ANDROLOGY ORIGINAL ARTICLE

Transcription

1 ISSN: ORIGINAL ARTICLE Correspondence Yvonne Lood, MSc, National Board of Forensic Medicine, Department of Forensic Genetics and Forensic Toxicology, Artillerigatan 12, S , Link oping, Sweden. Keywords: doping, gender dysphoria, hypogonadism, saliva, testosterone Received: 3-Mar-2017 Revised: 8-Sep-2017 Accepted: 20-Sep-2017 doi: /andr Relationship between testosterone in serum, saliva and urine during treatment with intramuscular testosterone undecanoate in gender dysphoria and male hypogonadism 1,2 Y. Lood, 3 E. Aardal-Eriksson, 4 C. Webe, 1,2 J. Ahlner, 4 B. Ekman and 4 J. Wahlberg 1 National Board of Forensic Medicine, Department of Forensic Genetics and Forensic Toxicology, Link oping, Sweden, 2 Department of Medical and Health Sciences, Link oping University, Link oping, Sweden, 3 Division of Clinical Chemistry, Department of Clinical and Experimental Medicine, Link oping University, Link oping, Sweden, and 4 Department of Endocrinology, Department of Medical and Health Sciences, Link oping University, Link oping, Sweden SUMMARY Long-term testosterone replacement therapy is mainly monitored by trough levels of serum testosterone (S-T), while urinary testosterone (U-T) is used by forensic toxicology to evaluate testosterone doping. Testosterone in saliva (Sal-T) may provide additional information and simplify the sample collection. We aimed to investigate the relationships between testosterone measured in saliva, serum and urine during standard treatment with 1,000 mg testosterone undecanoate (TU) every 12th week during 1 year. This was an observational study. Males with primary and secondary hypogonadism (HG; n = 23), subjects with gender dysphoria (GD FtM; n = 15) and a healthy control group of men (n = 32) were investigated. Sal-T, S-T and U-T were measured before and after TU injections. Sal-T was determined with Salimetrics â enzyme immunoassay, S-T with Roche Elecsys â testosterone II assay and U-T by gas chromatography-mass spectrometry. Sal-T correlated significantly with S-T and calculated free testosterone in both controls and patients (HG men and GD FtM), while Sal-T to U-T showed weaker correlations. Trough values of Sal-T after 12 months were significantly higher in the GD FtM group ( nmol/l) compared to HG men ( nmol/l) and controls ( nmol/l), while no differences between S-T and U-T trough values were found. Markedly elevated concentrations of salivary testosterone, 7 14 days after injection, were observed, especially in the GD FtM group. This study demonstrates that Sal-T might be a useful clinical tool to monitor long-term testosterone replacement therapy and might give additional information in forensic cases. INTRODUCTION Androgen deficiency and gender dysphoria in biological female with male identity (GD FtM) require lifetime testosterone therapy which has to be safe and effective. Male hypogonadism, defined as the failure of the testes to produce adequate amount of testosterone caused by testicular disorder (primary) or resulting from hypothalamic pituitary gonadal dysfunction (secondary), has been treated with testosterone for more than six decades (Schubert et al., 2004; Nieschlag, 2015). Testosterone is also used in cross-sex hormone treatment of GD FtM subjects, to develop and maintain male secondary sex characteristics (Mueller et al., 2007; Meriggiola & Berra, 2013; Pelusi et al., 2014), generally following the same dosage regimen scheme as for replacement therapy in hypogonadal men. There is a need for appropriate and accurate methods to determine the free or bioavailable testosterone fraction in serum in a simple and economical defensible way in clinical assessment. The circulating testosterone is in about 97 98% bound to plasma proteins. In normal men, 44% of the total concentration of testosterone is bound to sex hormone-binding globulin (SHBG) and the remaining major part is with much lower affinity bound to albumin, leaving about 2% of free circulating testosterone. In women, 66% of the total testosterone is bound to SHBG and only about 1% is unbound (Dunn et al., 1981). The combination of albumin-bound plus free testosterone is referred to as bioavailable testosterone (Arregger et al., 2007; Shea et al., 2014). Free testosterone is considered the most physiologically relevant fraction, but is rarely measured in clinical routine because these approaches are technical difficult, time-consuming and 86 Andrology, 2018, 6, American Society of Andrology and European Academy of Andrology

