General Characteristics.

Details and key findings extracted from the included studies are reported in Fig 2 and S3 Table. A majority (67%) of the research was conducted in the USA, Australia, and the United Kingdom. Most (71%) were published within the last 10 years and none were older than 20 years. The average sample size of the included studies was 23 out of a total 1,118 participants across all studies with ages in included samples ranging from 18–52 years. A diverse range of participants were represented in the included articles (Fig 2A), with 9 studies involving military participants [17,33,49–55], as well as a variety testing athletes of sports like rugby [56–60], soccer [61–65], running [66–68], rock climbers [69,70], weight-lifting [71,72], and cycling [20,73–75] amongst other sports [76,77]. Many also simply tested healthy adults [1,18,19,78–85]. In addition, 2 studies considered emergency physicians and investigated the exposure of health care personnel to acute stressors and possible implications on their performance and decision making [86,87]. However, of the 49 studies, only 13 included both sexes, with a total of 194 males and 101 females in these studies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Distribution of articles by key characteristics.

Panel A: Distribution of articles by population type studied. Panel B: Distribution of articles according to salivary biomarkers classification. Note: In 7 articles more than one type was examined. Panel C: Distribution of articles by acute stressors. Note: Muscle fatigue induced stressors include any exercise, endurance events, high intensity intervals, weightlifting, and rock climbing. The ‘Multi-stressor’ category includes combinations of different stressors, i.e., sleep deprivation and exercise. Panel D: Distribution of operational readiness outcomes measured. Note: In some studies, more than one outcome type was reported. ‘Others’ category includes outcomes like RPE, self-reported recovery, HRV, EEG, muscle thickness, driving simulation, proteomics, and metabolomics. Eight articles included self-reported measures (i.e., mood) with 3 of these reporting only self-reported measures. Abbreviations: mtDNAcn: Mitochondrial DNA Count Number; RPE: Rate of Perceived Exertion; HRV: Heart Rate Variability; EEG: Electroencephalogram.

https://doi.org/10.1371/journal.pone.0321223.g002

Experimental methodology.

Details on the designs, measured biomarkers, performance measures, and key findings are reported in S3 Table. Twenty-two out of the 49 studies were prospective cohort studies in that they attempted to describe how salivary markers change in response to stressors that are seen in a naturalistic setting but lacked comparative control groups [17,33,49–51,53,54,57–60,64–66,68,70,77,86–90]. Twelve studies were quasi-experimental in that the researchers measured response to imposed stressors pre-post [19,67,69,71,74,76,78,79,82,83,85,91]. Eleven studies were randomized cross-over designs [18,20,56,61,63,72,73,80,81,84,75] while 3 randomized participants into one of multiple stressor groups (parallel design) [52,55,62]. Only 1 study randomized participants into a stressor and true control group [92]. Three of the studies reported a predictive model of a participant’s outcome based on levels of salivary biomarkers [19,58,73]. Of these 3, a model was constructed to predict performance following fatigue [73] was later validated on a different population with a fatigue-inducing stressor (sleep deprivation) [92]. Distribution of key methodological characteristics of identified studies are visualized in Fig 2B-D.

Methodological details and reported outcomes for all identified studies are summarized in S3 Table. Fatigue, followed by sleep deprivation, multi-stressors, dehydration, hypoxia via altitude, and environment (cold stress) were the most studied acute stressors in the studies included (Fig 2C). As stated previously, stressors were determined a priori to be classified into 5 categories: sleep deprivation, muscle damage (fatigue), dehydration, environmental stress, and combined stressors (studies that involved more than one of the prior listed stressors). Additionally, there was a tendency to include specific populations relevant to the stressor. For example, altitude induced hypoxia studies involved trekkers [90] or aviators [55], two populations typically exposed to high altitude. Interestingly, a few studies have included sport game matches as stressors and attempted to predict match outcome (wins/losses) [58,59,88]. Of the 12 studies investigating biomarkers of sleep quality, 4 induced full sleep deprivation [78,92,61,19] while the rest were partial [49,57,68,87,86,93,94]. Of the 26 studies measuring biomarker response to muscle damage and/or fatigue 8 used exercise protocols [20,33,71–73,79–81], 4 involved military training [50–53], while the remaining 14 investigated biomarker response to sports training [58–60,62–66,69,70,74,77,88,89]. Of the 5 studies investigating biomarker response to environmental factors, 4 investigated the effects of varying altitudes during training [18,55,67,90], while 1 investigated the effects of training in the cold compared to normal temperature [75]. Of the 5 studies measuring response to dehydration, 3 measured biomarker response to exercise-induced (active) dehydration only [85,91,75], while 2 measured biomarker response to both active and passive dehydration [82,83]. Eight studies measured biomarker response to combined stressors, 6 of which examined military populations during training [33,52,53,94,51] while 2 involved exercise paired with another stress source like sleep efficiency [76], and sleep and contextual variables (e.g., crowd size, play time, number of matches) [89].

