Biomarkers of Creatine Metabolism in Humans: From Plasma to Saliva and Beyond

Abstract

The literature on creatine biomarkers in various bodily fluids remains limited. The purpose of this review is to explore the available data regarding the presence of molecules considered biomarkers of creatine metabolism—namely creatine, guanidinoacetate, and creatinine—across different bodily fluids and matrices. In addition to providing reference values for each biofluid, the paper reports concentrations of these biomarkers in different pathologies. The impairment of creatine metabolism is most extensively studied in creatine deficiency syndromes, which are characterized by genetic deficiencies in either the enzymes involved in creatine biosynthesis or creatine transport. However, other conditions may also influence creatine metabolism to some extent. Our paper also focuses on the transport pathways of these metabolites from their originating tissues to various bodily fluids, typically mediated by the creatine transporter (SLC6A8), with evidence suggesting the involvement of other transporters as well. Gas and liquid chromatography have replaced traditional methods for the analytical detection of biomarkers of creatine metabolism and are now commonly used for this purpose. The paper also discusses the differences and variations between these analytical methods.

1. Introduction

Although creatine (2-methyl-guanidinoacetic acid) is structurally a simple compound, it plays an important role in complex metabolic pathways. This nitrogenous organic acid is primarily known for its function in energy storage and transmission, particularly in the skeletal muscle. However, its activity extends to various other cells and tissues, including the heart, spermatozoa, brain, brown adipose tissue, kidneys, and liver [1]. Many of these tissues accumulate creatine from the bloodstream, either through de novo biosynthesis or intestinal absorption. The initial step of creatine biosynthesis occurs mainly in the kidneys, where arginine-glycine amidinotransferase (AGAT) catalyzes the synthesis of guanidinoacetic acid (GAA) from L-arginine and glycine. After being released into the bloodstream, GAA is absorbed by hepatocytes, where it undergoes further transformation through the action of guanidinoacetate-N-methyltransferase (GAMT). Once synthesized, creatine is transported from the liver into the bloodstream and taken up by cells expressing the creatine transporter (SLC6A8, CRT, or CrT1) [1,2] (Figure 1). Other tissues, such as adipose tissue, brain, skeletal muscle, and testes, also exhibit some capacity for creatine synthesis [3,4,5].

The primary metabolic function of creatine is to provide a rapidly available source of high-energy phosphate for ATP regeneration in the form of phosphocreatine [1]. Within cells, creatine is converted to phosphocreatine by the cytosolic enzyme creatine kinase (CK), which exists in three isoforms: CK-MM (skeletal muscle), CK-MB (cardiac tissue), and CK-BB (brain tissue). Typically, only one cytosolic CK isoform is expressed per tissue [6]. Phosphocreatine is primarily stored in skeletal muscle, although it is equally crucial in regulating cellular energetics within cardiac muscle tissue [1,7]. Although the specific functions of creatine in the central nervous system are not yet fully understood, it likely participates in various ATP-dependent mechanisms essential for brain function and development [8]. Beyond its energy-related functions, creatine also has secondary roles. It exhibits significant antioxidant properties, both directly and indirectly, by upregulating oxidative stress defense enzymes [6,9,10]. More recently, the thermogenic potential of creatine has emerged, with a proposed pathway specific to brown adipose tissue that promotes unique ATP turnover, resulting in heat generation [2,11]. Additionally, creatine activates cell signaling pathways, enhances muscle cell differentiation, and induces the expression of certain transcription factors and other genes [12,13]. Creatine may also function as a neuromodulator and potentially as a neurotransmitter [8,14].

Creatine deficiency syndromes (CDS), first reported in a single patient in 1994 [15], result from defects in creatine biosynthesis or transport, highlighting the importance of creatine as a diagnostic biomarker in these conditions [16,17]. Variations in creatine profiles exist among different population groups due to factors such as pathological conditions (neuromusculoskeletal, cardiovascular, or kidney diseases, carcinoma, or genetic disorders), age, diet, malnutrition, lifestyle, and supplementation [18,19]. Analyzing creatine biomarkers in biological fluids offers a non-invasive, practical, and cost-effective approach to assessing creatine profiles across different populations, serving as potential indicators of creatine metabolism on a broader scale. This method is more accessible compared to invasive biopsy techniques [20] or the technically demanding proton magnetic resonance spectroscopy used for in vivo determination of creatine levels in brain and muscle tissue [21,22,23].

Both creatine and its precursor, GAA, along with creatinine (a product of spontaneous, non-enzymatic creatine degradation), can serve as biomarkers of creatine metabolism. These metabolites are commonly found in plasma, urine, and cerebrospinal fluid (CSF) and can provide insights into physiological conditions [17,24,25,26]. Recent studies have also identified these metabolites in other bodily fluids [27,28,29]. The CK enzyme is also present in some of these fluids. However, CK levels often non-specifically indicate various physiological disorders, and creatinine (Crn) is an established renal function parameter, raising questions about their sole use as indicators of creatine metabolism [30,31,32,33].