2 SALIVARY LEVELS OF TESTOSTERONE expensive (Matsumoto & Bremner, 2005; Rosner et al., 2007; Shea et al., 2014). Total serum testosterone (S-T) is used in the assessment of androgen dysfunction in clinical routine and sometimes in combination with the free androgen index (FAI), calculated as S-T/SHBG, as a measurement of estimated free testosterone. However, this ratio is not generally considered a reliable index of bioavailable testosterone, especially in men (Kapoor et al., 1993). Moreover, various formulas, based on S-T, SHBG and albumin, can be used to calculate free-and bioavailable testosterone (S odergard et al., 1982; Vermeulen et al., 1999; de Ronde et al., 2006). These calculations are, however, strongly dependent on the validity of the included parameters. In forensic medicine and anti-doping testing, urine is currently the primary biological material used to detect illegal intake of testosterone (Sj oqvist et al., 2008; Lood et al., 2012), based on the determination of the testosterone glucuronide/ epitestosterone glucuronide ratio (T/E; Konieczna et al., 2011), affected by numerous variations in renal excretion and metabolism (Schulze et al., 2008). Salivary testosterone (Sal-T) has recently been described as a reliable approach in adult males and females in clinical treatment (Arregger et al., 2007; Bui et al., 2013; Keevil et al., 2014) and discussed as a potential complement in doping control (Thieme et al., 2013; Anizan & Huestis, 2014). Sal-T is assumed to reflect the free circulating testosterone levels in serum (Rey et al., 1988; Rilling et al., 1996) and may therefore be a valuable index of the biological activity of testosterone, independent of variations in circulating SHBG and albumin (Keevil et al., 2016). Significant increase in Sal-T concentrations following therapeutic transdermal administration was observed in a previous study, and the oral fluid was assumed to be a more sensitive marker of transdermal testosterone administration than changes in urinary testosterone (U-T) concentrations (Thieme et al., 2013). Moreover, saliva collection has advantages of being non-invasive and stress-free, and the easy self-sampling procedure might facilitate frequent monitoring to adjust for any intra-individual variations. The primary aim of this study was to investigate the relationship between Sal-T, S-T, calculated free testosterone (Free-T) and U-T during treatment with intramuscular injections of testosterone undecanoate 1000 mg every 12th week during 1 year and to evaluate the potential of saliva in monitoring longterm testosterone replacement therapy. The secondary aim was to obtain concentration ranges of Sal-T during standard treatment with testosterone undecanoate compared with control subjects with normal testosterone secretion. METHODS Subjects Forty patients, 23 males with primary or secondary hypogonadism (HG) and 17 GD FtM subjects were recruited and a control group of 32 healthy men was included. Seventeen of the patients (10 HG, 7 GD FtM) had been treated with testosterone undecanoate (1000 mg every 12th week i.m) for at least 1 year (non-naive) and 23 patients (13 HG, 10 GD FtM) initiated their therapy at the study start (naive). Seven of the GD FtM had undergone bilateral oophorectomy before the start of the study. Two non-naive GD FtMs, who discontinued the study program on their own initiative, were excluded. Ethics The study was approved by the regional Ethical Board in Link oping (DnrM64-09), and a written consent was obtained from all subjects participating in the study. Study design The patients were treated with long-acting testosterone undecanoate (TU), Nebido â, Bayer 1000 mg i.m. administered every 12th week, according to current clinical routine. Saliva, blood and urine were collected at baseline prior to injection and 4, 7, 14 and 28 days after the first injection and thereafter prior to injection every 12th week during 1 year. The naive patients had the second TU injection after 6 weeks and thereafter at 12-week intervals. Thirty of the patients (79%) extended the test period with another 28 days after the last TU injection 1 year after inclusion in the study, following the same sampling sequence as described above. Venous blood for determination of serum testosterone (S-T) and S-SHBG was collected (before 10:00 AM) in clot activator tubes and in lithium heparin tubes for determination of P-albumin. Saliva was collected prior to blood sampling using Salimetrics â oral swabs