Direct-to-Consumer Epigenetic Age Tests: Accuracy and Limits

What this page covers

Several companies sell mail-in or clinic-ordered "biological age" tests that read DNA methylation from blood, saliva or cheek swabs. This page looks only at validity evidence: does the test measure reliably, does the result mean what it claims, and does acting on it help? It does not rank or recommend any company or product. Evidence labels: promising but limited, mixed, insufficient evidence, marketing claim. Education only.

How the tests got to consumers

Epigenetic clocks began as research tools. A 2026 review traces their transition from research use into direct-to-consumer testing and notes that their reliance on large numbers of DNA sites and high-throughput technology limits clinical scalability because of cost, complexity and sample processing. [1] Several of the review's authors are founders of companies in the biotech field, so read it with that interest in mind. [1]

Analytical validity: does the test measure reliably?

One study found that technical noise produced deviations of up to 9 years between replicate samples for six prominent epigenetic clocks. A computational method brought most replicates within 1.5 years. [2] Its authors disclosed ties to a company that licenses related metrics. [2]

Sample type matters too. Researchers compared common clocks within the same 83 people (284 samples, ages 9 to 70) across blood and oral samples. They saw average within-person differences of almost 30 years in some clocks and low correlation between blood-based and oral-based estimates. [3] The authors noted growing interest in using saliva and cheek samples in research and commercial settings, and concluded that applying blood-derived clocks to oral samples may not give comparable results. [3]

Label: mixed. Reliability can be improved, but a test that does not report how it handles noise and sample type is hard to trust.

Clinical validity: does the number mean what it claims?

In groups of people, higher clock scores have been linked to later illness and death; see Biological Age Clocks Compared. For a single person it is less clear. A 2025 critique argued that clocks fail to meet common clinical-utility standards compared with established biomarkers, and that using them for individual decisions can be uninformative and potentially harmful. [4] A 2024 review said no consensus exists on how to validate aging biomarkers before clinical use. [5] Reviewers writing in Cell said the lack of standards and consensus on what makes a reliable aging biomarker hinders validation for clinical applications. [6]

Clinical utility: does acting on it help?

Proper evidence not found. We searched PubMed for trials or reviews showing that people who changed behaviors or took products after a consumer epigenetic age result had better health outcomes, and found none we could read. The closest item was a 2026 conceptual article proposing how epigenetic age feedback might support counseling, which states that the idea remains conceptual and requires empirical evaluation. [7] In a randomized calorie-restriction trial, one clock moved slightly, others did not, and the authors said conclusive tests need long-term follow-up on disease and death. [8] A score change is a biomarker change, not a health outcome.

How the FDA describes consumer tests

The FDA says direct-to-consumer tests for non-medical, general wellness or low-risk medical purposes are generally not reviewed by the FDA before they are offered, and that such tests have varying levels of evidence for their claims. [9] It describes analytical validity (accurate, reliable measurement) and clinical validity (the measurement predicts a state of health) as the questions it examines for tests it reviews. [9] Rules change, and this is not legal advice. We did not check other countries or any specific company's regulatory status.

Questions to ask before paying

• Has the exact test been checked for repeat-sample reliability, and on which sample type?
• Which health outcomes was the score validated against, in people like me?
• What would I change based on the result, and is that supported by trials?
• Who funds and owns the clock being used?