Issues with Saliva Hydration Testing

Based upon anecdotal observations, assessment of worker hydration status through field saliva testing has become more prevalent over recent years. This 2-part article addresses the key questions regarding this method, research findings and observations/feedback from the field.

Let’s start with what’s being measured - what are saliva hydration scores?

The concentration of dissolved particles in saliva for a given sample volume (osmolarity) or given sample mass (osmolarity) is what’s being measured, with osmolarity and osmolality used interchangeably. In basic terms, a lower concentration of particles in the saliva sample (lower score) reflects a greater fraction of fluid in saliva and therefore, a more “hydrated" sample.

Figure 1. A portable osmometer (MX3 Diagnostics)

Saliva as a potential marker of hydration status is not new (Friedberg and Doyle, 1959). Yet, it has been plagued by variability and limitations associated with laboratory-based analysis. The commercial release of the portable MX3 Hydration Testing System in 2018 (Figure 1) addressed the need for laboratory testing, however, reports of variability persist. Here’s a summary of the research findings regarding saliva as a marker of hydration:

“Given the inherent variability and profound effect of oral intake, use of saliva osmolality as a marker of hydration status is dubious” (Ely et al., 2011).

"…… its (saliva osmolality) large intra- and inter- individual variability limited its predictive power and sensitivity, rendering its utility questionable within a universal dehydration monitor” (Taylor et al., 2012).

“Saliva might be an effective index to evaluate hydration status but seems to be highly variable” (Villiger et al., 2018).

“The diagnostic utility of saliva osmolality is affected by oral artefacts such as recent fluid consumption and factors influencing saliva flow rate which include neural control and inherent inter-individual variability. It is therefore of limited value in the assessment of hydration status” (Lacey et al., 2019).

"Considering the poor reliability and large number of confounding factors associated with salivary variables, the use of this technique is questionable” (Barley et al., 2020).

And finally, the following is from research assessing the reliability of the MX3 Hydration Testing System:

“………. the categorization of hydration status according to SOSM (Saliva Osmolarity) from the MX3 HTS (Hydration Testing System) should be interpreted with caution given inter- and intra-individual variation in baseline saliva SOSM (Saliva Osmolarity) and potential interference from recent food, drink, gum, or tobacco consumption” (Winter et al., 2024).

While, worksite staff that use saliva as a hydration marker via the MX3 system may not be aware of the aforementioned quotes and associated research, in most cases, variability of test scores are not new to them. For example, Figure 2 represents back-to-back tests (two tests performed within ~40 seconds on a worker, no fluid consumption in the 5 mins prior or in-between tests) with the same MX3 osmometer in a climate-controlled clinic. 

Figure 2. Saliva osmolarity values from two tests performed within ~40 seconds on a worker, no fluid consumption in the 5 mins prior or in-between tests, with the same MX3 osmometer in a climate-controlled clinic

Test 1, 115 milliosmoles (mOsm), test 2, 63 mOsm. Is this worker considered hydrated or moderately dehydrated? How do you interpret these scores? The paramedic that performed these tests stated that they would routinely perform 3 or 4 tests in quick succession on a given worker and calculate an average of the scores to account for the observed variability. It’s this variability that has led some organisations with large field-based workforces to discourage or even ban the use of saliva hydration testing on their sites. So, what’s contributing to this variability?

Saliva collection in the laboratory and the field

Laboratory based studies utilise desktop osmometers that generally test 20 microlitre samples. That’s a sample of just 1/50^th of 1mL. Prior to the availability of a portable osmometer, saliva had generally been collected via a swab placed under the tongue for two minutes (salivette method) or by accumulating saliva in the mouth for two minutes prior to expelling (spitting) into a tube (expectoration method). Both the salivette and expectoration sampling methods produce adequate saliva to sample the 20 microlitres from and they also produce comparable results (Ely et al., 2013).

Field testing is achieved with a portable, handheld osmometer that does not require expectoration or a salivette. The MX3 Hydration Testing System tests a 1 microlitre sample or just 1/1000^th of 1mL of saliva sampled directly from the tongue with those being tested required to swallow all saliva and generate a fresh sample before each measurement, as per the MX3 manual (MX3 Hydration Testing System User Manual, 2022). It’s likely that the minute saliva sample and differences in the generation of the saliva sample account for some of the reported and observed variability.

