Research progress on exercise fatigue from the perspective of fatigue biomarkers

Adenosine triphosphate and creatine phosphate

The initial energy demands during exercise are primarily met through the rapid depletion of high-energy phosphagens, notably adenosine triphosphate (ATP) and creatine phosphate (CP), collectively termed the phosphagen system (Fig. 2). As physical activity progresses, the decline in these intramuscular energy reserves correlates directly with diminished contractile capacity.

High-energy phosphate compounds central to exercise metabolism include ATP and phosphocreatine (creatine phosphate, PCr). During physical activity, ATP breaks down to release energy, producing adenosine diphosphate (ADP). Phosphocreatine then donates its phosphate group to regenerate ATP from ADP. Concurrently, two ADP molecules can combine via the enzyme myokinase to form one ATP and adenosine monophosphate (AMP), releasing inorganic phosphate ions (Pi) (Hackney, 2016). Generally, energy consumption in a short period of time is mainly based on CP, and the decline can reach more than 90%. It has been well documented that a 70 kg person consumes about four calories (kcal) per minute when taking on a 1-hour walk. As exercise time increases, the calories consumed gradually increase (Thompson et al., 2013). Therefore, ATP and CP and their related metabolites, AMP, ADP, and Pi, can be used as one of the biomarkers for the preliminary assessment of exercise-induced fatigue.

Glucose and glycogen

During prolonged strenuous exercise, the main energy substance consumed is saccharides. After long exercise, glucose in the blood is consumed. With further extension of exercise duration, muscle glycogen and liver glycogen are greatly consumed. At this stage, up to 75–90% of muscle glycogen and more than 90% of liver glycogen stores may be depleted. Glycogen cannot maintain normal blood glucose levels and often resulting in hypoglycemia (Hackney, 2016; Nelson & Cox, 2005).

Fat

As exercise progresses, fat will also begin to be consumed; however exercise-induced fatigue will not lead to a large reduction of body fat (Romijn et al., 1993; Van Loon et al., 2001). Although the overall fat content does not change much, the levels of fatty acids and triglycerides in the blood will increase (Liao et al., 2024). Jie (1998) demonstrated that plasma free fatty acid concentrations in the blood during surface body exercise could rise from 0.1 mmol/L to 2 mmol/L. It is generally believed that fatigue onset can be delayed to some extent if endurance training can improve the utilization of fat during exercise, thereby reducing the consumption of glycogen and preventing a decline in blood glucose levels.

Amino acid related to energy metabolism

During exercise, free amino acids and intracellular amino acids in the blood are consumed and utilized. Among these, glutamine, leucine, isoleucine and valine are the amino acids currently considered to be related to energy metabolism. These amino acids in blood are consumed as exercise progresses and can also serve as biomarkers of exercise-induced fatigue (McGlory & Phillips, 2016).

Long exercise consumes glutamine, reducing its levels in the blood and muscle. The enzyme activity of glutamine is reduced due to the decrease in glycogen and blood glucose (Mioko, Yuji & Kazuto, 2013).

Valine, leucine and isoleucine are all branched-chain amino acids. These amino acids can be catabolized in muscle tissue and can be used by oxidative energy supply. After marathon runners supplemented branched chain amino acids, weariness decreased significantly. The results of comprehensive literature show that the branched chain amino acids are beneficial during short-term, moderate-intensity exercise; after prolonged exercise (>3 h) or when exercise-induced fatigue occurs, their depletion shows a high correlation with the state of fatigue.

While the depletion of energy substrates like glycogen and CP provides a logical and historically significant explanation for fatigue, their utility as real-time, predictive biomarkers is limited by several factors. Firstly, the direct measurement of intramuscular glycogen is highly invasive and impractical in most athletic settings. Secondly, there is considerable inter-individual variability in substrate utilization rates and fatigue thresholds, meaning a ‘one-size-fits-all’ critical level of depletion does not exist (Ament & Verkerke, 2009). For instance, well-trained athletes exhibit enhanced glycogen sparing and fat oxidation capabilities, delaying the point at which glycogen depletion triggers fatigue (Zafrilla, Marhuenda & Arcusa, 2019). Therefore, while these measures are mechanistically crucial, future research should focus on developing non-invasive proxies (e.g., via breath or sweat analysis) or dynamic metabolic models that can predict individual substrate depletion kinetics, rather than relying on static, post-hoc measurements.