and Salimetrics â storage tubes for determination of Sal-T. Blood and saliva samples were centrifuged at 2500 g for 10 min. Serum and plasma samples were assayed in daily routine at the laboratory, and saliva samples were stored at 20 C until assayed in batch. The first morning urine were collected and frozen at 80 C until batch analysis. Analytical methods S-T was determined with the Roche Elecsys â testosterone II assay using a high-affinity sheep MAB with electrochemiluminescence detection on a cobas e602 (Roche, Basel, Switzerland). Between-day precision (CV) during 1 year was 5% at concentrations of 2 and 20 nmol/l, and the measuring range was nmol/l. SHBG was determined using cobas e602 and P-albumin by turbidimetry on ADVIA 1800 (Siemens, Munich, Germany). Sal-T was determined with the Salimetrics â expanded range enzyme immunassay kit (Salimetrics LLC, State College, PA, USA) applied on a Tecan Freedom Evolyzer (Tecan AG, M annedorf, Switzerland). Between-day precision (CV) during 1 year was 12% at 0.05 nmol/l and 6% at 1 nmol/l. The measuring range was nmol/l. Saliva was checked for haemoglobin with Hemocue Plasma/Hb Low (Hemocue AB, Angelholm, Sweden), and no samples were found to contain haemoglobin >0.1 g/l. Determination of Sal-T, S-T, S-SHBG and P-albumin was performed at the Laboratory of Clinical Chemistry, University Hospital in Link oping. Free and glucuronide testosterone in urine (U-T) was determined by gas chromatography-mass spectrometry (GC-MS) at the Department of Forensic Toxicology in Link oping. The U-T concentrations were adjusted to creatinine determined by Jaffe-assay on an ADVIA 1800 (Siemens, Munich, Germany). The urinary levels of testosterone were determined after hydrolysis of the conjugates (Masse et al., 1989). GC-MS analysis was carried out on a 7890A gas chromatograph coupled with a 5975C mass spectrometer (Agilent Technologies, Palo Alto, California, USA). The separation was carried out on a HP Ultra 1 cross-linked methyl silicone capillary column (20 m mm 2017 American Society of Andrology and European Academy of Andrology Andrology, 2018, 6,

3 Y. Lood et al. i.d., film thickness 0.11 lm). The between-day precision (CV) during 1 year was 12.4% at 5 ng/ml (17.3 nmol/l) and 10.1% at 100 ng/ml (346.7 nmol/l). The measuring range was ng/ml ( nmol/l). Free-T was calculated using the Vermeulen equation, based on the measurement of total testosterone, SHBG and albumin concentrations (Vermeulen et al., 1999). The maximum testosterone concentrations were derived from the subject s individual data, seen between days 7 and 14, in accordance with C max in a previous study (Behre et al., 1999). Statistical analyses Statistical calculations were mainly taken with IBM SPSS Statistics 23. Between-groups comparisons were performed by independent samples t-test and paired t-test. Correlations between the different matrices and against body composition were evaluated using Pearson 0 s coefficient. Results are given in mean SD, and 95% CI in text, tables and figures. Statistical significance was considered at 5% level (p 0.05). In comparison with mean ratio of testosterone per kg bodyweight between HG men and GD FtM, one GD FtM subject with a body weight of 148 kg was excluded. RESULTS Anthropometric characteristics The mean age SD of the HG, GD FtM and controls was years (range: years), years (range: years) and years (range: years), respectively, at the start of the study. Significant differences in height (p < 0.001) and age (p = 0.035) were noted between GD FtM and both HG and controls and in weight (p = 0.021) between GD FtM and HG, but there were no significant differences in BMI or waist between the groups (Table 1). Control and baseline values Testosterone mean values SD in the control group were as follows: S-T nmol/l ( nmol/l), Sal-T nmol/l ( nmol/l) and U-T nmol/l ( nmol/mmol creatinine; Table 2). Baseline S-T levels were significantly lower in both HG (p < 0.001) and GD FtM (p = 0.004) groups than in the control group. In the HG group, Free-T (p = 0.001) and SHBG (p = 0.016) were also significantly lower than in controls. No significant differences in Sal-T or U-T levels were noted (Table 2). High baseline Sal-T concentrations ( nmol/l) were measured in four subjects, three GD FtM and one HG man. Table 1 Anthropometric characteristics of patients and control subjects HG n = 23 GD FtM n = 15 C n = 32 Age (years) * Height (cm) * Weight (kg) BMI (kg/m 2 ) Waist (cm) Mean 1 SD, HG, hypogonadal men; GD FtM, gender