How reliable is a saliva test for biological age?

A validated test for biological age via DNA methylation in cheek swabs achieves an average margin of error of approximately 4.4 years. By comparison, the same test on blood comes in at approximately 4.0 years. That difference is small, but it is not zero. Results were reproducible across multiple laboratories, although each lab must first validate the protocol internally before it can trust the results it produces.

A practical problem that receives little attention in commercial saliva tests is blood contamination. Blood contains far higher concentrations of most substances than saliva does, meaning that even a small amount of blood entering the sample can significantly distort the measurement. No reliable method yet exists to consistently detect this. This means that sample quality has a major influence on the result.

Outside of age estimation via DNA methylation, the picture for saliva tests is considerably less optimistic. Saliva generally contains low concentrations of signalling molecules compared with blood. In research into endometriosis diagnostics, for example, only 3 usable signalling molecules were found in saliva, compared with 13 in blood. For neurodegenerative conditions such as Alzheimer's disease and Parkinson's disease as well, saliva is so far much less well supported than blood or cerebrospinal fluid.

One study in 230 people found that a particular protein in saliva is elevated in Alzheimer's patients and correlates with cognitive decline. That is an initial, interesting indication, but it concerns a single marker in a single study. Whether this tells us anything about biological ageing has not been demonstrated. Other saliva biomarkers for brain diseases have not yet been validated for diagnostic use.