For many years, chemical methods were the preferred choice for creatine and creatinine detection, gradually being replaced by enzymatic methods [1]. Today, more advanced analytical methods, such as liquid and gas chromatography, often coupled with mass spectrometry, are employed to determine creatine, GAA, and creatinine levels in biological fluids and tissues [34]. Some studies have also demonstrated the use of capillary electrophoresis for detecting creatine and creatinine [35,36,37]. The present review aims to summarize the current knowledge on the presence of creatine metabolites in biological fluids and to highlight the potential of less commonly studied fluids as sources of these metabolites. The prevalent analytical detection methods are also discussed.

2. Creatine Biomarkers in Blood

Once synthesized in the liver, creatine is released into circulation and travels toward creatine-requiring tissues and cells. In clinical settings, creatine levels can be measured from either serum or plasma following blood sampling. Studies have shown that results from both matrices are comparable, suggesting minimal matrix-related variability [38]. A large-scale study involving over 60,000 subjects reported normal serum creatine values of 62 ± 24.3 µM for males and 63.1 ± 37.1 µM for females, regardless of age [39]. These findings are consistent with results from smaller studies [17,40,41,42,43]. Serum creatine levels are age-dependent, with higher values observed in younger children that gradually decrease with age [17,39,40]. Other factors influencing serum creatine levels include diet [44], physical activity [45], oral supplementation [1], and various pathological conditions [1,17,46].

CDS, rare inborn errors of creatine biosynthesis or transport, result from deficiencies in AGAT, GAMT enzymes, or the creatine transporter (CRT). These disorders typically present with neurological symptoms, including intellectual disability, speech disorders, epilepsy, and, in some cases, autistic behaviors. Plasma creatine levels are notably reduced in patients with GAMT and AGAT deficiencies. Almeida et al. reported plasma creatine levels ranging from 1–5 µM in GAMT-deficient individuals [17], while Verhoeven et al. found levels between 0–7 µM [47]. Bodamer reported creatine values of 5.37 and 8.15 µM in two GAMT-deficient children [40]. Although there is limited research on AGAT deficiency, the condition appears to result in decreased creatine concentrations, with combined creatine and creatinine levels ranging from 15–29 µM in three patients [47]. Conversely, CRT deficiency does not seem to affect plasma creatine levels [17,46].

In patients with urea cycle disorders (UCD), including argininosuccinic aciduria (ASL), citrullinemia (ASS), and ornithine transcarbamylase deficiency (OTC), plasma creatine levels are slightly reduced, ranging from 28 to 100 µM across nine patients [42]. A case study by Threlfall et al. documented decreased plasma creatine levels following renal failure, with concentrations normalizing after a half-month of hemodialysis [48]. Neuromuscular diseases, such as Duchenne muscular dystrophy, Becker muscular dystrophy, spinal muscle atrophy, myositis, and diabetic myopathy, typically result in elevated circulating creatine levels due to disturbances in creatine metabolism. In contrast, a reduction in plasma creatine levels is observed in gyrate atrophy (GA) of the choroid and retina, attributable to ornithine accumulation, which inhibits AGAT activity [1].

Following its synthesis in the kidney via the AGAT-catalyzed reaction, guanidinoacetic acid (GAA) is released into the bloodstream and transported to the liver, where it undergoes methylation to complete creatine synthesis. GAA is detectable in blood samples, albeit at lower concentrations than creatine. Joncquel-Chevalier Curt et al. reported reference plasma GAA values of 1.5 ± 0.6 µM in males and 1.4 ± 0.6 µM in females [39], consistent with findings from other studies [17,40,41,42,43,47,49,50]. Serum GAA levels do not appear to be influenced by sex [39] but tend to increase with age [17,39,47]. However, GAA concentrations decline in older adults, with lower values reported in bedridden elderly patients, such as 1.8 µM [51]. GAMT deficiency leads to GAA accumulation, with reported plasma levels ranging from 12 to 39 µM [46], consistent with other studies [17], [40,47]. In contrast, AGAT deficiency results in significantly reduced GAA levels, with concentrations between 0.01 and 0.04 µM [47]. CRT deficiency does not seem to impact blood GAA levels [46]. Decreased GAA levels have also been observed in uremic patients [52] and in individuals with untreated epilepsy compared to treated patients (1.6 vs. 2.1 µM) [53]. Additionally, patients with UCD and GA exhibit reduced plasma GAA levels, with a five-fold decrease reported in GA cases [1,54].

Approximately 1.7% of total creatine is non-enzymatically converted to creatinine daily, which is then exported from cells to the bloodstream and subsequently excreted by the kidneys. The reference range for serum creatinine, based on a study of 252 healthy subjects, is 63–107 µM for males and 46–103 µM for females [55]. These findings align with those of Finney et al. [56], who conducted a similar study on over 300 healthy individuals. Huang et al. [57] corroborated these values, additionally providing reference ranges for the elderly population, with median serum creatinine levels slightly higher in older adults than in middle-aged individuals, regardless of sex. Serum or plasma creatinine concentrations are known to be dependent on muscle mass and serve as indicators of skeletal muscle mass [25,26]. Abnormal creatinine concentrations are typically indicative of kidney dysfunction. Chronic kidney disease (CKD), characterized by impaired glomerular filtration, can cause serum creatinine levels to rise significantly from normal values to over 1500 µM, depending on the stage of the disease [58].