Measured and reported outcomes types are grouped and depicted in Fig 2. Panel D. Seven studies included sports or exercise skills as an outcome. These included hand measures like grip strength [70,80], countermovement jumps [60,63,80], isometric dynamometry [81] and rugby game skills [89,93]. Five studies included game matches, competition outcomes, or sports event completion or success, for example rugby match win or loss, as outcome measures [66,77,58,59,88]. Heart rate was another common performance outcome being measured in 4 studies [18,70,74,20], while 1 study also reported heart rate variability (HRV) [70]. Six studies included self-reported fatigue in their outcomes [17,60,64,70,73,92]. Three studies reported self-perceived muscle soreness [60,64,72] and 1 study has reported self-reported recovery [53].

Four studies considered pulmonary variables as performance measures, including pulmonary function test [66], ventilatory efficiency and oxygen consumption [79], and oxygen saturation [90,18]. Seven studies measured body mass change as an indicator of hydration during dehydration protocols [33,52,85,91,82,83,95]. Eight studies included serum-based biomarkers as measures of stress alongside salivary markers [50,65,69,71,18,82,83,95]. Two of these measured blood lactate levels [69,18], 2 measured muscles damage enzymes [71,65], a further 2 reported blood hormonal markers [50,95], while the final 2 reported serum osmolality [83,82]. Six studies reported sleep quality [57,60,62,64,76,94] and 3 articles reported urine measures (osmolality etc.) [82,83,95] as stress measures. Additionally, 1 study reported muscle thickness [72] while 1 study reported electroencephalogram measures [86].

In addition to physical stress measures listed above, 6 studies measured cognitive tests outcomes [33,76,78,87,92,19] while another 6 studies measured mood states [33,60,63,64,74,92]. Stress levels were measured in 4 studies [53,64,77,94]. Other measures like success on a driving simulation [19] and fatigue established by level of electroencephalogram (EEG) theta waves [86] were recorded in 4 studies. Seven studies did not include a paired performance outcome but instead only reported the change in level of salivary biomarkers in response to respective stressors [49,50,55,57,67,75,51].

Of the 49 studies, 33 used unstimulated methods to extract saliva while 6 used stimulated methods [33,54,67,76,19,86] with the remaining not reporting. Hormonal salivary biomarkers were the most commonly studied (Fig 2 Panel B), with 27 and 19 articles measuring cortisol and testosterone, respectively. The salivary properties of osmolality and flow rate were most commonly measured to study hydration [52,66,85,91,75,20,83,95]. The majority of studies (n=34) attempted to track two analytes at most (of the 34, 15 tracked a single analyte) while the remainder (n =15) tracked multiple analytes in response to stress. Excluding multi-omic studies, the greatest number of analytes studied was by Rutherfurd-Markwick et al. who measured 8 different analytes [20]. Several studies involved a clustering technique [51,73,86] that attempted to identify groups of proteins, metabolites, and peptides out of thousands identified in saliva, and relate the change in the expression of these groups to stress.

The identified studies employed a variety of analytical methods to quantify biomarkers, that can be broadly categorized into ten assay groups. The most frequently used were ELISA-based assays [50,53,55,60,64,66,68–70,72,87,89,75,61,20], being widely utilized in 15 out of 49 studies for their sensitivity and specificity in detecting salivary and plasma biomarkers like cortisol and testosterone, while radioimmunoassay’s were applied for hormone measurements like DHEAS and cortisol in 3 [17,33,79]. Enzyme immunoassays (n = 9) provided versatile applications for hormone and protein measurements [49,56,57,59,63,78,81,88,90]. Techniques involving chromatography and mass spectrometry were used in 5 studies [73,76,86,91,92] as they offer high precision in proteomic and metabolomic analyses. Similarly, general immunoassays (n = 5) [17,58,67,74,80] and osmometry-based assays (n = 5) [52,82–85] were essential for hormone quantification and osmolality measurements, respectively. Colorimetric and kinetic assays were used in 4 studies [19,62,65,77] to analyze enzyme activities, such as salivary alpha-amylase. Additionally, specialized techniques like spectrophotometry [71] and multi-omic approaches [51] highlighted innovative applications in biomarker detection.