Other factors that influence saliva hydration testing

Controlling for recent fluid consumption is a key concern for accuracy as determined through research conducted by the U.S. Army Research Institute of Environmental Medicine (Ely et al., 2011). Saliva testing with a laboratory-based desktop osmometer one minute following a brief mouth rinse with water, produced a substantial decrease in test scores (~35mOsm) without any change in actual hydration status. The next test at 15 minutes post mouth rinse produced scores similar to the pre-mouth rinse value (see Figure 3), providing some guidance regarding the time course of fluid consumption impact on test scores. 

Figure 3. Mean saliva osmolality values (error bars represent standard deviation) prior to and at 1- and 15-minutes post brief mouth rinse with water. Hydration classifications according to MX3 Diagnostics

The impact of a mouth rinse or consumption of a small volume of water on saliva hydration testing is an important point. And based upon our observations and site discussions, a point that is understood and used by workers to improve their saliva hydration scores. When informed of the impact of recent fluid consumption or mouth rinsing, a worksite staff member that routinely performs field-based saliva hydration testing posed the question “You’re not going to tell the workers how to game the system, are you?” The answer was “There’s no need to, they are already aware”. Given the frequency of testing, it should be no surprise that workers understand what factors contribute to acceptable test scores, including a quick sip of fluid ahead of their testing.

How reliable are MX3 HTS saliva osmolarity scores?

Reliability is basically a measure of consistency, or in other words, the degree to which you trust the score. Higher reliability translates to greater trust. The sole published study of MX3 HTS reliability assessed SOsm for 75 individuals via two tests conducted within five minutes of each other (Winter et al., 2024). Importantly, tests were conducted indoors with participants providing a fresh saliva sample according to manufacturer instructions. In this regard, tests were more controlled than our observations of those collected in the field, likely resulting in conservative estimates of SOsm reliability when measured on worksites.

Figure 4 plots the mean (average) of the two test scores and the difference between those scores. Despite the controlled testing, the researchers state that “it is with 90% confidence that any repeat measurements that fall within ± 21.3mOsm of an initial measurement may not represent a real change in SOsm". For example, if a workers pre-shift test score was 59mOsm, a score of 81 or greater, or 37 or smaller is required for you to be 90% confident that a true change in test score has occurred. But, the kicker here is that a similar reliability study needs to be done in the field using worksite test collection procedures. With less testing rigour, it’s highly likely that test reliability, and therefore, trust of test scores would worsen from that reported by Winter et al. (2024).

Figure 4. Difference in SOsm test scores 1 and 2 plotted by the average of the two measurements. The thin horizontal lines represent ± 21.3mOsm (Winter et al., 2024)

Does it really matter if test scores are not consistent?

From a scientific viewpoint, yes, it does. A lack of trust in a test result spawns the question, “why measure it?”

Based upon our observations, there appears to be two broad organisational groups performing saliva hydration testing. One group seem to have implemented saliva hydration testing as a means to remind workers to consider their hydration status. Whereas the other group seem to be using saliva testing as a key component of their heat stress management, with specific actions classified against MX3 HTS scores.

Back to the original question “Does it really matter if test scores are not consistent?”, the answer is dependant upon which group an organisation falls into. For those using MX3 HTS scores to determine heat stress and in some cases fitness for work in the heat, a lack of test score reliability has greater implication than for those using it as a means to discuss fluid consumption with workers. Using a measure of hydration status as a surrogate of heat stress or heat strain is not supported by the evidence. We consider that approach an abuse of hydration testing, regardless of the measure. Many factors need to be considered in this regard, including work rate (body heat production) and environmental conditions (determine potential for body heat exchange with the environment), for example (Vega-Arroyo et al., 2019).

What do saliva hydration scores actually mean?

We’ve received some feedback regarding a lack of context for SOsm scores, particularly for organisations that have historically measured urine specific gravity (USG) to estimate worker hydration status. Changes in body mass are considered a useful reference to provide context for other measures of hydration. So, below we provide SOsm and USG referenced against change in body mass from work by the US Army Research Institute of Environmental Medicine (Cheuvront et al., 2010). To understand these graphs (Figures 5 and 6), the term euhydration refers to the hydration state where you’re neither hyperhydrated or dehydrated. We’ve heard euhydration referred to as “normal” hydration, whatever that means. For the results depicted in the graphs below, euhydration was achieved by habitual food intake and consuming 2L of sports drink from 1800 to 2200 on the evening prior to testing. No food or fluid was consumed from 2200 until after the 0630 testing of hydration markers was complete the following morning. Dehydration was achieved through 3-5h of physical activity in hot and dry conditions (40ºC / 20% RH).