In summary, serum and plasma creatine concentrations are influenced by a multitude of factors. While no significant gender-related differences have been observed, age-related differences are evident, particularly between younger and older populations. Creatine levels are typically reduced in cases of CDS and abnormally elevated in disorders affecting the muscular system. Circulating GAA levels are low and tend to decrease further with aging. AGAT deficiency significantly reduces GAA concentrations, whereas GAMT deficiency causes a marked increase. Reduced GAA levels are also observed in uremic patients, as well as in individuals with UCD and GA of the choroid and retina. Creatinine, a product of creatine degradation and a marker of kidney function and skeletal muscle mass, shows slight increases in the elderly population and exhibits gender-related variations (

Table 1. Blood concentrations of creatine biomarkers in various conditions.

Different Physiological Conditions	Biomarker Concentration in Blood
Creatine
Normal	62.0 ± 24.3 µM ♂ 63.1 ± 37.1 µM ♀
AGAT deficiency	0–7 µM
GAMT deficiency	15–29 µM (combined with creatinine)
CRT deficiency	Not altered
UCD	Slightly decreased (28–100 µM)
Neuromuscular diseases	Increased *
Gyrate atrophy of the choroid and retina	Decreased *
GAA
Normal	1.5 ± 0.6 µM ♂ 1.4 ± 0.6 µM ♀
AGAT deficiency	0.01–0.04 µM
GAMT deficiency	12–39 µM
CRT deficiency	Not altered
Uremia	Decreased *
Untreated epilepsy	Decreased *
UCD	Decreased *
Gyrate atrophy of the choroid and retina	Decreased (up to five-fold) *
Creatinine
Normal	63–107 µM ♂ 46–103 µM ♀
AGAT deficiency	N/A
GAMT deficiency	N/A
CRT deficiency	N/A
CKD	Normal to greatly increased (depending on the stage) *

* Exact range not provided in the literature; N/A—not available; ♂—in males; ♀—in females.

).

3. Creatine Biomarkers in Urine

Kidney reabsorption of creatine plays a crucial role in maintaining homeostatic control of serum creatine concentration. The mRNA of the creatine transporter (SLC6A8) is highly expressed in mammalian kidneys, particularly in the tubules, and is implicated in this reabsorption process [59,60,61] (accessed April 2024). Although the kidneys efficiently reabsorb creatine in the tubules, trace amounts of this organic acid still appear in excreted urine [59]. To account for individual variability in fluid intake, urine output, and renal function, urinary excretion of various molecules is often normalized to creatinine excretion [62]. Urinary creatine values are sometimes reported as the sum of creatine and creatinine (Cr + Crn) or as the creatine/creatinine ratio (Cr/Crn). Marescau and co-workers (1992) reported urinary creatine excretion levels ranging from 3.44 to 364.25 mmol/mol creatinine in 35 healthy men. A large population study by Joncquel-Chevalier Curt et al. found values of 344.8 ± 350.5 mmol/mol creatinine in men and 371.8 ± 345.5 mmol/mol creatinine in women [39]. Urinary creatine levels can vary with age, as Nasrallah and co-workers (2010) reported concentrations ranging from 157 to 6850 µM in children under 4 years old, 42 to 5120 µM in those aged 4–12 years old, and 33 to 2421 µM in individuals over 12 years old. These age-related findings are consistent with those of Almeida et al. [17]. The Cr/Crn ratio ranges from 0.02 to 0.4 in adults and can be as high as 1.9 in younger subjects [42], in agreement with other studies [17,43]. Cr + Crn was reported as 8.1 ± 3.9 mmol/L across all ages [47]. Reduced urinary creatine concentrations have been observed in four GAMT-deficient patients, as reported by Verhoeven et al. [46]. However, other authors found urinary creatine levels within the normal range for GAMT-deficient patients [17,61]. Cr + Crn levels were also decreased in one GAMT patient [47], while Arias and co-workers (2006) reported a Cr/Crn ratio of 0.80 in one patient. Cr + Crn was also reduced in three AGAT-deficient patients [47]. A deficiency in the creatine transporter impairs efficient creatine reuptake, resulting in significantly increased urinary excretion. Although absolute or relative creatine concentrations may not be markedly altered, the Cr/Crn ratio often provides more informative diagnostic data, as it is noticeably elevated in creatine transporter-deficient patients [17,43,46].

In a study by Struys and co-workers [49], urinary excretion of GAA was reported as 53.9 ± 25.9 mmol/mol creatinine. Joncquel-Chevalier Curt et al. [39] identified GAA excretion as 61.6 ± 37.8 mmol/mol creatinine in males and 73.1 ± 47.5 mmol/mol creatinine in females. Other studies have also reported similar reference values [42,43,50,63]. Age appears to influence urine GAA levels, with children showing higher urinary concentrations than older individuals [17,43,64]. In cases of GAMT deficiency, where GAA methylation is impaired, elevated serum levels are accompanied by increased urinary excretion. Almeida et al. [17] reported urinary GAA levels ranging from 529 to 4368 mmol/mol creatinine in eight GAMT-deficient subjects. These findings are consistent with other studies involving GAMT-deficient patients [43,46,47,49,63]. In AGAT deficiency, where GAA synthesis is disrupted, reduced serum levels are mirrored by decreased urinary excretion. Carducci et al. [47] reported urinary GAA concentrations ranging from 2.4 to 5.8 µM in three AGAT-deficient patients, compared to control values of 313 ± 199 µM. In contrast, urinary GAA excretion is not significantly altered in CRT deficiency [43,46].