Lactic acid

The human body consumes a lot of ATP and CP during intense exercise. When these substrates become insufficient, the body begins to use the lactic acid system (anaerobic glycolysis) for energy supply in a short duration. During this process, ATP is produced by glucose under anaerobic conditions. Due to the low efficiency of ATP production via anaerobic glycolysis, the body will conduct a large amount of anaerobic glycolysis in order to produce enough energy, producing large amounts of lactic acid. Notably, this rise in blood lactate concentration is primarily driven by muscle contraction demands, with a significant contribution from the recruitment of fast-twitch muscle fibers (Sánchez-Medina & González-Badillo, 2011). These fibers are characterized by high force output but low fatigue resistance, and their heightened expression of glycolytic enzymes (e.g., phosphofructokinase, lactate dehydrogenase) accelerates glycogen breakdown. This causes pyruvate production rates to exceed the oxidative capacity of the mitochondria, leading to disproportionately high lactate generation compared to slow-twitch fibers (Garcia-Sillero et al., 2022). Yan et al. (2022) found through bicycle experiments in the 80s that exercise-induced fatigue was associated with elevated lactic acid after exercise.

Lactic acid itself does not cause fatigue; rather fatigue is associated with the H+ dissociated from lactic acid. The decrease in pH affects many processes, including the ability of troponin to bind calcium, as well as the activity of many enzymes. Previous studies have confirmed that the pH value decreases, which reduces the activity of kinases such as creatine kinase, ATPase, and phosphofructokinase (PFK), thereby affecting the metabolism of the lactate system (McCown, Reibel & Micozzi, 2010). This mechanism of lactate production, dominated by fast-twitch fiber recruitment, explains the rapid surge in blood lactate observed during high-intensity exercise (Garcia-Sillero et al., 2022). Consequently, lactate and the associated pH decrease remain established biomarkers for assessing exercise-induced fatigue and exhaustive exertion.

Ammonia

Studies have confirmed that during long-term high-intensity exercise, proteins and amino acids in the human body are consumed to participate in the energy supply. Due to the decomposition of proteins during long-term exercise, amino acid decomposition produces ammonia (Khadartsev, Fudin & Badtieva, 2022). Studies have shown that the increase in ammonia concentration plays an important role in both central and peripheral fatigue (Meeusen et al., 2006). In general, high concentrations of ammonia can affect ATP synthesis. At the same time, the increase in ammonia concentration will inevitably lead to an increase in osmotic pressure, resulting in internal environmental disorders. In addition, ammonia can enter the brain tissue through the blood–brain barrier, which has a toxic effect on the brain and affects the function of the central nervous syste. Current research suggests that ammonia hinders the synthesis of inhibitory GABA (γ-aminobutyric acid). Due to GABA deficiency, nerve control is reduced, resulting in fatigue (Foley et al., 2006).

Urea

Urea is closely associated with the metabolism of amino acids within the human body. Consequently, it enables us to assess the physical functioning and fatigue levels of athletes in a more comprehensive manner. It is widely acknowledged that the quantity of blood urea tends to rise in proportion to the exercise load, and its recovery process is relatively sluggish. The extent of exercise-induced fatigue is determined by measuring the degree of increase after exercise and the subsequent rate of recovery. When the level of urea in the blood following exercise is three mmol/L higher than that before exercise, it can be construed as an indication of a substantial amount of exercise, signifying that the athlete has reached the fatigue threshold (Qian et al., 2024).

Currently, urine and saliva are widely used as non-invasive biomarkers for fatigue monitoring in sports. To improve their measurement accuracy, it is essential to overcome the challenges in standardizing sample processing and advancing detection technologies. The key steps include: first, sample collection and storage processes must be strictly standardized. For urine, analysis must be conducted within 2 h (or within 3 h if refrigerated) to prevent biomarker degradation. For saliva, the addition of RNA stabilizers (such as RNAlater) enables preservation at room temperature for up to one year, significantly reducing the risk of biomarker degradation caused by repeated freeze-thaw cycles (Zheng et al., 2025).

Secondly, high-sensitivity detection technologies were applied. Automatic online solid-phase microextraction coupled with liquid chromatography-tandem mass spectrometry (SPME-LC/MS/MS) achieved a precision of 4.9% for cortisol detection in 40 µL of saliva (quantification limit: 0.03 ng/mL), while magnetic bead-assisted peptide mass analysis (MALDI-TOF MS) can identify fatigue-specific differential peptides within the molecular weight range of 2,000–15,000 Da, with a cross-validation rate of 95.49%, providing new targets for the development of portable devices (Gervasoni et al., 2018). Further elimination of systematic errors through multimodal data integration: combining salivary cortisol/α-amylase with urine urea/uric acid ratios and integrating psychological scales (such as POMS) to establish a machine learning assessment model, reducing the overall error rate by 32%. Additionally, constructing an individualized baseline database (such as resting salivary cortisol ranges) helps to avoid misjudgments based on group standards, ultimately achieving precise quantification of fatigue states (Sequeira-Antunes & Ferreira, 2023).