The reported sensitivities and coefficients of variation (CVs) from the studies highlight the precision and reliability of the assays utilized. ELISA-based assays demonstrated high sensitivity, with cortisol detection limits as low as 0.007 µg/dL [61] and testosterone assays capable of detecting as little as 3.67 pmol/L [87]. Many assays reported intra-assay CVs below 10% and inter-assay CVs below 15%, indicating consistent performance across repeated measures. Enzyme immunoassays showed similar high sensitivity and low variation, further establishing their utility in precise biomarker quantification.

Muscle Damage and Physiological Fatigue.

Twenty-six studies investigated salivary biomarker response to stressors that can potentially cause muscle damage and fatigue such as military training, weight-lifting, rock climbing, ultra-endurance events, and high intensity interval training [33,50,52,53,60,62–64,66,69–72,74,77,79–81,89,20,51,58,59,88,65,73]. The biomarkers studied in response to these stressors included cortisol [53,60,62–64,69,70,72,74,81,89,20,58,59,88], testosterone [53,60,63,64,70,74,80,81,89,58,59,88], secretory immunoglobulin A (s-IgA) [50,64,66,74,20], alpha-amylase [64,66,77,20], proteins and peptides [52,92,86,51], flow rate [66,20], metabolites [51], osmolality [52,20], creatine kinase and aspartate aminotransferase (AST) [71], potassium and sodium [20], uric acid [65], total antioxidant capacity (TAC) [65], thiobarbituric acid-reactive substances [65], mitochondrial DNA copy number (mtDNAcn) [79].

Cortisol levels decreased in 6 studies [53,63,64,72,74,88], increased in 3 [62,69,70], and remained unchanged in 3 [20,59,81] as a response to induced muscle damage and/or physical fatigue. Other studies quantified cortisol levels as predictors of performance with greater cortisol levels being associated with sporting performance [89] or win probability [58]. Testosterone levels decreased in 3 studies [53,74,58], increased in 4 [63,80,88,59], and remained unchanged in 3 studies [64,70,81]. Like with cortisol levels, several studies investigated testosterone concentration as predictor of performance and/or match outcome finding that both increases in testosterone levels after mid-week training exercises [58,59] as well as lower testosterone on the morning of a match predictive of match success [88]. Several studies examined the testosterone to cortisol ratio (T:C ratio) [61,64,65,71,75,82,20], with 4 reporting an increase [64,65,75,20], 1 reporting a decrease [71], and 2 reporting no change [61,82].

Changes in the expression of proteins and peptides after exercise, physically demanding training, and/or work-related fatigue were explored in 4 studies with one study demonstrating reduced total salivary protein after 5 weeks of military training compared to baseline [52]. Another 2 studies involved the identification of novel fatigue biomarkers from changes in the abundance of peptides/proteins that predicted fatigue status at 83% [51] and 96% [86] accuracy while the last developed a ‘fatigue biomarker index’ from a ratio of ion intensities from 2 novel peptides that was discriminatory of rating of perceived exertion during an extended exercise trial [92]. In the same study that identified novel protein-based markers of fatigue related to military training, a panel of small molecules (metabolites) were also identified yielding a discriminatory value of 96% [51].