Creatinine, an end product of creatine metabolism, is excreted in urine to prevent toxicity. This excretion occurs through both glomerular filtration and tubular secretion. The human organic cation transporter 2 (hOCT2), a member of the SLC22A superfamily, mediates the tubular secretion of creatinine at the basolateral membrane of renal proximal tubular cells [65]. Creatinine is present in urine at much higher concentrations than creatine and GAA. Bader et al. reported median urinary creatinine levels of 1.37 g/L in males and 1.00 g/L in females based on a study of more than 6000 samples [62]. These values are consistent with other studies conducted on larger sample sizes [66,67].

In summary, urinary creatine concentrations exhibit significant variability and are often normalized to creatinine levels to provide more meaningful diagnostic information, particularly in the context of creatine metabolism disorders. Urinary GAA levels show gender differences and are markedly elevated in GAMT-deficient patients but significantly reduced in AGAT deficiency. Urinary creatinine is a valuable marker for assessing creatine and GAA excretion but is primarily linked to kidney function, limiting its specificity as a biomarker for creatine metabolism. Values for creatine and GAA are summarized in

Table 2. Urinary concentrations of creatine biomarkers in various conditions.

Different Physiological Conditions	Biomarker Concentration in Urine
Creatine
Normal	344.8 ± 350.5 mmol/mol Crn ♂ 371.8 ± 345.5 mmol/mol Crn ♀
AGAT deficiency	0–7 µM
GAMT deficiency	Normal to slightly decreased *
CRT deficiency	Greatly increased *
GAA
Normal	61.6 ± 37.8 mmol/mol Crn ♂ 73.1 ± 47.5 mmol/mol Crn ♀
AGAT deficiency	2.4–5.8 mmol/mol Crn
GAMT deficiency	529–4368 mmol/mol Crn
CRT deficiency	Not altered

* Exact range not provided in the literature; ♂—in males; ♀—in females.

.

4. Creatine Biomarkers in CSF

Creatine appears to enter the cerebrospinal fluid (CSF) from two sources: circulating creatine that crosses the blood-cerebrospinal fluid barrier (BCSFB) and creatine originating from the brain. The choroid plexus epithelial cells (CPEC), which form the luminal lining of the BCSFB, express SLC6A8 on their basal side, facilitating creatine uptake from the blood [68] (Figure 2). Brain tissue creatine may either be synthesized de novo within brain cells or enter through the blood–brain barrier (BBB), although this passage is limited. While AGAT and GAMT are expressed in all brain cell types, they are rarely co-expressed in the same cell, necessitating the transport of precursors from AGAT-expressing to GAMT-expressing cells for creatine synthesis. This transport is likely mediated by SLC6A8 [69]. The mechanism of subsequent creatine transport from brain tissue to CSF remains unclear [70,71], although ependymal cells lining the CSF from its abluminal side have been reported to express SLC6A8 in rats [72].

During fetal and perinatal development, the brain predominantly relies on creatine from circulation due to low brain GAMT levels at early stages. The earlier development of the BCSFB compared to the BBB suggests that the BCSFB is the primary structure responsible for exchange between the periphery and brain tissue during these stages, especially given the much higher expression of SLC6A8 in CPEC during development compared to the adult brain [70]. The normal range for CSF creatine has been reported as 17–87 µM in 25 subjects of all ages [17], consistent with other studies that have used CSF to determine creatine concentrations [73,74]. In patients with GAMT deficiency, CSF creatine is significantly decreased, with levels found to be below 2 µM due to impaired enzyme function and GAA accumulation [73,74]. Although creatine levels are decreased in the gray matter of AGAT-deficient patients [75], no studies have specifically addressed CSF creatine profiles in these patients. SLC6A8-deficient patients appear to have normal to slightly elevated CSF creatine levels, suggesting another possible transport mechanism different than that through the SCL6A8 transporter, as proposed by deGrauw and co-workers [76,77,78]. No reports were found concerning creatine export from the BCSFB.

Regarding GAA, the BCSFB primarily functions as a clearance system in the mammalian brain, with the elimination clearance of GAA via the BCSFB being approximately 33-fold greater than its influx clearance from blood to CSF through the BCSFB, as shown in rat studies [68,79]. GAA transport in and out of CSF may be mediated, at least in part, by SLC6A8 and the taurine transporter (TauT, SLC6A6). Studies on isolated rat choroid plexus cells show that GAA can be taken up from blood by the BCSFB through three different mechanisms, two of which are consistent with SLC6A8 and TauT transport, with the third being a much smaller transmembrane diffusion [68]. Both SLC6A8 and TauT appear to be expressed in CPEC, though more dominantly on the apical side [68,80] (Figure 2), favoring export over import through the BCSFB. Uptake of GAA from blood by SLC6A8 is limited under physiological conditions at both the BBB and BCSFB due to 2–50 times higher concentrations of creatine, which competes for the same receptor with higher affinity for creatine transport (the Km value for GAA is ten times higher than that for creatine) [70,71]. The extent to which GAA export is mediated by SLC6A8 at the BCSFB is uncertain, considering the 100- to 1000-fold higher CSF creatine concentration compared to GAA and the consequent transporter saturation [68,71]. No reports were found on GAA transport from ependymal cells to CSF. Almeida and co-workers [17] reported the normal range of CSF GAA as 0.02–0.56 µM in 25 subjects of all ages, consistent with other studies reporting normal ranges [49,50,73,74]. In GAMT-deficient patients, where blood creatine concentrations are low, GAA uptake through the BBB (and possibly at both sides of the BCSFB) via SLC6A8 is favored [70]. Due to the increased transport rate and the brain’s AGAT activity, CSF GAA concentrations in these patients reached values of 14 and 15 µM, as reported by Almeida and Ensenauer [17,73]. In SLC6A8 deficiency, GAA transport between different brain regions necessary for completing creatine synthesis is expected to be disrupted. Indeed, GAA accumulated in the brain of one SLC6A8-deficient patient [81]. However, this is usually not the case, as van de Kamp and co-workers reported normal cerebral GAA levels in the brains of nine SLC6A8-deficient subjects [82]. Van de Kamp and co-workers also reported normal to slightly elevated CSF GAA levels in this deficiency, with values ranging from 0.05–0.44 µM [83].

