We recently published a peer-reviewed paper investigating the best way for providers to monitor testosterone creams and gels.¹ The conclusive winner of the investigation is serum testing. This seems an odd paper from a lab that has no intention of performing serum testing, so I wanted to explain our rationale. While we are a testing lab, we are passionate about promoting best practices in our industry, in this case, the practices of monitoring transdermal testosterone replacement therapy (TRT).  

Recently, I attended an educational event during which an experienced educator gave a terrific lecture on the benefits of TRT in men. There was a reference to using serum testosterone (measuring both free and total) to follow TRT, with one exception. While the instructor did not call this out verbally, the slide subtly encouraged the learners to use saliva testing in cases where the testosterone was a transdermal cream. I set up a call with the instructor a few weeks later and had a great conversation about many of the studies in our publication. While saliva testing wasn’t the instructor’s personal practice, the recommendation remained in the lecture as a vestige of historic teaching within some integrative medicine education. At this same educational event, if you moved through the hall of exhibitors and asked the hormone testing labs their view, most would also affirm the use of saliva testing in these scenarios. This has a huge impact on practitioner practices.  Effective TRT requires effective dosing, which can be greatly impacted by which lab tests are used, especially since the response in saliva is dramatically higher than the response seen in serum (and urine). 

The purpose of our paper was to examine the existing body of evidence to determine whether the practices of monitoring TRT taught in integrative medicine can hold up to an evidence-based evaluation, now that more data has been published in the last 5-10 years. Our conclusions about serum and saliva in the paper are based on an in-depth review of a large body of peer-reviewed research studies, clinical guidelines, and real-world evidence from laboratory databases. Lastly, this review paper also includes new clinical data. These data evaluate hundreds of female patients on transdermal testosterone. The women showed high levels of salivary testosterone and yet do not report higher levels of high androgen symptoms compared to similarly treated women who do not have high salivary testosterone levels. 

I hope this paper serves as an opportunity for our industry to reapproach this topic. Our industry will serve patients better if we can narrow in on standardized practices for effective dosing for men and women. We must understand the laboratory science well to do this accurately. Our publication is our attempt to add productively to this conversation, beginning with testosterone where there is the most clinical data to overlay with lab data (compared to estradiol and progesterone).

Reconsidering saliva testing in TRT monitoring: A personal journey through the evidence

At last check, many labs in our industry promote saliva testing as THE way to monitor transdermal hormone creams/gels, including testosterone. This seems a genuine response to much of the teaching within certain integrative medicine circles throughout the past 20 years. As the owner of a lab that offers urine testing and only uses saliva testing for cortisol assessment, my opposing view may seem a biased perspective. As it turns out, there was a day when I affirmed the saliva-based approach for monitoring hormone creams/gels. I want to walk you through the process that led to my change in thinking and encourage others to examine these claims anew. This blog and our published paper are an open invitation for reconsideration but also for good-faith responses and rebuttals, be they informal or preferably with published data. 

It is well documented and uncontroversial that when patients use transdermal testosterone, estradiol, or progesterone creams/gels, the response in saliva is much higher than what is seen in serum or urine.^2-5 But which results best represent clinical reality? I was trained within the integrative medicine community 20 years ago to use saliva testing when patients were taking exogenous, transdermal hormone creams and gels. Like many providers, I trusted what was being taught by authoritative voices without always independently reviewing the underlying evidence (a common and understandable tendency that can make it challenging to later revisit and revise long-held beliefs). I learned these arguments well enough that I began to teach them to providers myself.

For monitoring hormone creams and gels, the saliva-based approach argues that free hormones measured in saliva act as our best surrogate representing tissue (e.g., brain, breast, muscle, endometrium) exposure even if serum levels do not increase. I made this argument many times after which a long-time colleague (also a co-author on our publication), who himself owned a saliva lab, challenged me on this view, arguing something that seemed anathema to me at the time — saliva values, he argued, in this HRT context have no clinical relevance. 