dysphoria, female to male; C, controls. *Significant differences compared to controls; p < Significant differences compared to HG men; p < Table 2 Concentrations of testosterone and SHBG at baseline and at trough level after 12 months testosterone undecanoate treatment (prior to the fifth injection for the non-naive patients and prior to the sixth injection for the naive patients) HG GD FtM C Mean SD Mean SD Mean SD S-T (nmol/l) Baseline * * months Free-T (nmol/l) Baseline * months * Sal-T (nmol/l) Baseline months * SHBG (nmol/l) Baseline * months * S-T/SHBG (FAI) Baseline months * * U-T (nmol/mmolcrea) Baseline months HG, hypogonadal men (n = 23); GD FtM, gender dysphoria; female to male (n = 15); C=controls (n = 32). S-T, serum testosterone; Free-T, calculated free testosterone; Sal-T, salivary testosterone; U-T, urinary testosterone; FAI, free androgen index. *Significant differences compared to controls; p < Significant differences compared to HG men; p < Correlations Sal-T correlated significantly to S-T and Free-T in both controls and patients. In controls, the correlation between Sal-T and Free-T was stronger (r = 0.70, p < 0.001) than the correlation between Sal-T and S-T (r = 0.35, p = 0.047). In patients, the correlation between Sal-T and Free-T (r = 0.74, p < 0.001) for all measured values for all patients was the same as between Sal-T and S-T (r = 0.74, p < 0.001), with a weaker correlation between Sal-T and U-T (r = 0.40, p < 0.001). Within the GD FtM and HG groups, the correlations between Sal-T and Free-T were stronger in the GD FtM group compared to the HG men (p < in all cases; Fig. 1). When tested at each sampling point, the best correlations between Sal-T and S-T were seen at higher levels of testosterone (Fig. 2). Sal-T, S-T and U-T showed comparable kinetic pattern over time for all patients during one year, but at the highest concentration levels after injections, large inter-individual variations were noted (Fig. 3). Sal-T and S-T were highly negatively correlated to weight, BMI and waist at maximum levels 7 14 days after injection for all patients (p < 0.01 in all cases), but not at trough levels after 12-month TU treatment. The correlation between testosterone and body constitution was strongest between S-T and waist in the HG group (r = 0.85, p < 0.001), between S-T and BMI in the GD FtM group (r = 0.42, not significant), and in the control group, the only significant correlation noted was between S-T and waist (r = 0.41, p = 0.021). Trough values Trough values measured after 12 months in all HG and GD FtM patients, prior to the fifth injection for the non-naives and 88 Andrology, 2018, 6, American Society of Andrology and European Academy of Andrology

4 SALIVARY LEVELS OF TESTOSTERONE Figure 1 Correlations between salivary testosterone vs. serum, calculated free and urinary testosterone in the control group (left, dotted lines represent 95% CI) and all measured values in the hypogonadal men (HG) and subjects with gender dysphoria (GD FtM, right). prior to the sixth injection for the naives, were by mean: S-T nmol/l, Sal-T and U-T nmol/mmol creatinine. GD FtM showed significantly higher Sal-T values after 1 year ( nmol/l) compared to HG men ( nmol/l) and controls ( nmol/l), but no significant 2017 American Society of Andrology and European Academy of Andrology Andrology, 2018, 6,

5 Y. Lood et al. (A) Pearson s correlation (r) between salivary testosterone vs. serum testsoterone (B) Day Figure 2 Correlation between salivary vs. serum (A) and calculated free testosterone (B) measured at baseline (day 0) and after 4, 7, 14 and 28 days after the first testosterone injection and the injection after 12 months (340, 343, 350, 366) and prior the injections at day 84, 168, 252 and 336. Shadow bars indicate significant correlation (p < 0.05). [Colour figure can be viewed at wileyonlinelibrary.com]. Pearson s corre lation (r) between salivary testosterone vs. calculated free testosterone Day Figure 3 Testosterone concentrations (mean 95% CI) for all measured values in saliva (left) and serum and urine (right) in patients during intramuscular testosterone undecanoate injections every 12th week during 12 months and first and last injections. [Colour figure can be viewed at wileyonlinelibrary.com]. differences in S-T or U-T were found between the groups (Table 2). S-T, Free-T and Sal-T were significantly elevated in the naive HG and GD FtM group (p < 0.001, p < and p = 0.050), compared to baseline. Ninety-five per cent of the subjects had an S-T concentration within the reference range of a normal man after 12 months (<50 years: nmol/l, 50 years: nmol/l). Trough Sal-T, S-T and U-T were by mean 13, 16 and 27% lower among the GD FtM individuals being oophorectomized compared to the group with remaining ovaries, but the differences were not significant. Maximum values The highest (maximum) measured values of Sal-T, S-T and U-T were observed at day 7 14 after TU injection, except in one 90 Andrology, 2018, 6, American Society of Andrology and European Academy of Andrology