The accumulation of creatinine, the end product of creatine metabolism, in the brain is mitigated through its elimination via the blood-cerebrospinal fluid barrier (BCSFB). The elimination clearance of creatinine is approximately five times greater than its influx clearance in isolated rat choroid plexus [84]. These elimination clearance values are comparable to the total blood-to-brain and blood-to-CSF influx clearances, indicating a well-regulated transport balance of creatinine between the brain and blood. The production rate of creatinine in the rat brain is estimated to be 0.11 nmol/min·g brain, resulting in higher creatinine concentrations compared to serum. This elevation may be attributed to the constant non-enzymatic degradation of brain creatine and the continuous influx of creatinine from the blood to the brain [84].

SLC6A8 is thought to be involved in the transport of creatinine (both brain-to-blood and vice versa) based on its structural similarity to creatine. However, it is likely that SLC6A8 is significantly saturated with creatine, which has a higher transport priority over creatinine due to the differences in Km values for these molecules and the concentrations of creatine in serum and CSF [84]. Alternatively, organic cation transporters such as human OCT2 and OCT3 (hOCT2 and hOCT3) might contribute to creatinine transport, mediating low-affinity creatinine transport (Figure 2). In rats, a greater contribution of rat OCT3 (rOCT3) over rat OCT2 is estimated [71,84]. rOCT2 and rOCT3 mRNAs are expressed in the choroid plexus, with rOCT2 proposed to be localized on the apical membrane of CPEC [85], while the localization of rOCT3 remains to be determined [84]. Transport of creatinine via rOCT3 was not inhibited in the presence of creatine [84]. In the human choroid plexus, OCT2 and OCT3 mRNA expression is low but comparable [61] (accessed May 2024).

CSF creatinine concentrations were reported as 67.6 ± 27.4 µM in 24 healthy subjects [86]. Schulze reported a range of 30–130 µM for CSF creatinine in 33 healthy subjects [74]. Swahn and Sedvall [87] observed lower CSF creatinine concentrations in 45 schizophrenic patients compared to controls. CSF creatinine levels are also suggested to correlate to some extent with the severity of depression [24]. In one GAMT-deficient patient, the CSF creatinine concentration was 4.5 µM [74]. As with serum and urine, creatinine concentrations in CSF increase in patients with renal insufficiency. In non-dialyzed patients with renal insufficiency, CSF creatinine values ranged from 168–521 µM in 8 patients [86].

CSF primarily functions as a medium for the elimination of these biomolecules, though some uptake from surrounding blood vessels may occur. CSF creatine concentrations are generally within the range observed in blood, while cerebrospinal fluid GAA levels are at least an order of magnitude lower than those in serum or plasma. The transport of these biomolecules through brain tissues is complex, with SLC6A8 appearing to mediate the majority of this transport. The involvement of other transporters, such as SLC6A6 and OCT3, is also plausible. The concentrations of these biomarkers in CSF exhibit changes consistent with expectations in CDS and certain other neurological disorders.

Table 3. CSF concentrations of creatine biomarkers in various conditions.

Different Physiological Conditions	Biomarker Concentration in CSF
Creatine
Normal	17–87 µM
AGAT deficiency	N/A
GAMT deficiency	<2 µM
CRT deficiency	Normal to slightly increased *
GAA
Normal	0.02–0.56 µM
AGAT deficiency	N/A
GAMT deficiency	~14 µM
CRT deficiency	Normal to slightly increased
Creatinine
Normal	30–130 µM
AGAT deficiency	N/A
GAMT deficiency	4.5 µM
CRT deficiency	N/A
Schizophrenia	Decreased *
Renal insufficiency	168–521 µM

* Exact range not provided in the literature; N/A—not available.

shows how different conditions may affect the CSF concentration of these biomarkers. While other methods for assessing creatine biomarkers in biological fluids exist, obtaining CSF remains relatively invasive.