The traditional view of lactate as a mere fatigue-causing waste product has been profoundly revised. The ‘lactate shuttle’ hypothesis re-frames lactate as a crucial energy carrier and signaling molecule (Brooks, 2018). This paradigm shift challenges the oversimplified acidosis model, as the relationship between pH decline and fatigue is not always direct or causal (Westerblad, Allen & Lännergren, 2002). Similarly, while ammonia and urea levels correlate with protein catabolism and fatigue, their specific mechanistic roles remain inadequately defined. A critical research gap lies in understanding the dynamic interplay between these metabolites. Future studies should move beyond correlative measurements and employ interventions that selectively manipulate individual metabolites to establish causality in fatigue development.

Creatine kinase

Serum creatine kinase (CK) is a crucial enzyme in the biochemical processes that has the remarkable ability to catalyze the formation of adenosine triphosphate (ATP). primarily through the CK/phosphocreatine (PCr) system. This system minimizes [ATP]/[ADP] fluctuations during high-intensity activities to sustain contractile function (Dahlstedt et al., 2003). It serves as a reaction-catalytic enzyme for the recovery of ATP, a process that holds significant importance in relation to the maintenance of the energy balance following physical exercise. The presence of serum creatine kinase is primarily attributed to the movement of creatine kinase from within the muscle cells into the serum across the cell membranes. Typically, the content of serum creatine kinase is ordinarily maintained at a relatively low level. However, when the body undergoes exercise-induced fatigue, it leads to an increase in the permeability of the cell membranes. As a result, creatine kinase is liberated from the cells and enters the bloodstream, thereby causing a notable elevation of the serum creatine kinase content (Qian et al., 2024; Shijing et al., 2016).

Additionally, creatine kinase exhibits antioxidant properties by inhibiting lipid peroxidation and protein oxidation during intense exercise. This protective function may mitigate oxidative stress-induced fatigue, suggesting a dual role in both energy metabolism and cellular protection (Miglioranza Scavuzzi & Holoshitz, 2022). In acute fatigue scenarios, changes in neuromuscular parameters (e.g., reduced peak power, prolonged contraction time) correlate more strongly with performance decline than creatine kinase elevation, indicating CK’s limited sensitivity as a real-time fatigue biomarker during short-term exhaustive exercise (Yánez, 2023).

Hemoglobin

Hemoglobin, the main component of red blood cells, plays a crucial role in delivering oxygen and carbon dioxide. In cases where the level of hemoglobin is reduced or the demand for oxygen increases, the oxygen supply will fall short. This shortage will subsequently lead to a decrease in exercise capacity (Yin et al., 2024). During intense physical activity, exercise-induced fatigue can occur, which may cause damage to red blood cells. As a result, hemoglobin will be released from these damaged cells. This release can lead to a situation where the level of hemoglobin drops below the normal range (Shijing et al., 2016).

Nitric oxide and nitric oxide synthase

Nitric oxide (NO) in the body is enzymatically synthesized by nitric oxide synthase (NOS). Nitric oxide plays a crucial role in promoting the augmentation of blood flow and the dilation of blood vessels, thereby regulating the blood supply within the body (Draghici et al., 2024). In the realm of sports, the supplementation of L-Arginine proves to be beneficial. It helps in reducing the muscle injury that athletes may encounter and also contributes to the enhancement of their performance (Lomonosova et al., 2014). This is particularly significant as it directly impacts the athletes’ ability to perform at their best and minimize the risk of injury that could potentially hamper their progress and career (Jackman et al., 2010). Therefore, the decline in nitric oxide can cause exercise-induced fatigue.

Succinate dehydrogenase

Succinate dehydrogenase (SDH) is a key enzyme in the tricarboxylic acid cycle. Its activity can be used to assess the aerobic oxidation capacity of athletes (Lewis et al., 2010). Succinate dehydrogenase is located on the inner mitochondrial membrane and does not enter the tissue fluid and thus into the bloodstream. Due to exercise-induced muscle tissue damage, the increased permeability of the mitochondria leads to an increase in the content of succinate dehydrogenase in the coating slurry, which can be used to reflect the status of the tricarboxylic acid cycle (Rodrigues et al., 2010).

