Limitation and Caveats on Saliva Testing

by Mark Newman, MS

Our last educational email was about the limitations and caveats of our dutch testing (as well as urine testing in general). In this post, we will look at saliva testing. This entire series, which we are calling “The Caveat Series,” can be summed up by viewing these three videos.

• Saliva Testing Caveats (today’s focus)
• Serum Testing Caveats
• Dutch and 24hr Urine Testing Caveats

Those who favor saliva testing often point to these apparent advantages.

• Convenient (true, saliva testing is fairly easy to do)
• Allows for the testing of diurnal free cortisol (true, but you NEED cortisol metabolites offered by DUTCH to make confident clinical conclusions)
• Allows for the testing of “free” sex hormones like estradiol (this is technically true but practically a massive deficiency of saliva testing…see below)
• Allows for the monitoring topical hormones (this is a complicated issue, but the data does not support this conclusion)

It is my position that salivary testing has very limited utility given the current state of technology. It can be used to test sex hormones as well as cortisol, but clinical and analytical issues create problems on multiple fronts. The following is a listing of issues to be aware of when using salivary testing (each is discussed in detail in the Saliva Testing Caveats video above):

• Cortisol testing is accurate for free cortisol, but many times can be clinically misleading without concurrent measurements of metabolites.
• Most assays use immunoassays which are prone to cross-reactivity issues (taking 7-keto-DHEA can artificially raise testosterone and DHEA levels, taking estriol can falsely increase estradiol, etc.)
• Estradiol testing is largely of poor quality. We just completed a study that showed very disappointing results for salivary testing. Progesterone and estradiol were tested throughout the menstrual cycle for 6 different US labs. Serum and urine testing was completed at the same time. Serum and urine results correlate beautifully. Saliva testing was good for progesterone for ½ of the labs. Estradiol was marginal for one lab, quite poor for three and abysmal for two. This is obviously a fairly negative conclusion, but if you watch the video and examine the data, you’ll see that it is a very fair assessment of the data. We continue to test additional individuals (at great pain and expense) and continue to see similar issues repeated with each. Salivary estradiol simply does not show proper ability to differentiate between “normal” and “low” for estrogens.
• Sublingual, vaginal and transdermal hormones as well as oral progesterone are not monitored well (see video for details).
• Of the above issues, DUTCH offers increased accuracy. Sublingual hormones are difficult with DUTCH testing as well.

Transdermal hormones continue to be a complicated topic with more research needed to assess whether dosages can be effectively adjusted using urine testing with particular products. To date, DUTCH testing MAY reflect dosing well in some scenarios and may not in others. Pragmatically we see some providers with dutch levels significantly increased (even well into the premenopausal range) with reasonable doses of estradiol (<2mg) and progesterone (<50mg). Other providers see little or no increase with similar dosing. Clearly there is still more to learn about this route of administration, but the data is clear that salivary values are not clinically reproducible and do not reflect systemic hormone exposure.