Salivary IgA decreased in 2 studies [50,64] and remained unchanged in 1 [20]. In another, its response varied on the status of the athlete, decreasing if participants finished a 36-hour ultra-endurance event and increasing if they dropped out or were stopped before 24 hours [66]. Alpha-amylase increased in 2 studies [77,20], remained unchanged in 1 [66], and decreased only at the 8^th and final mesocycle of a 40-week professional soccer season [64]. Salivary flow rate showed varying responses in both studies investigating its change after exercise, with 1 study showing an increase in females but not males after exercise [20] and the other showing an increasing in non-finishers of a 36-hour endurance event but no change in finishers [66]. Osmolality was increased by approximately 25% at the end of 5 weeks of basic military training [52] but remained unchanged after a single bout of exercise [20]. Rutherfurd-Markwick et al. found that potassium ions (K+) increased for females after exercise but remained unchanged for males, while sodium ions (Na+) did not change for either sex [20]. In response to accentuated eccentric training protocol among resistance trained sports science students, both creatine kinase and aspartate aminotransferase were elevated post-exercise [71]. Uric acid levels remained depressed 38-hours after muscle damage induced by professional soccer games, while total antioxidant capacity remained unchanged and thiobarbituric acid-reactive substances were elevated [65]. Another study examined ventilatory efficiency during an 8-minute cardiopulmonary exercise on a treadmill (cycling 4–6 minutes at 80–90% maximal heart rate) and found that mitochondrial DNA content (mtDNAcn) was inversely correlated with VE/V_CO2 [79].

Sleep Deprivation.

Twelve studies measured the response of salivary biomarkers to sleep deprivation [33,49,57,68,78,87,19,61,86,92–94]. These included cortisol [33,57,68,78,87,61,93], testosterone [33,49,57,68,61,93,94], T:C ratio [57,68,61], alpha-amylase [19], DHEA [94], melatonin [87], and untargeted proteins identified via mass spectrometry [92,86]. Cortisol levels were significantly increased with sleep deprivation in 4 studies [57,68,61,93], while in 2 studies cortisol level changed relative to a factor like work-shift time [87] or time of saliva sampling [33]. For example, overnight shift workers showed a trend toward elevated cortisol in the evening compared to day-shift workers [87], and after a 53-hour military training exercise, morning cortisol levers were lower compared to baseline while evening levels were higher [33]. Another study showed a decrease in cortisol awakening response after a 24-hour sleep deprivation protocol [78].

Testosterone levels were decreased in 3 studies [33,49,94] and remained unchanged in 4 [57,68,61,93]. T:C ratio decreased in 2 studies [57,61] and increased in 1 study [68]. Alphas-amylase was found to decrease in 1 study [19] while DHEA was also decreased in another [94]. A single study examined melatonin and reported no change due to sleep deprivation from shift work [87]. One study identified a pair of peptides that could discriminate the sleep deprived group from control [92]. Another proteomic study was able to identify a group of 30 proteins related to fatigue-status and established a model that discriminated between fatigue (after 18–24 hours work shift) and non-fatigue (before work shift) status at 96% accuracy that is likely largely related to shift work-based sleep deprivation [86].

Dehydration.

Five studies examined dehydration [85,91,75,83,82] with biomarkers investigated including osmolality [85,91,75,83,82], total protein concentration [85,91,75], flow rate [85,75] and alpha -amylase, K+, chloride ions (Cl-), and cortisol [91]. Osmolality [85,91,82,83,95] and total protein concentration was found to increase with dehydration in all studies in which it was tested, while flow rate also consistently decreased [85,95] in response to dehydration. The range increase of osmolality was 38% -110% (pre/post) for active dehydration and 9% for passive dehydration. Total protein increased in the range of 133% - 500%. Meléndez-Gallardo et al. reported that Cl^- level is effective (p<0.05) in identifying dehydration in even mild changes in body mass of 1.5% [91]. The study also found an increase in levels of cortisol, K+, and alpha-amylase due to moderate (2% water loss) dehydration [91].

Environmental Stress.

Five studies examined environmental stress. Among them, 4 investigated the impact of high altitude [67,90,18,55], while 1 examined the effects of cold temperature [75]. Markers studied included cortisol [67,90,55], pH [18], DHEA-s [55], IgA, osmolality, and flow rate [75]. High altitude was found to elevate cortisol [67,90,55], DHEA-s [55], and pH levels [18] of saliva following exercise. Another study demonstrated a relationship between ambient temperature and salivary IgA response but found no influence of cold temperatures on saliva flow rate of IgA secretion [75].

Multi-Stressor.