Figure 5. Individual and mean values of saliva osmolality, plotted as a function of change in percentage body mass when euhydrated (orange circles) and dehydrated (red circles) (Cheuvront et al., 2010).

Figure 6. Individual and mean values of urine specific gravity, plotted as a function of change in percentage body mass when euhydrated (orange circles) and dehydrated (red circles) (Cheuvront et al., 2010).

What’s all this mean for worker hydration status assessment?

Firstly, given the inherent variability of SOsm and that the MX3 HTS utilises a minute volume of saliva for analysis, rigorous collection procedures are necessary to maximise reliability. This includes at least a 5-min period between fluid/food consumption, smoking/vaping etc and SOsm measurement. Where rigorous sample collection is not practical nor considered worthwhile, the value of saliva hydration status testing must be questioned. 

Where the above is achieved, the results require careful interpretation. Variability in test scores between individuals has been a key theme throughout the saliva osmolality research. Test scores are not suitable to be compared between individuals and large changes (greater than ±21mOsm) are needed before you can trust an actual change has occurred in controlled settings. Research addressing the reliability of MX3 HTS as used on worksites is warranted. As is research on other saliva based measures of hydration, including Salhy (Figure 7) prior to their commercial release to permit organisations to make informed decisions regarding their use.  

Figure 7. Salhy Personal Hydration Test (www.salhy.com.au)

References

Barley OR, Chapman DW, Abbiss CR. Reviewing the current methods of assessing hydration in athletes. J Int Soc Sports Nutr. 2020;17(1):52

Cheuvront SN, Ely BR, Kenefick RW, Sawka MN. Biological variation and diagnostic accuracy of dehydration assessment markers. Am J Clin Nutr. 2010;92(3):565-73

Ely BR, Cheuvront SN, Kenefick RW, Sawka MN. Limitations of salivary osmolality as a marker of hydration status. Med Sci Sports Exerc. 2011;43(6):1080-4

Ely BR, Cheuvront SN, Kenefick RW, Spitz MG, Heavens KR, Walsh NP, Sawka MN. Assessment of extracellular dehydration using saliva osmolality. Eur J Appl Physiol. 2014;114(1):85-92

Friedberg SJ, Doyle EM. Osmotic pressure of saliva. Clin Research. 1959;7:150

Lacey J, Corbett J, Forni L, Hooper L, Hughes F, Minto G, Moss C, Price S, Whyte G, Woodcock T, Mythen M, Montgomery H. A multidisciplinary consensus on dehydration: definitions, diagnostic methods and clinical implications. Ann Med. 2019;51(3-4):232-251

MX3 Hydration Testing System User Manual. 2022. Accessed 2 October 2025. Available at https://mx3diagnostics.com/wp-content/uploads/2024/05/MX3ManualLayout_Jan-2022.pdf

Taylor NA, van den Heuvel AM, Kerry P, McGhee S, Peoples GE, Brown MA, Patterson MJ. Observations on saliva osmolality during progressive dehydration and partial rehydration. Eur J Appl Physiol. 2012;112(9):3227-37

Vega-Arroyo AJ, Mitchell DC, Castro JR, Armitage TL, Tancredi DJ, Bennett DH, Schenker MB. Impacts of weather, work rate, hydration, and clothing in heat-related illness in California farmworkers. Am J Ind Med. 2019;62(12):1038-1046

Villiger M, Stoop R, Vetsch T, Hohenauer E, Pini M, Clarys P, Pereira F, Clijsen R. Evaluation and review of body fluids saliva, sweat and tear compared to biochemical hydration assessment markers within blood and urine. Eur J Clin Nutr. 2018 Jan;72(1):69-76

Winter I, Burdin J, Wilson PB. Reliability and minimal detectable change of the MX3 hydration testing system. PLoS One. 2024;19(11):e0313320