Gamma-hydroxybutyric acid (GHB) is an endogenous substance produced during the metabolism of gamma-aminobutyric acid (GABA), a major inhibitory neurotransmitter, in the human brain. GHB exerts sedative, hypnotic, and anesthetic effects by binding to the B sub-type of the GABA receptor and to a GHB-specific receptor in the central nervous system (CNS). It is also present in small amounts in natural food, such as some wines, meats, and fruits but exerts negligible effect on the human body upon ingestion. The sodium salt of GHB, sodium oxybate, used as a CNS depressant in the treatment of narcolepsy and alcoholism, is often misused in drug-facilitated crimes (Busardo and Jones, 2019; Trombley et al., 2019).

Numerous clinical and forensic studies on the distribution of endogenous GHB in human urine or blood have been performed to differentiate between endogenous and exogenous GHB in the body (Andresen et al., 2010; Busardo and Jones, 2019; Strickland et al., 2020). According to the reference range for urine samples, a cut-off value of 10 μg/ml GHB is generally accepted for forensic purposes; however, the analytical results in actual cases are at risk of false positive or false negative interpretation due to the in vitro production and rapid metabolism, with a half-life of 30–50 min, of GHB (Kang et al., 2014; Busardo and Jones, 2019). To circumvent these issues, metabolomics studies have been performed to identify alternative GHB intoxication markers (Jung et al., 2021). Urine metabolomics following GHB exposure have demonstrated that GHB administration to rats alters the metabolism of amino acids (Seo et al., 2018), polyamines (Lee et al., 2019), and organic acids related to the tricarboxylic acid (TCA) cycle (Seo et al., 2016b). Another study confirmed glycolic acid and succinic acid as potent urinary markers of GHB exposure (Palomino-Schatzlein et al., 2017). Further, the potential of metabolites such as 2,4-dihydroxybutyric acid (2,4-OH-BA) and 3,4-dihydroxybutyric acid (3,4-OH-BA) in phase 1 and GHB-glucuronide and GHB-sulfate in phase 2, as alternative markers of GHB exposure, has been investigated (Jarsiah et al., 2020; Jung et al., 2021). Moreover, new urinary GHB biomarkers, including succinylcarnitine and GHB-conjugates (GHB-carnitine, GHB-glutamate, and GHB-glycine), have been identified (Steuer et al., 2019).

However, despite the identification of potential diagnostic biomarkers of GHB intoxication, GHB continues to be used primarily in clinical and forensic settings. One of the main reasons for this is that the reference concentrations of the potential biomarkers, which are also endogenous metabolites, have not been standardized yet. Moreover, intra-individual variations that depend on circadian rhythm or diet and inter-individual variations, such as gender and age, possibly affect the concentrations of the endogenous metabolites (Slupsky et al., 2007; Bertram et al., 2009; Dallmann et al., 2012; van Duynhoven and Jacobs, 2016). Therefore, the feasibility of the potential biomarker metabolites of GHB exposure needs to be examined keeping the inter- and intra-individual variations under consideration. In this study, we selected the most promising biomarker metabolites of GHB exposure (glutamic acid, GABA, succinic acid, 2,4-OH-BA, 3,4-OH-BA, glycolic acid, and succinylcarnitine) [Figure 1, (Kim et al., 2022)] based on the results of previous metabolic studies (Jarsiah et al., 2020; Jung et al., 2021) and investigated their urinary profiles in 206 healthy adult Korean females. The concentrations of endogenous GHB and biomarker metabolites in 472 urine samples collected from these females at various time points were quantified using fully validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) followed by comparisons based on age and day and time of sample collection. Furthermore, metabolites having a high correlation with endogenous urinary GHB were identified.