After resisting this challenge for a time, I decided to dive into the literature to see if I could make sense of things, starting with the hormone that seemed to have the most clinical data — testosterone.  5-10mg of transdermal testosterone produces consistently supraphysiological saliva results in men (if tested 8-24 hours after therapy). On the other hand, it takes at least 25mg (usually 50-100mg) of transdermal testosterone to significantly increase male serum testosterone levels. Salivary results imply that nearly 100% of the applied testosterone is absorbed, while serum (and urine) imply that about 10% is absorbed. So, what was I looking for in my investigation? I looked for any studies I could find where clinical parameters were reported for patients using different dosages of transdermal testosterone. I was testing two competing hypotheses: 

Hypothesis #1: Saliva more accurately represents tissue uptake of hormones. This would mean that doses below 25 mg (5-20 mg) should behave as “male doses” and result in clinical responses similar to those seen with other TRT scenarios like injections. 50-100mg should consistently behave as a very high dose in men with clinical outcomes implying an overdose. If this hypothesis is correct, I should find studies that imply that serum underestimates clinical impact when taking testosterone gels. 

Hypothesis #2: Serum levels more accurately represent tissue uptake of hormones, and doses >25mg would be needed to see typical male responses to TRT. If serum levels don’t increase significantly, clinical changes should not be expected. Additionally, 50-100mg should not have evidence of excess testosterone in men unless serum levels increased excessively. Analogously, 5-10mg of TRT in a female should not include reports of high androgen symptoms unless serum results are elevated. 

I honestly expected to find mixed results. I expected to find some studies that supported both hypotheses, and I was curious which one would have more support in the >100 studies I read. To my surprise, I was unable to find a single study where the clinical response exceeded what the serum results implied about testosterone exposure. Every study I found reporting clinical outcomes aligned, to at least a first approximation, with serum levels, including for bone mineral density, lean body mass, sexual function, hematologic parameters (erythrocytosis), and endocrine feedback (luteinizing hormone suppression). Additionally, when treatment with transdermal testosterone failed to significantly increase serum testosterone levels, clinical improvements weren’t experienced. For example, in a study with hypogonadal men where treatment with testosterone gel failed to significantly increase serum testosterone levels, no appreciable changes were reported in sexual function, mood, bone mineral density, or lean body mass.⁶  

I wanted to dig further, so I turned my attention to female studies. I looked at 5-20mg doses used in women and 25-100mg doses used in transmen. These studies created a different angle on the same question. Saliva testosterone levels imply that 5-10mg is a “male dose” of testosterone gel. If you give a male dose to a female, you expect to see high androgen symptoms. Once again, I found that studies consistently concluded that until female serum levels rose above the female range, high testosterone symptoms weren’t found. This was true despite the fact that many of these women had saliva levels exceeding typical male levels of testosterone.  This was also echoed by studies with transmen, where a 50mg testosterone gel that significantly increases serum testosterone levels into the male range correspondingly produces androgenic results as expected. On the other hand, symptoms associated with high testosterone in women were not seen in studies using 5-10mg or in any studies not also showing elevated serum levels. 

To sum, these were the key findings from my review of the literature: 

• Serum testosterone correlates with clinical outcomes. Changes in serum testosterone during transdermal TRT consistently aligned with changes across multiple clinical endpoints, including bone mineral density, lean body mass, sexual function, erythrocytosis, luteinizing hormone suppression, and androgen-excess symptoms.
• Clinical effects require raising physiological serum levels. Clinical effects were observed only when serum testosterone levels were raised into the normal range (typically using transdermal TRT doses of 25-100mg gel in men or 5-10mg gel or cream in women). When serum testosterone levels did not adequately increase, clinical effects were absent.  
• Elevated saliva levels lack clinical correlation. Not a single published study shows that the elevated saliva responses to transdermal TRT are accompanied by corresponding changes in any clinical outcomes or symptoms. 
• New data confirm no relationship between elevated saliva and symptoms. In addition to reviewing the existing published literature, our publication also presents new data from hundreds of women on transdermal TRT, showing that supraphysiological saliva testosterone levels did not correspond to greater androgenic symptom severity, and no correlation was found between saliva testosterone levels and any of six androgenic symptoms (excess facial or body hair, acne, irritability, scalp hair loss, low sex drive, and oily skin).¹ Collectively, the evidence suggests that saliva does not reflect tissue-level hormone activity during transdermal TRT.  

My former view that saliva levels best represented tissue exposure of hormone creams and gels crumbled. Within this same chapter of my life, I started Precision Analytical and began work on the DUTCH Test. That is a story for another day, but part of what inspired DUTCH was to find an improved method to assess hormones. We commercialized urine assays and salivary cortisol but left our salivary estradiol, progesterone, and testosterone in the research category. It may surprise some to find our team publishing a paper that concludes that serum testing (which our lab does not perform) is the ideal choice for a particular scenario. Our intention is to help our industry improve our reputation by pursuing truth collectively, wherever that may lead. In this case (TRT monitoring), our evidence-based conclusions include using serum as a primary test, leaving DUTCH as a very useful complementary test. 