5. Creatine Biomarkers in Saliva

Saliva represents a promising matrix with considerable potential in diagnostic medicine [88,89]. However, there is limited literature concerning salivary creatine biomarkers. As with other biological fluids, SLC6A8 may be responsible for the presence of these metabolites in saliva, given its significant mRNA expression in salivary glands [61] (accessed May 2024). Martinez et al. reported salivary creatine concentrations as 9.24 ± 2.76 µM in males and 10.67 ± 5.07 µM in females over the age of 10 years old. For individuals aged 0–10 years old, the values were 7.65 ± 3.12 µM in males and 8.07 ± 2.98 µM in females. These levels were significantly lower than those found in blood [28]. Suzuki et al. [90] found salivary creatine to be 6.18 ± 2.59 µM in 25 subjects aged 21–54 years old. In patients with chronic kidney disease (CKD), the average salivary creatine level was 14.49 ± 9.15 µM prior to hemodialysis and slightly decreased following treatment. Only one study assessing guanidinoacetate (GAA) concentrations in saliva has been identified. Martinez and co-workers reported salivary GAA levels as 2.38 ± 1.44 µM in males and 2.69 ± 1.60 µM in females over the age of 15 years old. In individuals under 15 years of age, GAA concentrations were 4.02 ± 2.16 µM in males and 3.94 ± 2.92 µM in females. GAA and creatine concentrations in saliva showed no significant gender-related differences, although age-related differences were observed, particularly in females [28]. Salivary creatinine was found to be 4.59 ± 4.33 µM in 25 healthy subjects [90]. In CKD patients, salivary creatinine levels were elevated, with stage-dependent increases. Values were reported as a median of 11 µM and showed a strong correlation with serum creatinine concentrations, suggesting its potential as a diagnostic tool [58]. Suzuki and co-workers [90] reported salivary creatinine concentrations of 64.53 ± 61.88 µM in 11 CKD patients prior to hemodialysis, which decreased to 22.10 ± 16.8 µM after treatment. Overall, these biomarkers are present in saliva at significantly lower concentrations compared to other biofluids, as shown in

Table 4. Salivary concentrations of creatine biomarkers in various conditions.

Different Physiological Conditions	Biomarker Concentration in Saliva
Creatine
Normal	9.24 ± 2.76 µM ♂ (aged 15+) 10.67 ± 5.07 µM ♀ (aged 15+)
AGAT deficiency	N/A
GAMT deficiency	N/A
CRT deficiency	N/A
CKD	14.49 ± 9.15 µM
GAA
Normal	2.38 ± 1.44 µM ♂ (aged 15+) 2.69 ± 1.60 µM ♀ (aged 15+)
AGAT deficiency	N/A
GAMT deficiency	N/A
CRT deficiency	N/A
Creatinine
Normal	4.59 ± 4.33 µM
AGAT deficiency	N/A
GAMT deficiency	N/A
CRT deficiency	N/A
CKD	64.53 ± 61.88 µM

N/A—not available; ♂—in males; ♀—in females.

. While some can serve as indicators of CKD, the full diagnostic potential of saliva remains underexplored. Its ease of collection and rich biomolecular content make it a promising medium for further investigation. The existence and extent of correlations between salivary and serum parameters of creatine metabolism remain an intriguing area for future research.

6. Creatine Biomarkers in Semen

Creatine plays a crucial role in sperm energy metabolism [44]. However, research on creatine biomarkers in semen or seminal plasma is limited. While numerous studies have explored semen creatine kinase content in the context of fertility, only a few have investigated the presence of creatine itself. Although the exact mechanism by which creatine is released into seminal fluid remains uncertain, it is evident that a portion of creatine synthesis occurs within the male mammalian reproductive tract [4,91]. In the testes of rats and mice, creatine synthesis is specific to Sertoli cells in the seminiferous tubules [91], which also exhibit the highest expression of SLC6A8 mRNA in the human male reproductive system [61] (accessed May 2024). Conversely, seminal vesicles, which also express SLC6A8 mRNA, lack the capacity for creatine synthesis, suggesting that creatine may be absorbed from circulation or tubular fluid [4]. It is possible that some of the creatine in Sertoli cells originates from plasma [4]. Therefore, the presence of creatine in seminal plasma could result from both de novo synthesis in the testes and sequestration from plasma by certain reproductive tissues. The most recent study on this topic reported semen creatine levels of 791 ± 342 µM in 53 normozoospermic men, with no significant differences compared to a group of 35 men with asthenozoospermia [29]. An earlier study, which analyzed seminal plasma creatine in normospermic, azoospermic, infertile, and vasectomized men, found differences in creatine levels between normal controls and azoospermic or vasectomized subjects (median levels of 1818 µM, 1249 µM, and 544 µM, respectively) [92]. To date, there are no reports on the presence of guanidinoacetate (GAA) in seminal plasma. A study by Allahkarami et al. [93] appears to be the only one reporting creatinine levels in seminal plasma. In 50 infertile men, the creatinine level was found to be 316.48 ± 99.89 µM. Summarized values are shown in

Table 5. Concentrations of creatine biomarkers in various conditions in seminal fluid.

Different Physiological Conditions	Biomarker Concentration in Semen
Creatine
Normal (normozoospermia)	791 ± 342 µM 1 1818 µM 2
Asthenozoospermia	Normal 1
Azoospermia	1249 µM 2
Vasectomy	544 µM 2
GAA
Normal	N/A
Creatinine
Normal	316.48 ± 99.89 µM

¹ [29]; ² [92]; N/A—not available

. Creatine levels do not seem to have a significant correlation with male fertility, although they are undeniably important for spermatozoa energy utilization. Creatine supplementation may influence seminal creatine concentrations. However, there is no existing research on GAA concentrations in semen, suggesting it may be present in negligibly small amounts.