6 SALIVARY LEVELS OF TESTOSTERONE subject, with maximum S-T and Sal-T values noted after 4 days. The mean maximum S-T concentrations measured for all patients after the first TU injection and the last injection given after 12 months were nmol/l (n = 61). Corresponding Sal-T levels were nmol/l (n = 66) and U-T nmol/mmol creatinine (n = 64). There were no significant differences in mean Sal-T, S-T or U- T levels neither between the GD FtM and HG groups nor between the maximum levels measured after the first TU injection and the injection after 12 months. Markedly elevated testosterone values were, however, observed in some GD FtM (see Fig. 2). Significantly, higher mean ratio of maximum Sal-T per kg body weight was noted in GD FtM compared to HG men (p = 0.03). DISCUSSION This observational study is the first report evaluating concentrations of Sal-T, S-T and U-T simultaneously in HG men and GD FtM during treatment with injectable long-acting TU during one year. The main finding in our study was that Sal-T significantly corresponds to both S-T and Free-T during intramuscular TU therapy. However, the correlation between Sal-T and U-T was weaker and with large inter-individual variations. U-T concentrations are influenced by numerous variations in biotransformation and renal excretion that may contribute to the inter-individual variations (Coviello et al., 2006; Anizan & Huestis, 2014). We suggest that Sal-T has potential as a tool in clinical assessment and might add information for optimal dosing of long-acting testosterone injections to HG men and GD FtM. Testosterone doses given to GD FtM are usually based on the treatment schemes of HG men (Hembree et al., 2009). Using standard doses of long-acting TU, mean Sal-T trough levels after one year treatment were found to be significantly higher in GD FtM compared to HG men and controls. Salivary glands metabolism might have some impact on the alteration in Sal-T concentrations, but other factors such as body size, body composition (Mongrawe-Chaffin et al., 2015), remaining ovaries and gender differences in protein binding and steroid synthesis may affect the results as well. GD FtM very often exhibit reduced body mass due to their female phenotype, and perhaps, it is necessary to reduce the dose of TU in order to maintain physiological testosterone levels. It is noteworthy that slightly lower trough testosterone levels were observed in the oophorectomized GD FtM compared to the GD FtM with remaining ovaries maybe explained by the absence of ovarian testosterone production (Kotsopoulos et al., 2015; Rech et al., 2016). FAI normally increases as a function of the exceeded SHBG saturation limit (Krieg & Arning, 1978; Dunn et al., 1981) and was markedly elevated compared to S-T and Free-T in GD FtM. This might indicate a higher bioavailable testosterone fraction maybe more accentuated in GD FtM (Table 2), even though earlier studies have shown that FAI is not a reliable parameter of Free-T (Kapoor et al., 1993; Vermeulen et al., 1999). Mean S-T, Free-T, S-T/SHBG and U-T increased in HG men and GD FtM after 12 months treatment, while mean Sal-T at 12 months did not seem to increase in the same way. The extremely high initial Sal-T measured in four subjects, influenced the mean baseline values, which might be explained by self-administration of testosterone gel previously prescribed. Although Sal-T is supposed to mirror the serum-free fraction, we found higher Sal-T concentrations compared to Free-T in both HG men and GD FtM, which may be explained by crossreactivity with other steroids in saliva measured by immunoassay (DHT 36%). Immunoassay is the standard method used in clinical practice and was the only available method at our laboratory at the time of the study. However, a more specific and sensitive analytical technique, such as liquid chromatographytandem mass spectrometry (LC-MS/MS) is preferable. The steroid metabolizing enzymes 17b-hydroxysteroid