Isn’t saliva testing more commonly used for estradiol than for testosterone creams/gels?

There are providers and labs that state that “saliva is preferred for transdermal hormones.” In actual practice, our experience has been that providers tend to use saliva more for  estradiol and progesterone monitoring and less so for testosterone monitoring. This is likely because they have found serum to work adequately for TRT and have found the low testosterone doses associated with saliva monitoring to be ineffective. Salivary monitoring of transdermal estradiol remains more common. 

Estradiol saliva values tend to be elevated above serum/urine with creams and gels but not to the same degree compared to testosterone. This seems to result in dosing below what has been shown to be clinically effective but not dramatically lower, as with testosterone. However, if an analysis analogous to the one applied in our TRT publication was repeated with estradiol, the findings would be similar. We have done this analysis but have not published that work. Clinical outcomes related to estrogen deficiency (vasomotor symptoms, vaginal atrophy, and bone mineral loss) fail to improve beyond placebo at concentrations of estradiol gels that elevate saliva levels. These elevated saliva estradiol levels imply high tissue levels, but we are unable to find any clinical data that is congruent with the saliva picture. In contrast, serum and urine levels seem to be more consistent with clinical changes. 

Ultimately, for providers to use saliva to monitor estradiol creams and gels, we need data that affirms the clinical reliability of these relatively high values. In light of what we found with testosterone, a strong case would need to be made for salivary estradiol. One of the goals of our publication is to serve as an invitation for evidence. Currently, there are no peer-reviewed studies that lay out a case for using saliva for estradiol, even at a very basic level. A pragmatic approach I see often is that providers test patients 24 hours or even longer after the last dose. This often provides elevated values but providers can still justify the process because they have missed the fact that saliva estradiol is extremely high for most of the day before that point. A sample collected between doses (12 hours after last application) should represent overall exposure well. Providers typically avoid samples in this time frame because the values are disturbingly high.

What does our publication conclude about serum testing? 

While serum has limitations, it is a proven option for baseline testosterone evaluation as well as for monitoring TRT, including testosterone creams and gels. We evaluated the numerous studies looking at changes in serum testosterone levels and clinical outcomes during transdermal TRT and found strong alignment. We were not able to find studies where serum seemed to underestimate clinical efficacy leading to overdosing. 

It is worth stating that when using serum for estradiol gels and creams, there is real potential for overdosing. This is not because estradiol doesn’t show up in serum but rather because the daily up-and-down pattern can be faster than with testosterone. It is common for a provider to give an effective dose of estradiol and see a serum estradiol lower than what they are trying to achieve because the value returns close to baseline between the administration and the test, even if the levels are adequate in between. Serum is not a great match for monitoring estradiol creams or gels. DUTCH may be a better option in these cases by representing most of the day instead of just a moment in time. 

What does our publication conclude about saliva testing?

For baseline testing, saliva’s accuracy depends highly on the methodology. Most assays that are not LC-MS/MS have serious accuracy questions, while well-developed LC-MS/MS assays have been proven effective. If you want to use a saliva assay for testosterone, try to find one using LC-MS/MS with solid validation data, hopefully including serum correlation data.

When patients apply transdermal testosterone creams and gels, saliva values become elevated well above the rise seen in serum or urine. We were unable to find a single study that implied that these on-therapy saliva values change in alignment with clinical outcomes. Consistent with the literature, new data included in our publication showed that women on transdermal testosterone with high saliva values did NOT show an increase in high testosterone symptoms.¹  At this time, there is no evidence-based rationale to use saliva testosterone testing as a surrogate for serum during transdermal TRT monitoring in men or women.    

What does our publication conclude about DUTCH Testing?

Studies show that urine testosterone is correlated to serum testosterone at baseline as well as with transdermal testosterone use (except in individuals with a UGT2B17 deletion, which leads to falsely low levels of testosterone in urine). Testosterone metabolites also influence responses to TRT and its effects on clinical outcomes in hypogonadal men. While serum testing is better suited than urine for assessing testosterone deficiency before initiating TRT, urine androgen testing is a useful adjunct to serum testing during transdermal TRT, to obtain additional complementary information on hormone metabolism relevant for TRT outcomes.  Providing information on estrogen production and metabolism, as well as thorough evaluation of the HPA axis, adds to the complementary value of DUTCH in the context of TRT evaluation. 