7. Creatine Biomarkers in Human Milk

Human milk remains one of the least studied biological fluids in this context. The mechanism by which creatine is present in human milk is not fully understood; it is unclear whether creatine is synthesized actively within the breast or diffuses from circulation, potentially via active transport [27]. However, the Human Protein Atlas reports that the expression of SLC6A8 mRNA in the breast is comparable to that in the male reproductive system [61] (accessed May 2024). Human milk provides approximately 9% of the infant’s recommended daily intake of creatine, with the remainder being synthesized de novo by the infant [94]. A cross-sectional study found creatine concentrations in breast milk and maternal plasma to be similar, ranging from 60–70 µM, with no significant changes observed between 1–2 weeks and 5–6 weeks of breastfeeding [95]. However, other studies suggest that creatine is actively enriched in human milk compared to plasma [96,97,98]. An earlier study by Hulsemann et al. reported average creatine concentrations in human milk of 77 µM [96]. In a longitudinal study, creatine concentrations were highest in colostrum and subsequently decreased during the transition to mature milk, with no significant changes thereafter [27]. Another study also observed a decrease in creatine concentration during the first month of lactation [97]. Conversely, a study utilizing NMR-based metabolite profiling reported a slight increase in creatine concentrations from days 31–87 of lactation compared to the initial 9–24 days [98]. Peral and co-workers [99] reported exceptionally high creatine levels, with a median of 865 ± 40 µM in three subjects. The only study addressing guanidinoacetate content in human milk found that GAA levels in breast milk were significantly lower (approximately 0.3 µM) compared to plasma [95]. Creatinine levels in human milk are generally lower than those of creatine and show minimal variation over the course of 6 months of lactation [27]. Spevacek et al. [97] reported an average breast milk creatinine concentration of 45 ± 7 µM, which is consistent with findings from other studies [96,98]. Overall, there appears to be an active enrichment of human milk with creatine, with the highest levels observed in the initial milk, which then decrease over the lactation period. The presence of both creatine and its precursor, as well as its waste product, suggests ongoing creatine metabolism in the mammary tissues. A summary of reference values for creatine and GAA is presented schematically in Figure 3.

8. Analytical Assessment of Creatine Biomarkers

In the early stages of research on creatine metabolism, chemical methods provided valuable tools for analytical detection, some of which continue to be used in modified forms [55,100,101]. The Barritt reaction is utilized to determine creatine concentration by coupling creatine with diacetyl to form a red-colored compound, which is stabilized with α-naphthol. The absorbance of this compound is measured at 546 nm [102]. The Sakaguchi reaction offers a potential chemical method for detecting guanidinoacetate (GAA), though it lacks specificity as it can also detect monosubstituted guanidines such as arginine, argininosuccinic acid, and homoarginine. Despite this limitation, it has proven to be a rapid, efficient, and cost-effective screening method for GAMT deficiency [103,104]. The Jaffe method, which involves the formation of a red-colored molecular complex between creatinine and picric acid in an alkaline solution, has been the most commonly used chemical method since its development in 1886 [105]. It remains widely used for urine samples due to their lower interference compared to serum samples [106]. However, the Jaffe method’s main disadvantage is its lack of specificity due to interference from compounds containing a methylene group, such as glucose, ketones, and proteins, which can lead to elevated readings, while bilirubin can cause reduced values [107]. Consequently, more accurate enzymatic assays are being explored as potential solutions. Significant differences have been reported between creatinine measurements in serum using the standard Jaffe method and enzymatic methods [107].

The most commonly employed enzymatic method involves a series of reactions catalyzed by creatininase, creatinase, sarcosine oxidase, and peroxidase [108]. Another enzymatic approach, which measures the degradation of both creatine and creatinine, uses reactions catalyzed by creatininase, CK, pyruvate kinase, and lactate dehydrogenase. The degradation process is stoichiometrically proportional to the oxidation of NADH to NAD+, which is then measured spectrophotometrically [109,110]. Although enzymatic methods are more specific, they can still be affected by interference from bilirubin or ascorbate when peroxidase is used [1,111]. Enzymatic methods are also more expensive and have lower storage stability compared to chemical methods, which continue to be used, often in modified forms. Knapp and Mayne [112] developed a modified Jaffe method to minimize bilirubin interference. Junge and co-workers [55] reported a modified method that produced results comparable to the enzymatic creatinine PAP (p-amino-phenazone) method.

Gas chromatography coupled with mass spectrometry (GC/MS) offers a more powerful analytical tool for detecting creatine, GAA, and creatinine. This method involves preparing samples by mixing them with toluene, hexafluoroacetylacetone, sodium bicarbonate, and internal standards for each biomarker [17,28,42,43,49,50,113,114,115]. Various derivatization agents, such as mono(trimethylsilyl)trifluoroacetamide (MTSFA) [113,114], trimethylamine and pentafluorobenzyl bromide in acetonitrile [17,28,49], bis(trimethylsilyl)trifluoroacetamide (BSTFA) [42,43], or N-tert-butyldimethylsilyl-N-methyltrifluoroacetamide (MTBSTFA) [115] have been used. Trimethylsilyl derivatization compounds often produce two peaks (monoTMS and diTMS), resulting in lower sensitivity [50,115]. Prieto et al. reported greater stability of tert-butyldimethylsilyl derivatives compared to trimethylsilyl derivatives, with less fragmentation and more sensitive peaks [115]. BSTFA has also been shown to produce reliable results with appropriate internal standards [43]. The choice of derivatization agent depends on the procedure and type of column used.