dehydrogenase and 5a-steroid reductase has been found to be present in human submandibular glands and may in men convert testosterone to androstenedione and dihydrotestosterone (DHT). In women, on the contrary with naturally lower testosterone levels, androstenedione may be converted to testosterone, a metabolic activity which most probably is reversed by testosterone administration. This metabolism seems not to be a problem in measurement of Sal-T, due to the rapid passage through the gland (Rey et al., 1988; Blom et al., 1993). It has been stated that Sal-T cannot be considered a surrogate marker of serum-free testosterone in women, explained by testosterone binding to saliva proteins. This effect will significantly affect the much lower Sal-T in normal women and children, but is negligible in healthy adult men (Fiers et al., 2014). As testosterone replacement therapy, in both HG men and GD FtM, is intended to achieve serum testosterone levels within physiological ranges of normal men, this effect of protein binding would not be apparent in Sal-T measurement. However, when immunoassays are used to determine Sal-T, the total testosterone fraction in saliva is measured and is therefore independent on protein binding. The correlation demonstrated in this study, between Sal-T and Free-T, was in agreement with the findings in a recent study on Sal-T and S-T in normal men and women determined by LC- MS/MS and calculated Free-T using the Vermeulen equation (Keevil et al., 2014). We show that Sal-T correlates just as well to S-T as to Free-T and even better at the higher concentration levels after the injection. Although Sal-T is not directly comparable to serum-free testosterone, it is a direct measurement, while calculated Free-T and FAI is based on the validity of a number of parameters. The higher Sal-T concentration compared to Free-T is systematic and can be taken into account when interpreting results. Sal-T might be a good surrogate parameter of the unbound testosterone fraction in plasma which can be useful for clinical use in men and especially whenever testosterone supplementation should be undertaken. The secondary aim of the study was to investigate Sal-T levels reached during conventional replacement therapy. This is a valuable knowledge also in forensic toxicology, obtaining ethical approval for studies of in vivo, non-therapeutic doses, which are often used in abuse situations of testosterone, would be difficult. An important finding in this study was the high and divergent testosterone concentrations measured 7 14 days after the TU injection in the majority of the treated subjects, with S-T and Sal-T levels clearly above the physiologically levels found in control subjects. Thus, doses and dose intervals might have to be adjusted in some individuals to prevent the risks of undesired 2017 American Society of Andrology and European Academy of Andrology Andrology, 2018, 6,

7 Y. Lood et al. side effects. Sal-T can also provide a useful alternative to the collection of urine in forensic cases and anti-doping testing of testosterone, because it easily can be collected without risk for manipulation by the suspected offender. Moreover, Sal-T mirrors the serum testosterone better than U-T and is also independent on genetic inter-individuals variations in urinary excretion (Schulze et al., 2008) and might therefore be a useful screening tool in forensic investigations in future. The benefits of saliva sampling include the non-invasive, stress-free and easy collection procedure with minimal training for sample takers. Saliva collection could be self-administered by patients with no need of attendance at a clinic and to less cost. Although saliva could be preferable to serum and urine and of value in several fields, problems concerning blood contamination and loss of adsorption have been reported (Granger et al., 2004; Gr oschl, 2008; Wood, 2009). A study by Kivlighan et al., 2004 confirmed the presence of haemoglobin and transferrin in saliva samples 30 min after tooth-brushing, and Sal-T was increased during that time. Thus, saliva is routinely checked for haemoglobin at our laboratory, and no further analysis is performed at haemoglobin >2 g/l. This was not an obvious problem in our study, as none of the saliva samples were found to contain blood. Some limitations in this study need to be addressed. The relatively small number of GD FtM subjects available for recruitment at the clinic might affect the ability of generalization. Lack of genetic