In the 2025 study, the first saliva sample was collected in a clinic under supervision and the second at home [11], so the two samples were not collected the same way.

In the 2026 customer study, 755 people had tested more than once. More of their results went down than up, but the authors say these results "should be interpreted with caution" [8]. The authors reported links between some supplements and falling results in this group. The only one that stayed significant after adjusting for age, smoking, sex and general health lost significance once they also allowed for starting values and regression to the mean [8].

In the sources we reviewed, we found no randomised trial that used this test. This page is not a recommendation to take supplements.

The intraclass correlation coefficient (ICC) measures agreement between two tests of the same person, where 1 is perfect agreement. In the 2026 study, tests taken a few weeks apart had an ICC of about 0.45 to 0.5 [8]. In the paper's supplement chart, the ICC never rises above about 0.51, and it falls to about 0.39 once pairs of tests up to 500 days apart are included [8]. The authors say this is lower than for blood-based epigenetic clocks [8]. Very few customers took two tests on the same day, so the authors say same-day agreement is "difficult to determine" [8].

A widely used reading guide calls ICC values below 0.5 "poor" and 0.5 to 0.75 "moderate", judged on the confidence range around the estimate [12]. The paper gives no such range. By that guide, the agreement was poor to moderate. In the paper's chart of repeat results for 100 randomly chosen customers, a few age-adjusted results moved by roughly 20 years between tests plotted close together in time [8]. This is our reading of the chart; the paper does not discuss these cases.

A change of a few years between two TruAge results can therefore come from test variability alone. The Rate of Aging divides that change by the years between tests, so the shorter the gap, the more any variability is magnified. TruMe told us that changes should be read against the test's precision and the time between tests. In the sources we reviewed, we found no published figure, beyond that one agreement estimate, for how much one person's result can vary between repeat tests, and no threshold for a meaningful change.

A TruAge result is one estimate from methylation at nine sites, made by a formula built to predict calendar age, with an average error of about 5.4 years in TruMe's customers [8]. A result younger or older than your calendar age is a gap at one point in time, not a speed of aging. The percentile compares you with people of the same age in TruMe's reference data [1]. TruMe calls this a "diverse global database" [9], but we found no published details of its size or make-up. The test cannot diagnose disease, and TruMe says its testing is for wellness and educational purposes only [9]. The reviewed studies do not show whether using this test to guide decisions improves health.

The 2026 cohort was a convenience sample, "predominantly Caucasian" and unusually healthy (4% smoked), and the authors say they do not know how the findings generalise to other populations [8]. The 2026 paper declares no conflicts of interest, although a TruMe co-founder is an author and its corresponding author is a scientific advisor and board member of a supplement company whose customers were in the cohort [8][11]. The 2021 study analysed customers of a supplement company that had a testing agreement with TruMe [10]. The 2025 study was funded by the company that owns the supplement it tested [11].

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