Testosterone/cortisol

Testosterone is an androgen hormone that accelerates the synthesis of substances in the body, while cortisol is a glucocorticoid that promotes the catabolism of substances in the body. The testosterone/cortisol (T/C) ratio serves as an indicator of the anabolic and catabolic balance of nutrients in the body. Unlike free testosterone concentrations that significantly decrease when the body is depleted due to movement, increased cortisol and its receptors cause protein breakdown beyond synthesis levels (Meeusen, 2014; Shimomura et al., 2009).

Blood-based biomarkers such as CK and hemoglobin are widely used yet frequently misinterpreted. A major limitation is their lack of specificity for fatigue. For instance, elevated CK levels serve as a robust indicator of muscle damage but do not directly reflect the acute state of fatigue that impairs performance within a single exercise bout (Brancaccio, Maffulli & Limongelli, 2007). Similarly, a decrease in hemoglobin may result from hemolysis or changes in hydration status, rather than solely indicating a diminished oxygen-carrying capacity. The T/C ratio, while valuable for assessing long-term anabolic-catabolic balance, is affected by diurnal rhythms, nutrition, and psychological stress, thereby complicating its interpretation in the context of acute fatigue (Küüsmaa, Sedliak & Häkkinen, 2015; Vaamonde et al., 2022). Thus, these biomarkers are most informative when used as part of an integrated panel rather than as standalone diagnostic tools for fatigue. Future research should focus on identifying novel, more specific blood-borne factors.

Malondialdehyde

Malondialdehyde (MDA) is a product of degradation by peroxidation in vivo. To some extent, its concentrarion can reflect the severity of free radical attack and damage to motor cells. The content of malondialdehyde in lipid peroxidation products increased after exhaustion exercise, which proved that malondialdehyde could be used to determine exhaustion exercise. Mitchell’s post-run plasma analysis of ultra-long marathon runners confirmed that malondialdehyde levels in body were significantly elevated after exercise (Maxwell et al., 2001; Mohammadi et al., 2025).

Superoxide dismutase

Superoxide dismutase (SOD) is an important antioxidant enzyme in the free radical scavenging system. Its activity level can indicate the body’s free radical load. When the body has a high free radical content during exercise-induced fatigue, the high activity of superoxide dismutase is required. After prolonged exercise, the results showed a significant increase in malondialdehyde content in the plasma, as well as an increase in the activity of superoxide dismutase. By analyzing SOD/MDA, it can reflect the free radical production and clear rate in the body, and further analyze the actual changes of free radical metabolism in depth and objectively, thereby indicating the degree of exercise-induced fatigue of the body (Zhao et al., 2024).

Catalase

Catalase (CAT) is one of the key enzymes for scavenging intracellular H₂O₂. H₂O₂ is the reduction product of O₂ andhas strong oxidation properties. It can directly oxidize the hydrophobic groups of some enzymes, which can make the enzyme inactive. Catalase can bind to and clear the hydrogen peroxide in vivo. Quintanilha found that catalase activity increased in skeletal and cardiac muscle of rat after aerobic endurance training, indicating that catalase activity in muscle can be improved as exercise progressed (Lew & Quintanilha, 1991). The activity of catalase in the human body is highly sensitive to exercise stimulation. Aerobic exercise can significantly promote the increase of catalase activity in the body, and if the exercise intensity increases, the catalase activity will be further increased (Ekström et al., 2025).

Glutathione peroxidase

Glutathione peroxidase (GSH-PX) is an enzyme that catalyzes the reduction reaction of H2O2, thus protecting membrane structural integrity. Most studies suggest that exercise-induced fatigue lesds to elevated glutathione peroxidase activity. Heavy exercise can cause a significant increase in glutathione peroxidase activity in muscle tissue. Lew et al. reported that when rats ran to exhaustion, glutathione peroxidase, glutathione S-transferase, and glutathione reductase activities increased in the liver and bone, while their activities decreased in plasma (Lew, Pyke & Quintanilha, 1985; Wang et al., 2024; Wu, Liu & Chen, 1999).