information (Schulze et al., 2008; Bang et al., 2013) and the fact that GD FtM differed in some anthropometric characteristics (i.e. body weight) compared to controls may have affected the comparison results. A female control group would have been appropriate for evaluation of the initial baseline values for GD FtM. As our study had an observational design, further dose finding studies in GD FtM and HG men are needed. CONCLUSION This study suggests that Sal-T has the potential to be a useful biomarker in the assessment of testosterone replacement therapy, offering an easy, non-invasive sampling and might also add information to urine on the circulating testosterone levels in non-medical use of testosterone in forensic cases. Additional studies are needed before considering introduction of Sal-T in clinical and forensic use, including evaluation of reference ranges, intra-individual variations of oral fluid testosterone influenced by gender, age, body size, genotype and hormonal disorders. ACKNOWLEDGEMENTS This study was supported by the Medical Research Council of Southeast Sweden and the Link oping University, Sweden. DISCLOSURES The authors declare no conflict of interests. AUTHORS CONTRIBUTION JW and BE designed the research study. YL, EA and CW collected the data, YL wrote the draft manuscript and EA, JA, BE and JW revised the manuscript for intellectual content. All authors approved the final manuscript. REFERENCES Anizan S & Huestis MA. (2014) The potential role of oral fluid in antidoping testing. Clin Chem 60, Arregger AL, Contreras LN, Tumilasci OR, Aquilano DR & Cardoso EM. (2007) Salivary testosterone: a reliable approach to the diagnosis of male hypogonadism. Clin Endocrinol 67, Bang AK, Jørgensen N, Rajpert-De Meyts E & Juul A. (2013) UGT2B17 genotype and the pharmacokinetic serum profile of testosterone during substitution therapy with testosterone undecanoate. A retrospective experience from 207 men with hypogonadism. Front Endocrinol 4, 1 8. Behre HM, Abshagen K, Oettel M, H ubler D & Nieschlag E. (1999) Intramuscular injection of testosterone undecanoate for the treatment of male hypogonadism: phase I studies. Eur J Endocrinol 140, Blom T, Ojanotko-Harri A, Laine M & Huhtaniemi I. (1993) Metabolism of progesterone and testosterone in human parotid and submandibular salivary glands in vitro. J Steroid Biochem Mol Biol 44, Bui HN, Schagen SE, Klink DT, Delemarre-van de Waal HA, Blankenstein MA & Heijboer AC. (2013) Salivary testosterone in female-to-male transgender adolescents during treatment with intra-muscular injectable testosterone esters. Steroids 78, Coviello AD, Lakshman K, Mazer NA & Bhasin S. (2006) Differences in the apparent metabolic clearence rate of testosterone in young and older men with gonadotropin suppression receiving graded doses of testosterone. J Clin Endocrinol Metab 91, Dunn JF, Nisula BC & Rodbard D. (1981) Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J Clin Endocrinol Metabol 53, Fiers T, Delanghe J, T 0 Sjoen G, Van Caenegem E, Wierckx K & Kaufman J-M. (2014) A critical evaluation of salivary testosterone as a method for the assessment of serum testosterone. Steroids 86, 5 9. Granger DA, Shirtcliff EA, Booth A, Kivlighan KT & Schwartz EB. (2004) The trouble with salivary testosterone. Psychoneuroendocrinology 29, Gr oschl M. (2008) Current status of salivary hormone analysis. Clin Chem 54, Hembree WC, Cohen-Kettenis P, Delemarre-van de Waal HA, Gooren LJ, Meyer III WJ, Spack NP, Tangpricha V & Montori VM. (2009) Endocrine treatment of transsexual persons: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab 94, Kapoor P, Luttrell BM & Williams D. (1993) The free androgen index is not valid for adult males. Steroid Biochem Mol Biol 45, Keevil BG, MacDonald P, MacDowall W, Lee DM, Wu FC & Team NATSAL. (2014) Salivary testosterone measurement by liquid chromatography tandem mass spectrometry in adult males and females. Ann Clin Biochem 51, Keevil BG, Fiers T, Kaufman JM, Macdowall W, Clifton S, Lee D & Wu F. (2016) Sex hormone-binding globulin has no effect on salivary testosterone. Ann Clin Biochem 53, Kivlighan KT, Granger DA, Schwartz EB, Nelson V, Curran M & Shirtcliff EA. (2004) Quantifying blood leakage into the oral mucosa and its effects on the measurement of cortisol, dehydroepiandrosterone and testosterone in saliva. Horm Beh 46, Konieczna L, Plenis A, Olȩdzka I, Kowalski P & Bazczek T. (2011) Optimization of LC method for the determination of testosterone and epitestosterone in urine samples in view of biomedical studies and anti-doping research studies. Talanta 83, Kotsopoulos J, Shafrir AL, Rice M, Hankinson SE, Eliassen AH, Tworoger SS & Narod SA. (2015) The relationship between bilateral oophorectomy and plasma hormone levels in postmenopausal women. Horm Cancer 6, Andrology, 2018, 6, American Society of Andrology and European Academy of Andrology