Eight studies examined multiple stressors with 6 studies examining military populations [33,49,52,53,94,51] and 2 studies involving athletes where exercise was coupled with another stress source including sleep efficiency [76], and sleep and contextual variables (crowd size, play time, number of matches) [89]. Biomarkers evaluated included cortisol (n=4), testosterone (n=3), DHEA-s (n=1), melatonin (n=1), alpha-amylase (n=1), osmolality (n=1), total protein (n=1), and proteomics and/or metabolites (n=2). Cortisol decreased following field-training but was significantly increased during the 3-day recovery period compared to baseline in one study due to stressors that included military training, exercise, and sleep deprivation [53] while it increased in another due to survival training associated [54]. Lieberman et al. reported a decrease in morning levels of cortisol but an increase in evening levels due to intense exercise with load carriage, sleep deprivation, dehydration, and heat environmental stressors [33]. Another study did not provide specific quantitative concentration levels of cortisol but found an association between cortisol levels and performance ratings by coaches and players [90]. Several studies reported that testosterone [33,53] and T:C ratio [53] decreased in response to military training, exercise, sleep deprivation, and heat environment stressors. Taylor et al. discovered that DHEA-s levels increased during both the morning and afternoon, while the DHEA-s to cortisol ratio exhibited an increase in the morning and a decrease in the evening. [54]. The digestive enzyme alpha-amylase was shown to increase shortly before, and then significantly decrease 1 day after a Taekwondo competition relative to baseline [77]. Fogt et al. reported that osmolality increased while total protein concentration decreased in Air Force basic military trainees due to exercise and dehydration stressors [52]. In one study, a proteomic and metabolomic analysis identified clusters of proteins and metabolites capable of distinguishing between various states of combat missions [51]. Additionally, another study identified clusters of metabolites associated with cognitive performance and their ability to differentiate between groups based on their sleep patterns [76]. A single study examined melatonin response to intense exercise with load carriage, sleep deprivation, dehydration, and heat environment stressors and found no change in concentration levels [33].

Muscle Damage and Physiological Fatigue.

The majority of studies examined the response of salivary biomarkers to muscle damage and physiological fatigue, identifying creatine kinase, aspartate aminotransferase, uric acid, cortisol, testosterone, and the T:C ratio as responsive markers. However, the direction of change (e.g., increase or decrease) and magnitude of response of these markers varied widely across studies. For instance, cortisol and testosterone, two of the most frequently measured biomarkers of muscle fatigue and damage, demonstrated mixed responses. Thus, the evidence suggests that hormonal markers like cortisol and testosterone alone may lack the necessary sensitivity and specificity to discriminate TA readiness to perform occupational duties due to their responsiveness to other factors like psychological and social stress [96] and diurnal rhythm [97].

Despite their individual limitations, there was evidence to support their combined utility. One study found that mid-week levels of cortisol and testosterone predicted rugby match success with 72.2% accuracy [58]. Another study found a significant increase (p <0.001) in pre- to post- mid-week T:C ratio before winning games compared to no change before losing games [59], suggesting that the T:C ratio may serve as a more reliable indicator of anabolic/catabolic balance which is reflective of body strain (low ratio) or recovery (high ratio) [98]. Similarly, the ratio of DHEA-s, an anabolic HPA regulator, to cortisol was explored in response to acute hypobaric hypoxia [55]. Despite the rapid increase in cortisol in response to hypoxia, the overall DHEA-s to cortisol ratio remained the same [55], highlighting the potential utility of ratios over single biomarkers for assessing stress resilience.

Other biomarkers frequently studied included IgA [64,66,74,20] and alpha-amylase [64,66,77,91,20], which also exhibited inconsistent directional results across studies. For example, Cook et al. reported decreased IgA levels in ultramarathon finishers, while non-finishers participating for more than 12 hours showed increased levels [56]. Additional salivary biomarkers like creatine kinase [71], thiobarbituric acid-reactive substances (TBARS) [65] and uric acid [71,65] demonstrated potential utility, albeit with varying response timelines. TBARS, a byproduct of oxidative stress, showed significant increases post-professional soccer games before returning to baseline 48-hours after [65]. In comparison, creatine kinase and uric acid exhibited delayed responses, with significant changes noted 48 [65] and 96 hours [71] post-exercise, respectively. Uric acid in particular is known as a key salivary antioxidant and has shown to reliably indicate stress during physical exertion as prior evidence has demonstrated that resistance training significantly elevates salivary uric acid levels that were well correlated with serum markers of oxidate stress [99].