The role of oxidative stress in fatigue is a field of significant debate. The ‘mitohormesis’ theory posits that moderate ROS production is essential for adaptive signaling and training responses (Ristow & Zarse, 2010). Therefore, simply observing an increase in MDA or antioxidant enzyme activity does not necessarily indicate deleterious fatigue; it could signify a positive adaptive process. A majority of studies measure these markers in blood, but their levels may poorly reflect the redox environment within contracting muscle fibers (Powers, Radak & Ji, 2016). Furthermore, the inconsistent outcomes of antioxidant supplementation studies in combating fatigue challenge the simplistic ‘oxidants are bad’ narrative (Merry & Ristow, 2016). Future research should focus on the targeted measurement of redox status in specific cellular compartments and during recovery to determine whether oxidative stress is a cause, a correlate, or a consequence of exercise-induced fatigue.

Hydroxytryptamine

Hydroxytryptamine (5-HT) is a metabolite of tryptophan and a crucial neurotransmitter in the central nervous system, involved in various physiological roles. Tryptophan is a substrate for hydroxytryptamine synthesis and a rate-limiting substance, and free tryptophan in plasma can enter the brain through the blood–brain barrier and affect hydroxytryptamine. During exercise, lipolysis increases free fatty acids, and free tryptophan increases, which in turn increases hydroxytryptamine synthesis in the brain. Studies have shown that exercise can increase hydroxytryptamine levels in the central system, and this increase is associated with the development of central fatigue. Newsholme, Gordon & Newsholme (1987) first proposed that hydroxytryptamine may be a regulator of central fatigue. Hydroxytryptamine acts as an inhibitory transmitter that reduces the impulse to be released from the center to the periphery and thus reduces exercise capacity. Studies have also confirmed that with the extension of exercise time, the anabolism of hydroxytryptamine, dopamine, etc. in the brain of the body will decrease (Newsholme & Blomstrand, 1995; Castrogiovanni & Imbesi, 2012).

Dopamine (DA)

Dopamine can regulate the tension degree of the muscle tissue, and dopamine was the first neurotransmitter confirmed to play an important role in exercise-induced fatigue (Liao et al., 2024). Usually, dopamine metabolism increases throughout the brain after exercise. Sutoo Allen, Lamb & Westerblad (2008) has found that there are two main reasons for the increase in dopamine in the brain after initial exercise, one is to promote the synthesis of dopamine, and the other is to promote the binding of dopamine receptors. However, studies have shown that the synthesis of dopamine in the rat midbrain is weakened during central fatigue, and that a high dopamine level can delay fatigue development (Kanter et al., 1985; Lu et al., 2024).These studies have shown that when exercise-induced fatigue, the amount of dopamine in the brain decreases.

Noradrenaline

Noradrenaline (NE) is a neurotransmitter synthesized and secreted by adrenergic nerve terminals. It is produced from dopamine through catalysis by the enzyme dopamine β-hydroxylase. Studies have confirmed that noradrenaline in the hypothalamus decreases after exercise and exhaustion, and the content of noradrenaline can affect the metabolic level of noradrenalin, both of which can inhibit the normal effect of the hypothalamus (Hackney, 2016; Lew, Pyke & Quintanilha, 1985), which is one of the causes of exercise-induced fatigue.

Acetylcholine

Acetylcholine (Ach) is a cholinergic neurotransmitter released within the central nervous system by cholinergic nerve endings. Its synthesis and release are vital for central nervous system. The synthesis rate of acetylcholine is affected by the precursor choline. After running a marathon, the level of choline in the plasma drops by 40%, and supplementing with choline drinks to maintain plasma choline levels will delay the onset of fatigue. Supplementing with choline drinks during marathons can delay the onset of fatigue (Fecik et al., 2024).

Amino acid (neurotransmitter correlation)

At present, it is believed that the amino acids related to central fatigue mainly include gamma-aminobutyric acid (GABA), glutamic acid (Glu), and branched-chain amino acids (BCAA). BCAAs, which include isoleucine, leucine, and valine, are important for energy supply. As discussed earlier, the role of BCAAs in fatigue has been introduced (Roelands et al., 2009).

Gamma-aminobutyric acid is an inhibitory neurotransmitter. One cause of central motor fatigue is an increase in gamma-aminobutyric acid levels. With the extension of exercise time, the body will appear hypoxia, making the gamma-aminobutyric acid oxidation process weakened, and the high concentration of gamma-aminobutyric acid will cause postsynaptic inhibition (Sutoo & Akiyama, 2003). Elevated levels of gamma-aminobutyric acid in the brain can lead to exercise-induced fatigue.

Glutamic acid is a neurotransmitter related to excitability in the central nervous system, which is very abundant in the brain, and normal levels of glutamic acid play an important role in maintaining neuronal excitability, and glutamic acid is the transmitter of most excitatory synapses. When the amount of glutamic acid in the brain changes abnormally, it leads to a decline in the function of the central system, which is onecause of exercise-induced fatigue (Li et al., 2020).