8 SALIVARY LEVELS OF TESTOSTERONE Krieg M & Arning C. (1978) Quantitative determination of the binding capacity of the sex hormone-binding globulin, using agar gel electrophoresis. J Clin Chem Clin Biochem 16, Lood Y, Eklund A, Garle M & Ahlner J. (2012) Anabolic androgenic steroids in police cases in Sweden Forensic Sci Int 219, Masse R, Ayotte C & Dugal R. (1989) Studies on anabolic steroids. I. Integrated methodological approach to the gas chromatographic-mass spectrometric analysis of anabolic steroid metabolites in urine. J Chromatogr 489, Matsumoto AM & Bremner WJ. (2005) Serum testosterone assays accuracy matters. J Clin Endocrinol Metab 89, Meriggiola MC & Berra M. (2013) Safety of hormonal treatment in transgenders. Curr Opin Endocrinol Diabetes Obes 20, Mongrawe-Chaffin ML, Anderson CAM, Allison MA, Ouyang P, Szklo M, Vaidya D, Woodward M & Hill Golden S. (2015) Association between sex hormones and adiposity: qualitative differences in women and men in the multi-ethnic study of atherosclerosis. J Clin Endocrinol Metab 100, E596 E600. Mueller A, Kiesewetter F, Binder H, Beckmann MW & Dittrich R. (2007) Long-term administration of testosterone undecanoate every 3 months for testosterone supplementation in female-to-male transsexuals. J Clin Endocrinol Metab 92, Nieschlag E. (2015) Current topics in testosterone replacement of hypogonadal men. Best Pract Res Clin Endocrinol Metab 29, Pelusi C, Constantino A, Martelli V, Lambertini M, Bazzocchi A, Ponti F, Battista G, Venturoli S & Meriggiola MC. (2014) Effects of three different testosterone formulations in female-to-male transsexual persons. J Sex Med 11, Rech CMZ, Clapauch R, de Souza MD & Bouskela E. (2016) Low testosterone levels are associated with endothelial dysfunction in oophorectomized early postmenopausal women. Eur J Endocrinol 174, Rey F, Chiodoni G, Gomez F & Felber JP. (1988) Interpretation of the discrepancy observed between plasma free and salivary testosterone levels in man. Steroids 52, Rilling JK, Worthman CM, Campbell BC, Stallings JF & Mbizva M. (1996) Ratios of plasma and salivary testosterone throughout puberty: production versus bioavailability. Steroids 61, de Ronde W, van der Schouw YT, Pols HAP, Gooren LJG, Muller M, Grobbee DE & de Jong FH. (2006) Calculation of bioavailable and free testosterone in men: a comparison of 5 published algorithms. Clin Chem 52, Rosner W, Auchus RJ, Azziz R, Sluss PM & Raff H. (2007) Position statement: utility, limitations, and pitfalls in measuring testosterone: an endocrine society position statement. J Clin Endocrinol Metab 92, Schubert M, Minnemann T, H ubler D, Rouskova D, Christoph A, Oettel M, Ernst M, Mellinger U, Krone W & Jockenh ovel F. (2004) Intramuscular testosterone undecanoate: pharmacokinetic aspects of a novel testosterone formulation during long-term treatment of men with hypogonadism. J Clin Endocrinol Metab 89, Schulze JJ, Lundmark J, Garle M, Skilving I, Ekstr om L & Rane A. (2008) Doping test results dependent on genotype of uridine diphosphoglucuronosyl transferase 2B17, the major enzyme for testosterone glucuronidation. J Clin Endocrinol Metab 93, Shea JL, Wong PY & Chen Y. (2014) Free testosterone: Clinical utility and important analytical aspects of measurement. Adv Clin Chem 63, Sj oqvist F, Garle M & Rane A. (2008) Use of doping agents, particularly anabolic steroids, in sports and society. Lancet 371, S odergard R, B ackstr om T, Shanbhag V & Carstensen H. (1982) Calculation of free and bound fractions of testosterone and estradiol 17 beta to human plasma proteins at body temperature. J Steroid Biochem 16, Thieme D, Rautenberg C, Grosse J & Schoenfelder M. (2013) Significant increase of salivary testosterone levels after single therapeutic transdermal administration of testosterone: suitability as a potential screening parameter in doping control. Drug Test Anal 5, Vermeulen A, Verdonck L & Kaufman JM. (1999) A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 84, Wood P. (2009) Salivary steroid assays research or routine? Ann Clin Biochem 46, American Society of Andrology and European Academy of Andrology Andrology, 2018, 6,