Innovative approaches involving proteomics and metabolomics have identified promising markers for fatigue detection. For instance, a panel of five proteins (e.g., periplakin, heat shock protein alpha family B member 9, myeloperoxidase, heat shock protein family A member 9, and transketolase) discriminated between fatigued and non-fatigued individuals with 83% accuracy following military training, while a panel of 5 metabolites (e.g., C₁₆H₁₃-ClN₂O, gamma-butyrobetaine, L-proline, proline-glycine, and pipecolic acid) achieved 96% accuracy in discriminating fatigue status in the same sample [51]. The identified molecules associated with innate immunity, protein cycling and processing, and metabolism of sugars and amino acids, were identified using a discovery approach in which samples were analyzed using mass spectrometry to quantify all detectable proteins and/or metabolites. From which a subset that best predicted fatigue status was identified using statistical techniques like linear discriminant analysis (LDA) [51] and Fisher discriminant analysis [86]. Other recent evidence supports the use of the proteome to evaluate physical stress as proteins linked to immune response, antioxidant activity, and structural integrity were found to be responsive to both exercise and sex [100]. Although not yet practical for field settings, use of multi-omic analyses and machine learning techniques may allow for the discovery of novel combinations of biomarkers that are more robust to some of the common confounders to predictions based off singular analytes like diurnal rhythm or race (e.g., genetics), which can then be used to develop a novel assay.

The greater responsiveness to fatigue of the markers found in these studies may be due to a variety of reasons including their greater antigenicity [86], reduced susceptibility to non-exercise factors such as dietary and health conditions, and perhaps most importantly, the availability of numerous proteins or metabolites involved in a wide variety of physiological pathways. However, these discovered proteomic markers have not been studied as extensively as testosterone and cortisol. Further research is necessary to characterize their responses to physical stress and establish their reliability for operational readiness assessments.

Sleep Deprivation.

Cortisol levels consistently increased in response to sleep deprivation [57], with a noteworthy decrease in the magnitude of the awakening response (e.g., the rapid cortisol rise 30 minutes after awakening) [78]. These findings align with previous evidence suggesting that sleep deprivation can impair HPA function. This impairment results in elevated evening cortisol levels due to disruptions in glucocorticoid feedback mechanisms [101] and reduced morning, cortisol secretion, reflecting alterations in normal circadian rhythm [102].

In comparison to cortisol, the impact of sleep deprivation on testosterone levels was more variable. While testosterone tended to decrease following sleep deprivation [33,49,94], several studies reported no significant change [56,61]. This inconsistency may be due to methodological differences in factors like population studied, chronic vs acute sleep-deprivation, lack of standardization in sleep loss protocol, and/or different saliva collection time points. Evidence suggests that the degree of sleep-deprivation is a critical factor where total sleep deprivation appears to reduce serum testosterone levels, while partial sleep-deprivation does not [103]. These trends were mirrored in the reviewed studies examining salivary testosterone as those with more intense sleep deprivation (e.g., from military training) [17,33,49] resulted in decrease testosterone while those with less intense sleep loss did not [56,61]. The precursor to testosterone and other steroid hormones, DHEA-s, also decreased following sleep deprivation from military training [17]. Additionally, decreased levels of alpha-amylase were found in one study after extensive sleep-deprivation and was suggested to be sensitive to the diurnal rhythm for arousal and capable of monitoring performance by closely tracking the diurnal drive for alertness [19].

Proteomics-based studies further highlight the potential of advanced biomarker panels in assessing fatigue related to sleep deprivation. Two studies investigating the impact of sleep deprivation on salivary biomarkers used proteomic markers to predict fatigue status with one identifying clusters of peptides that successfully distinguished between fatigued and non-fatigued groups with over 90% accuracy [92] and another identifying a group of 30 proteins and established a model with 96% accuracy rate in discriminating between subjects fatigued from long-duration shift-work compared to non-fatigued subjects [86]. These findings highlight the potential for proteomics to enhance the precision of fatigue detection caused by sleep deprivation. Panels of proteomic biomarkers may provide a more comprehensive and reliable method to assess sleep quality and/or response to poor sleep, particularly in operational and occupational settings.

Dehydration.

Research has consistently demonstrated dehydration’s detrimental effects on various fitness components, including aerobic performance [104–106], strength [107,108], power [109,110], and balance [111,112]. Additionally, reviews have found that a >2% BW loss impairs cognitive functions related to attention, executive function, and motor coordination [113–115]. This body of evidence supports the need for hydration monitoring to maintain operational readiness in the environments where TAs operate. As TAs frequently operate in environments where both physical exertion and heat exposure are high, early detection and management of hydration status are essential for maintaining both performance and operational effectiveness.