Although previous studies have identified diagnostic biomarkers of GHB intoxication (Seo et al., 2016b; Palomino-Schatzlein et al., 2017; Seo et al., 2018; Lee et al., 2019; Steuer et al., 2019; Jarsiah et al., 2020; Jung et al., 2021), the use of metabolic precursors or products of GHB as alternative markers to differentiate between exogenous and endogenous GHB has had some limitations in clinical and forensic practice, primarily due to the lack of human data. In this study, 472 urine samples from 206 healthy females were analyzed to propose population reference concentrations to facilitate clinical and forensic decisions related to GHB intoxication. To the best of our knowledge, this is the first report of unadjusted and creatinine-adjusted concentrations of GHB and its biomarker metabolites in a significantly large number of human urine samples.

Wide variations have been reported in the endogenous concentrations of GHB in human urine in previous studies (Brailsford et al., 2010). In the present study, the maximum concentration of endogenous GHB detected in urine was 0.75 μg/ml, which is much lower than that previously reported (Brailsford et al., 2010). It is presumed that in vitro production during transport or improper storage conditions was the reason for this discrepancy. In previous studies on endogenous GHB concentration in urine, the samples were stored at −20°C (for a maximum of 36 months), 4°C, or 5°C (for a maximum of 33 days) (LeBeau et al., 2002; Elliott, 2003; Yeatman and Reid, 2003; Crookes et al., 2004; LeBeau et al., 2006; Brailsford et al., 2010). The urine samples in the current study were stored at −70°C immediately following collection and handled in an ice bath during the entire course of sample preparation, based on the results of the previous stability experiment (Kang et al., 2014). Few studies have been previously conducted on the endogenous urinary concentrations of biomarker metabolites of GHB. In a previous study on determining 10 urinary neurotransmitters and their metabolites in urine samples of six males and four females, glutamic acid and GABA levels were measured, and their concentrations were 3.050 ± 0.150 and 0.543 ± 0.036 μg/ml, respectively (Yan et al., 2017). Further, phase 1 metabolite and GHB levels were measured in urine samples of 77 males and 55 females with ranges as follows: GHB, not detected–1.94 μg/ml, glycolic acid, 1.30–400 μg/ml; succinic acid, 1.17–273 μg/ml; 2,4-OH-BA, 0.72–26.2 μg/ml; and 3,4-OH-BA, 1.88–122 μg/ml (Jarsiah et al., 2020). Most concentrations determined in the current study were lower than the previously reported maximum concentrations.

Creatinine adjustment of the concentration of drugs or metabolites in urine improves the reliability of the quantitative results as it compensates for the individual variations in the physiological dilution of urine (Lafolie et al., 1991). In a previous study on changes in urinary drug concentrations after the administration of heroin, cannabis, or cocaine followed by water intake in healthy adults, normal water intake did not significantly affect the drug-positive ratio of the participants when normalized by creatinine concentration. On the contrary, the creatinine concentration decreased and the drug-positive ratios of the participants increased due to the correction by creatinine concentration in excessively hydrated subjects (Cone et al., 2009). This approach could misestimate urinary drug or metabolite concentrations at the extremes of age and body size as creatinine excretion is determined by muscle mass (Dong et al., 2021), but it is widely used as a correction method for the quantitative results of drugs or metabolites in urine samples from healthy individuals (Barr et al., 2005). The weak correlation between unadjusted and creatinine-adjusted urinary GHB concentrations has been reported (Kang et al., 2014). As urinary volume is influenced by various factors such as kidney function, fluid intake, and perspiration (Carrieri et al., 2001), creatinine adjustment of the concentrations of GHB and its biomarker metabolites is recommended to compensate for individual variability in spot urine sampling.

It has been previously reported that creatinine is one of the human urinary metabolites that enables the classification of participants under and over 40 years old; the concentrations of creatinine in the former were higher than those in the latter (Slupsky et al., 2007). In the current study, neither the unadjusted nor the creatinine-adjusted concentrations of GHB were affected by differences in age. Serum metabolite profiling studies based on gender and age differences among the Japanese population have demonstrated that gender and age are key contributors clustering the subjects according to serum metabolites. Age-associated differences were observed in 24 and 23% of 516 endogenous serum metabolites in men and women, respectively, and the concentrations of glutamic acid and succinylcarnitine was higher in subjects aged 55–65 years as compared to subjects aged 25–35 years (Saito et al., 2016). Another metabolomics study had reported that the plasma concentrations of glutamic acid and succinic acid significantly increased and that of glycolic acid significantly decreased in older mice (72 weeks) as compared to those in young mice (8 weeks) (Seo et al., 2016a). The results of the present study are consistent with those in these two previous studies. Further, in a previous review on the aging metabolome, the central role of the TCA cycle in signaling and metabolic dysregulation associated with aging was emphasized (Sharma and Ramanathan, 2020).