Regarding internal standards, 2-phenylbutyrate is commonly used [113,114], though it may produce lower sensitivity [50]. Labeled standards, such as [^13C_2]guanidinoacetate, are often employed for stable isotope dilution (SID) GC/MS analysis of GAA [49]. Martinez and colleagues [28] used β-guanidinopropionic acid as an internal standard, while N-methyl-D_3-creatine and D_3-creatinine are used for creatine and creatinine determination, respectively [17,43,116]. SID GC/MS enhances accuracy, sensitivity, and selectivity due to specific derivatization and compensates for variability in sample preparation [40]. This method also allows the analysis of small sample volumes, which is particularly advantageous for limited samples such as cerebrospinal fluid [49]. In urine samples, creatine and GAA concentrations may decrease upon freezing due to precipitation. Brief sonication after thawing can help redissolve precipitates, providing consistent results similar to those obtained with fresh samples [115].

Since the initial reports by Hiraga & Kinoshita [117] and Marescau et al. [118] on liquid chromatography (LC) for guanidine compound determination, this method has evolved and is now widely used alongside GC. LC is suitable for all sample types and is typically coupled with mass detection to achieve high sensitivity and specificity. It can be performed with or without derivatization. Underivatized methods, as reported by Young et al. [119] and Carling et al. [120], offer the advantage of reduced sample preparation time and simultaneous detection of creatinine, creatine, and GAA [90]. Derivatization, usually involving butanolic HCl, enhances sensitivity and produces detectable butyl esters [34,40,119,121]. Other studies also reported derivatization with benzoin for dried blood spot analysis [47,122]. Earlier variations included post-column derivatization with ninhydrin [41,117,118]. Common internal standards added before derivatization include isotope-labeled standards such as D_3-creatine and [^13C_2]guanidinoacetate [40,63,90,119,120], as well as [^2H_2]guanidinoacetate and [^13C, ^15N]guanidinoacetate for GAA [119] and [^2H_3]creatinine for creatinine [90,119,120]. Sample treatment typically involves deproteinization with an organic solvent, such as acetonitrile or methanol, which is particularly useful for urine samples when proteinuria or potential contamination is suspected [34].

Older studies often utilized fluorometric detection [41,47,117,118,122], or UV detection [123]. Modern LC methods are now predominantly coupled with tandem mass spectrometry (LC-MS/MS), which significantly enhances sensitivity and specificity, especially when paired with electrospray ionization (EIS) [34,40,90]. Hydrophilic or reversed-phase columns further improve sensitivity and specificity [34].

A comparative study by Mustafa et al. [124] evaluated three methods—GC/MS, UHPLC/MS, and NMR—for analyzing GAA and creatine in urine. This study found that UHPLC/MS had lower limits of quantification (LOQ) compared to many GC/MS and LC/MS methods, except for one study [121]. NMR spectroscopy showed greater interday and intraday variations in results compared to UHPLC-MS/MS, primarily due to poor performance at lower concentrations. Both UHPLC-MS/MS and NMR spectroscopy were comparable to GC/MS, with UHPLC-MS/MS being preferred for screening for CDS due to its superior performance [124].

The methodological complexity of instrumental techniques can result in slight, often negligible, variations in results across studies. The choice of method in a laboratory depends on the availability of and familiarity with the instrumentation. Comparison between methods could be relevant for diagnosing CDS. LC-MS/MS offers advantages such as simpler sample preparation, faster analysis, and simultaneous creatinine determination, while GC/MS provides superior sensitivity, which may be beneficial for detecting AGAT deficiency or analyzing biofluids with extremely low levels of creatine or GAA. For GAMT or CRT deficiency, either method is suitable for detecting increased levels of creatine and GAA. Plasma and CSF analysis can be useful for ruling out diagnoses, particularly for AGAT deficiency, as plasma creatine levels may be normal in CRT-deficient subjects. Moreover, increased research into other biological fluids has the potential to enhance our understanding and use of creatine metabolism biomarkers.

9. Conclusions

This review compiled current information on biomarkers of creatine metabolism in bodily fluids, including their physiological and non-physiological values and the evolution of detection methods. Creatine and its metabolites in biological fluids are influenced by various factors. Gender has minimal impact on creatine and GAA concentrations, with only urinary GAA excretion differing between genders. In contrast, creatinine levels in blood and urine show some variation between males and females. Age significantly affects levels, with higher values observed in developmental stages and childhood, decreasing with age. Diet and pathological conditions could also influence creatine metabolism, with enzyme and transporter deficiencies having notable effects. Although the precise influence of each factor on different biofluids remains unclear, this review is intended to support future research and hypotheses in the field. Current analytical methods are sufficient for expanding the range of sample types for creatine metabolism assessment beyond blood and urine. The primary concern remains whether biomarkers’ concentrations in various fluids accurately reflect physiological conditions and their intercorrelation.