Identification of hydration status using salivary biomarkers showed promised based on the results of the studies included. Salivary properties such as increased osmolality and total protein concentration, and decreased flow rate consistently indicated dehydration, particularly at moderate to severe levels [52,20,83,95]. These relationships are well-supported by physiological mechanisms, as dehydration alters salivary gland function through changes in serum osmolality and volume [116,117]. Additionally, dehydration triggers the release of acute phase proteins and an elevation in globulin concentration [118]. Other biomarkers, such as cortisol and Cl-, also responded to dehydration [91]. Notably, Cl- was particularly effective in detecting mild dehydration (~1.5% decrease of body weight), likely from to its stronger connection to water movement due to its ionic nature This sensitivity suggests Cl- could enable early intervention before more severe dehydration-related effects occur [91]. These findings are especially relevant as hydration is a critical concern for TAs [119], with physical performance declining at dehydration levels of >2% body weight, particularly under heat stress [120].

Environmental Stress.

The primary environmental stressors addressed in the studies included were temperature and altitude. Altitude-induced stress was associated with elevated levels of cortisol [67,90,55], DHEA-s [55], and salivary pH [18]. Cortisol and DHEA-s levels increase with elevated altitude as a physiological adaptation to the reduced oxygen availability at higher altitudes, prompting adaptation of the HPA axis [121,122]. Salivary pH may also rise due to changes in buffering capacity influenced by hypoxia [123]. While altitude exposure may be less relevant for monitoring the readiness of TAs compared to more common acute stressors like sleep deprivation, dehydration, and muscle fatigue/damage, the physiological effects of high altitude are well documented [124] and can still pose challenges to operational readiness in specific contexts. In addition, the effect of cold temperature was investigated in a single study whereby exercise in cold temperatures relative to normal temperatures resulted in decreased osmolality but not salivary IgA [75]. Although limited, this finding suggests that cold environments may influence certain salivary biomarkers differently than other stressors, warranting further exploration to understand the implications for TA readiness in extreme conditions.

Multi-Stressor.

Several studies incorporated multiple stressors, predominantly within military populations [33,52–54,51], enhancing ecological validity by more accurately reflecting the multifaceted nature of military duties better than studies with single stressors. These findings offer a more comprehensive understanding of the physiological responses to complex, real-world stressors. Notably, inconsistent cortisol responses observed in these studies [33,53,54], highlight the variability of stress reactions in different individuals and situations. The decreases in testosterone levels and the T:C ratio following military training suggest a potential overtraining effects [125–127], which is critical for developing effective training and recovery protocols. Conversely, the increase in DHEA-s levels increased during stressful captivity training, along with the variations in DHEA-s to cortisol ratio between morning and evening, indicates an adaptive response to prolonged stress [54].

Signs of dehydration during military training, evidenced by increased osmolality and decreased protein concentration [52], emphasize the need for proper hydration strategies to maintain TA operational readiness. Lastly, while most studies focused on standard hormonal panels and salivary properties, one study’s use of proteomic and metabolomic profiling to identify proteins linked to chronic stress and distinguish acute battle-related stress events underscores the potential for advanced biomarker analysis in understanding stress responses [51].

The collective findings from the studies reviewed have identified several salivary biomarkers, as depicted in Fig 3, which show promise for detecting exposure to specific acute stressors frequently encountered by TA. These salivary biomarkers appear to mirror either a systemic reaction to acute stressors or responses specific to certain conditions. Cortisol, a key marker of HPA axis activation, indicates physiological stress across various types of stressors, and affects nearly every tissue due to the widespread presence of cortisol receptors [10]. Cortisol also influences the levels of other significant biomolecules, such as alpha amylase [128], demonstrating its broad physiological impact. However, while cortisol serves as a general indicator of stress, it may lack the specificity needed to differentiate between individual responses or distinct stressors. Additionally, testosterone levels and the T:C ratio offer additional insights to the anabolic-to-catabolic balance, which is crucial for muscle recovery and adaptation [129]. While cortisol and testosterone reflect systemic responses and deliver valuable information, proteomic analyses provide further insight into cellular processes, revealing more distinct biological markers [130]. This finer resolution aids in understanding the unique signatures associated with different physiological stressors and may guide the development of targeted strategies to monitor and mitigate the effects of acute and chronic stressors in TAs.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Commonly reported stressors of Tactical Athletes (TA) and responding salivary biomarkers identified in this review.