Besides gender and age differences, a variety of intra-variable factors such as circadian rhythms (Slupsky et al., 2007; Bertram et al., 2009; Dallmann et al., 2012) and diet (van Duynhoven and Jacobs, 2016; Kessler and Pivovarova-Ramich, 2019) affect the human metabolome. In particular, a close relationship between the circadian clock and metabolic change has been demonstrated (Bellet and Sassone-Corsi, 2010; Masri and Sassone-Corsi, 2013; Davies et al., 2014). In the current study, significant inter-day variation of GABA concentration and intra-day variation of 3,4-OH-BA and succinylcarnitine concentrations were observed. A distinct time-of-day variation with a significant cosine fit during the wake/sleep cycle and during 24 h of wakefulness was observed in an untargeted and targeted analysis of two-hourly plasma samples in a previous study. Of the 171 metabolites quantified, daily rhythms were observed in the majority (n = 109), with 78 of these maintaining their rhythmicity during 24 h of wakefulness, most with reduced amplitude (n = 66) (Davies et al., 2014). It has been previously demonstrated that the metabolome of plasma and saliva is affected by the circadian clock; the concentrations of approximately 15% of all identified metabolites in plasma and saliva were altered during 40 h of extended wakefulness (Dallmann et al., 2012). In another study, the concentration of creatinine in urine collected during morning hours was significantly higher than that in the afternoon (Slupsky et al., 2007), which is consistent with our results.

As shown in Table 3, the urinary creatinine concentration increased significantly at 18:00–23:00 and 4:00–10:00 h. In previous human studies, elevated creatinine concentrations were observed in the morning, which could be because the urine is concentrated overnight (Barr et al., 2005; Slupsky et al., 2007). In another study, the creatinine concentrations were measured in urine collected from healthy adults at 9:30, 12:00, 14:30, 17:30, and 22:00 h, and the concentrations were higher at 9:30 and 22:00 h. The authors of the study found that the creatinine excretion rate in each individual was positively correlated with the urinary flow rate (Sallsten and Barregard, 2021). The creatinine concentrations in all urine samples in our study were acceptable for the criterion for screening drugs of abuse (≥ 5 mg/dl) set in the United States Department of Transportation Regulations (Barbanel et al., 2002). Therefore, the significant elevation in the creatinine concentration at 18:00–23:00 and 4:00–10:00 h was probably due to overnight urine concentration as well as decreased fluid intake and urine flow rate during evening and at night.

The biomarker metabolites quantified in the current study have been previously proposed as potential markers for GHB intoxication, but their correlation with GHB has not been investigated, which makes it difficult to utilize them in clinical and forensic practice. In this study, the urinary concentrations of 2,4-OH-BA, succinic acid, and 3,4-OH-BA showed the highest correlation with that of GHB. Succinic acid is not only one of the phase 1 metabolites of GHB but also an intermediate of the TCA cycle. Moreover, its role in pathological conditions such as inflammation and tumorigenesis has been emphasized (Mills and O'Neill, 2014; Tretter et al., 2016). Therefore, it appears to be unsuitable as a distinct marker for GHB intoxication. Nevertheless, the high correlation of 2,4-OH-BA or 3,4-OH-BA concentrations with those of GHB strongly supports their potential as direct markers of GHB intoxication.

In conclusion, this study provides data on urinary profiling of endogenous GHB and its biomarker metabolites in urine samples of 206 healthy females based on age and the time of sample collection. The data of the current study suggest population reference concentrations to facilitate clinical and forensic decisions associated with GHB-related intoxication and could prove useful in the verification of biomarker metabolites upon comparison with urinary profiles of GHB-administered populations.