How reliable is a saliva test for biological age?

DNA methylation tests on saliva measure how strongly certain sites on your DNA are chemically marked. That marking changes with age in a predictable way, which makes it possible to estimate your biological age. The well-known 'skin & blood clock' by Horvath does this via 391 sites in the DNA and shows a high correlation with calendar age. That clock also works in saliva, but was originally developed and validated on blood and skin cells, so saliva is not its primary focus.

A simpler test, measuring only three DNA sites on the ELOVL2 gene, was tested in more than 4,600 people. It performed comparably to more comprehensive models. Two of the authors have financial interests (a patent and shareholding), so treat that finding somewhat more cautiously than independent research.

The biggest practical problem is this: the more accurately a clock estimates calendar age, the less it appears to say about health outcomes. In the best-trained models there was no demonstrable association with premature death. The test then becomes technically more precise, but says less about your actual health. This is a fundamental tension that has not yet been resolved.

Saliva methylation does capture real biological signals. Girls who experienced four or more adverse childhood experiences had, on average, 1.65 years of extra epigenetic ageing, but in boys there was no significant effect. Prenatal smoke exposure was also still visible in the saliva of children at age fifteen. This shows that the clock measures more than age alone, but interpretation remains complex.

For consumer use, an additional limitation is relevant: when little DNA is available, which occurs more often with saliva, the available models systematically overestimate age. A mathematical correction helps, but that method is still in the research phase.