Tissue endothelin

NO can promote vasodilation and is also an important neurotransmitter, which has an important physiological role in exercise, especially in the cardiovascular system. NO is mainly manifested in tissue cells as intracellular messenger molecules, which can cause vascular smooth muscle relaxation through the cGMP interaction. Studies have confirmed that normally NOS is functionally active in brain tissue, and the expression of NOS can be significantly weakened after heavy-load exercise training, indicating that fatigue can reduce NOS expression in brain tissue (Anish, 2005).

ET is an active small molecule polypeptide that promotes vasoconstriction, and its constrictive effect on blood vessels is the most effective substance, contrary to the effect of NO. Previous studies have confirmed that exercise promotes the enhancement of tissue endothelin (ET) expression, causing vasoconstriction, which in turn leads to hypoxia in the body. The expression of tissue endothelin is highly correlated with exercise intensity, and exercise-induced fatigue occurs due to ischemia and hypoxia when the exercise load is too large (Meeusen et al., 2007).

A key limitation in this field remains the overreliance on peripheral measures to infer central neurotransmitter changes. Since the blood–brain barrier restricts free exchange of molecules between the periphery and the brain, such inferences are highly speculative (Meeusen et al., 2007). The classic “central fatigue hypothesis,” once centered on hydroxytryptamine, has been challenged: dopamine and noradrenaline are now considered equally important in regulating motivation and perceived exertion (Roelands & Meeusen, 2010). Methodological constraints also present a challenge. While animal studies permit direct brain measurement via techniques like microdialysis, human studies rely on indirect proxies. Notably, interactions between peripheral metabolites and central neurotransmission form a promising yet underexplored research frontier.

Saliva PH

Due to the increase in lactic acid production after long-term strenuous exercise, the amount of CO2 in the blood increases, and acidic substances such as ketone bodies and pyruvate accumulate, resulting in a decrease in the pH value of blood and thus a decrease in the pH value of saliva (Grzesiak-Gasek & Kaczmarek, 2022). When exercise-induced fatigue, acidosis often occurs, acidosis can reduce the body’s muscle exercise capacity and also cause symptoms of central nervous system fatigue. Adequate alkaline substances should be supplemented after exercise.

Salivary immunoglobulin A

Salivary immunoglobulin A (SIgA) level is one of the important indicators of human immune status. After intensive exercise or long and intense exercise, the immunosuppression caused by exercise leads to a decrease in immunoglobulin A level. Most research indicates that s-IgA level can serve as an indicator for evaluating exercise load, as it decreases following short periods of intense exhaustive exercise. Fahlman et al. (2001) performed 30 s full anaerobic work test for 3 min, which indicated that a temporary decrease in saliva immunoglobulin A levels in women. In summary, saliva immunoglobulin A is a valuable biomarker for assessing exercise-induced fatigue.

Other components in the saliva were used as biomarkers

With the development of high-throughput technology and their application in athletics, more components are identified in saliva, and they can replace the corresponding detection in the blood. These components are also markers of exercise activity. Guo et al. (2010) used the serum and saliva of athletes before and after exercise as the research object, and found the sports-related biochemical indicators in saliva. Giles et al. (2012) identified a class of small molecule proteins (sMW) in saliva after exercise through high-throughput proteome combined chromatography and mass spectrometry, and then correlated the fatigue degree of each peptide exercise, and obtained a positive correlation between a class of small molecule proteins and exercise-induced fatigue.

Enthusiasm for non-invasive biomarkers must be tempered by a critical awareness of their pre-analytical and analytical vulnerabilities. Hydration status significantly influences urinary analyte concentration, necessitating creatinine correction and rigorous sampling protocols (Cone et al., 2009). Similarly, salivary measurements of hormones such as cortisol and IgA are exquisitely sensitive to circadian rhythms, sampling techniques (stimulated vs. unstimulated), and oral health status (Grzesiak-Gasek & Kaczmarek, 2022). A key limitation is that many studies report changes in these biomarkers without first establishing validated, population-specific reference ranges under exercise conditions. Consequently, although non-invasive biomarkers are well-suited for longitudinal monitoring, their current utility for cross-sectional or single-time-point diagnosis of fatigue remains limited. The future direction lies in developing integrated sensor systems for real-time biomarker measurement, combined with machine learning methods to interpret the complex, multi-parameter data they generate.