^⤉Indicates that the salivary biomarker was included in predictive models of outcomes relevant to TA success.

https://doi.org/10.1371/journal.pone.0321223.g003

Cognitive Performance.

Six of the identified studies reported cognitive outcome measures [19,76,78,87,92,131]. Notably, Michael et al. [92] demonstrated that a panel of specific salivary peptides could differentiate between sleep-deprived individuals and controls with an area under the curve (AUC) of 92%, while cognitive performance on tasks such as the Stroop color test, could discriminate between sleep deprived individuals and controls with an AUC of 88%. While these peptides hold potential as a proxy measure of cognitive performance before and after sleep-deprivation, a direct relationship between these biomarkers and cognitive performance was not established [92].

Further insights came from Machi et al. [87], who observed a 28% reduction in the morning cortisol peak following a night shift as well as declines in cognitive performance on the Stroop color word test and the Southern California Repeatable Episodic Memory. Similarly, cortisol levels post-sleep deprivation was related to scores on the Acute Readiness Monitoring Scale (ARMS), particularly its cognitive subscales, which were associated with the psychomotor vigilance test results [78]. In another study, Lieberman et al. [131] found that elevated cortisol levels prior to field military training were inversely related to cognitive reaction times, indicating that higher cortisol may be indicative of cognitive strain under stressful conditions.

Other biomarkers also showed promise in predicting cognitive outcomes. Pajcin et al. [19] identified a positive association between alpha-amylase levels and performance on both the psychomotor vigilance test and driving simulation tasks following sleep deprivation, suggesting alpha-amylase is responsive to cognitive and motor skills under stress. Lastly, Akazawa et al. [76] discovered that several metabolites, including citrulline, were elevated in individuals with poor sleep quality. Citrulline showed a strong inverse correlation with cognitive performance during high-intensity exercises, suggesting its potential as a biomarker for cognitive decline in physically demanding situations. Collectively, the findings suggest that specific salivary biomarkers, including cortisol, alpha-amylase, and citrulline, are responsive to changes in cognitive performance, particularly under conditions of sleep deprivation and high-intensity activity. However, further research is needed to establish direct and consistent relationships between these biomarkers and cognitive performance metrics.

Physical Performance.

There were more studies that examined physical performance outcomes than cognitive (S3 Table). Testosterone levels increased significantly (91%) in response to a maximal aerobic velocity test [68], likely due to the short-term anabolic effects of high-intensity activity [132]. Similarly, testosterone increased by 20% in men and 15% in women following sprint cycling [80]. Several studies found the T:C ratio declined in response to a maximal aerobic test [61] and after two weeks at a sports climbing camp [70] but was primarily driven by increases in cortisol. Conversely, Mclean et al. [60] found that while post-rugby match countermovement jump performance declined, cortisol and testosterone remained unchanged, suggesting an adaptation to stress in highly trained individuals [133,134]. However, this finding contrasts with results from a 21-day cycling training regimen where daily T:C ratio fluctuated dramatically (0–400%), yet performance remained consistent [74]. Moreover, a number of other studies reported changes in cortisol and testosterone in relation to physical performance outcomes with mixed findings [63,89,58,59]. Such variability underscores the complex and context-dependent relationship between hormonal biomarkers and physical performance.

Non-hormonal biomarkers also showed associations with physical performance in several studies examining the relationship between changes in non-hormonal markers with physical performance [66,79]. Bellar et al. reported that that IgA levels dropped significantly amongst ultramarathon finishers reflecting a decline in mucosal immunity [66]. Additionally, Chen et al. [79] found exercise ventilatory efficiency was associated with mtDNAcn, and may be indicative of aerobic fitness.

Cumulatively, the findings emphasize the complex relationship between hormonal responses to stress and their relationship to physical performance. While hormonal markers such as testosterone, cortisol, and the T:C ratio provide insight into anabolic and catabolic balance, their responsiveness to physical performance outcomes remains inconsistent. Non-hormonal biomarkers like IgA and mtDNAcn may offer additional sensitivity and specificity for monitoring changes in physical readiness. Therefore, future research should explore non-hormonal biomarkers that show evidence of being able to respond to changes in physical performance more sensitively and specifically.