Do the findings in our review article extend to progesterone and/or estradiol?

Simply put, estradiol behaves somewhat analogously to testosterone while progesterone behaves somewhat differently. Serum estradiol scales up similarly with transdermal therapy doses, but there is one major difference. The up-and-down pattern with creams and gels is fast enough that serum is difficult to use and overdosing can occur if providers miss the peak levels and continue to increase the dose.⁷ This is an area where the DUTCH Test may really excel as urine captures levels over time. 

Serum progesterone levels do not scale up predictably as transdermal doses increase, so conclusions about testosterone and progesterone are not easily interchangeable.^8,9 As a complement to our publication, see our video series on transdermal hormone monitoring for a visual walk-through of the data on estrogen, progesterone, and testosterone. 

What do we hope is the response to this publication?

Interestingly, one reviewer of our paper recommended not publishing our work because they claimed that it is universally known that serum is the standard and our publication was not necessary. Clearly, in the space of integrative medicine, these questions are not considered “settled science.” 

When providers are taught to follow saliva testing to optimize hormone creams and gels, it can lead to quite different dosage uses. These doses may be ineffective and certainly not optimized if clinical changes are not related to lab values. This is critically important as the conversation of best practices evolves in our industry. 

My hope with this publication is that the fraction of our industry that has been taught that saliva is the best way to monitor patients in these scenarios will take a fresh look at the clinical data. If the saliva-based monitoring approach, that I once held, is correct, there should be outcome studies where clinical responses imply that serum underestimates clinical efficacy. Our publication is attempting to thoroughly address this issue as it relates to testosterone. 

We welcome those who wish to add to this dialogue. We invite studies we may have missed or data sets that might add clarity to answer these critical questions. Serum, saliva, and urine testing all have scenarios in which they are clinically useful and those where other tools might provide more clinical accuracy. Our hope is to encourage continued exploration of these questions.

Read the full study here and our DUTCH Article on the key takeaways.

References

1) Newman MS, Smeaton J, Gillson G, Jaferi A (2026). Alternatives to serum testing for transdermal testosterone monitoring: a review of clinical data and correlation with clinical response. Frontiers in Reproductive Health 8:1804311.

2) Thieme D, Rautenberg C, Grosse J, Schoenfelder M. Significant increase of salivary testosterone levels after single therapeutic transdermal administration of testosterone: suitability as a potential screening parameter in doping control. Drug Test Anal. 2013;5(11-12):819-825. 

3) Krebs A, Clement HW, Zimmerer J, et al. Transfer of Topical Testosterone to Subcutaneous Microdialysate, Blood and Saliva in Healthy Young Men. Exp Clin Endocrinol Diabetes. 2019;127(5):289-294. 

4) Ahern T, Adaway J, Monaghan P, Trainer P, Keevil B. Can salivary testosterone be used in the monitoring of men using transdermal testosterone replacement therapy? In: Endocrine Abstracts. Vol 49. Bioscientifica; 2017. 

5) Zava DT. Topical Therapy with Estradiol, Progesterone and Testosterone and their Distribution in Saliva, Capillary Blood, Serum, and Urine. Townsend Letter. Published online January 1, 2021.

6) Brockenbrough AT, Dittrich MO, Page ST, Smith T, Stivelman JC, Bremner WJ. Transdermal androgen therapy to augment EPO in the treatment of anemia of chronic renal disease. Am J Kidney Dis. 2006;47(2):251-262.

7) Newman, MS. Transdermal hormone monitoring update – Part 3: Estradiol. DUTCH Test website. https://dutchtest.com/tutorials/transdermal-hormone-monitoring-update-part-3-estradiol

8) Newman, MS. Transdermal hormone monitoring update – Part 2: Progesterone. DUTCH Test website. https://dutchtest.com/tutorials/transdermal-hormone-monitoring-update-part-2-progesterone

9) Newman, MS. Transdermal hormone monitoring update – Part 4: Testosterone. DUTCH Test website. https://dutchtest.com/tutorials/transdermal-hormone-monitoring-update-part-4-testosterone