Evaluating the four most important salivary sex steroids during male puberty: testosterone best characterizes pubertal development

Statement of ethics

Prior to data acquisition and analyses, the protocol was approved by the Ethics Committee at the Albert-Ludwigs-University of Freiburg (registration number: 184/09) and all participants or their parents signed an informed consent.

Subjects

Eligible patients were investigated at the Department of Pediatrics and Adolescence Medicine of the University Hospital Freiburg, Germany, and the Practice for Pediatrics and Adolescence Medicine in Butzbach, Germany. Inclusion of 165 peripubertal boys was performed after a physical examination carried out by experienced pediatric endocrinologists. Recruiting and screening of study participants were realized through our study investigators in the two participating study centers and the inclusion predominantly comprised healthy boys and adolescents ≥6 and ≤20 years of age. Moreover, male patients with diabetes type 1, familial short stature and familial hypercholesterolemia were also enrolled in the study. The exclusion criteria included diseases which might influence the normal development of male puberty or which required the administration of hormones (e.g. cortisone, TE, GnRH analogs) and medication affecting puberty (e.g. aromatase inhibitors).

The mean age of the boys was 12.7±2.8 years, mean height 150.5±17.8 cm, mean weight 46.2±18.0 kg and mean body mass index (BMI) 19.6±4.2 kg/m².

Clinical and laboratory examinations

BMI was defined as weight/height² (kg/m²). The development of PH [9] was assessed according to Tanner stages [4] and TV was measured using the Prader orchidometer [2].

Hormone analyses

Statistical analysis

Data were analyzed using SPSS 24 for Windows (IBM Corp., Armonk, NY, USA) and the statistical software R (R Foundation for Statistical Computing, Vienna, Austria) [10]. p-Values<0.05 were considered statistically significant. The Mann-Whitney U test was performed to evaluate the difference between groups, separately for serum and saliva steroid values and for PH stages and TV. Bland-Altman analysis was utilized for the evaluation of differences in validity between serum and saliva. Spearman rank correlation coefficients were calculated to assess the strength of association between steroid hormone levels in serum and saliva and between salivary hormone levels and pubertal development (PH stages, TV). Binomial logistic regression analyses were conducted with categorical variables to evaluate correlations of pubertal status with salivary steroids or TV and log estimates were transformed by exponentiation. Multiple linear regression analysis was applied for metric variables (TV and steroids).

Salivary ELISA

Table 1 compares self-assessed results with the corresponding values in the user manual regarding inter- and intra-assay CVs of used ELISAs for the determination of salivary sex steroids.

Development phases in peripubertal boys

Table 2 provides information about pubertal development and corresponding salivary and serum sex steroid levels covering the period between prepuberty and postpuberty. During this period, steroid levels in serum and saliva increased with age. The highest increases were observed between prepuberty and puberty, which applies in particular to TE. All total serum steroid concentrations were higher than the corresponding values in saliva.

Correlation of salivary sex steroid levels with PH stages and TV

Table 3 compares the individual steroids regarding the strength of the correlation between saliva levels and pubertal development (PH stages, TV). All correlations were statistically significant, but TE showed the strongest association.

Correlation between pubertal development stages and salivary sex steroids

A binomial logistic regression analysis revealed significant correlations between salivary steroids and pubertal maturation (PH stages, TV) during the development phases of prepuberty-puberty and puberty-postpuberty only for TE (Model 2) (Table 4). Inclusion of further salivary steroids did not improve analysis results (Model 1).

Correlation of male pubertal sex steroids between serum and saliva

Correlations between serum and saliva steroid hormones were all significant (<0.001), but to varying correlation coefficients with the highest one for TE and the lowest for 17-OHP (Figure 1).

Figure 1:

Relationship of male pubertal sex steroids between serum and saliva levels.

(A) dehydroepiandrosterone-sulfate (DHEAS)/dehydroepiandrosterone (DHEA), (B) androstenedione, (C) testosterone and (D) 17-hydroxyprogesterone (17-OHP) in normal pubertal boys.

Main findings

In addition to previous investigations, we explored the course of the four most crucial male salivary sex steroids throughout the peripubertal period to assess relationships between various sex hormones and pubertal development stages. A significant association was found between salivary concentrations of TE, DHEA, AE and 17-OHP and pubertal development in normal boys. Using binomial logistic regression analysis, however, correlations between salivary steroids and pubertal development only showed a significant result for TE. The inclusion of further salivary steroids did not improve the predictability of pubertal development.

Sex steroids and pubertal development

In general, steroid measurements are very useful not only for understanding the timing of normal pubertal maturation with its external manifestations, but also for the identification of pubertal disorders. However, some specifics must be taken into consideration in studies evaluating relationships between steroids and pubertal maturation. In particular, the diurnal rhythm of steroids, an individual variability in puberty hormone change and the simultaneous impact of various hormones should be respected [1], [9].

Sex steroids and PH stages

The growth of PH as part of the adrenarche is triggered by the activation of the hypothalamic-pituitary-adrenal axis with the consequent production of various steroids such as DHEA and DHEAS. Both adrenal steroids can be detected in the blood and are useful markers of the onset of adrenarche [1], [9], [11]. Large proportions of DHEA are sulfated in the zona reticularis of the adrenal gland by the enzyme SULT2A1 (DHEA-sulfotransferase) to form DHEAS, which is primarily detectable in the blood compared to DHEA [12]. In addition, the half-life of DHEA is indicated to be 15–30 min and that of DHEAS to be 7–10 h [13]. In contrast to the situation in the blood, salivary concentrations of DHEA are considerably higher than the concentrations of DHEAS. This is mainly due to the fact that lipoinsoluble, conjugated steroids in saliva (e.g. DHEAS) represent only 1% of their free unbound serum concentrations [7], [14]. In peripheral tissues, DHEA and DHEAS function as precursors for the androgens TE and dihydrotestosterone. PH growth in boys is also being stimulated by peripheral conversion of DHEAS to physiologically active TE [9]. A marked and progressive increase in circulating DHEAS can be found in the blood during adrenarche and was described as about 5 times higher at the end of puberty compared to the level before the onset [9]. Serum DHEAS levels of our healthy subjects increased about 2.5 times and salivary DHEA levels reached a four-fold increase comparing prepubertal values with those after the end of puberty. In addition, salivary TE and DHEA showed the best correlation with PH stages and TV. A good correlation between physical measures of puberty (PH stage, genital stage) and salivary steroids (TE, DHEA) was also described in 82 boys aged 9–14 years [11]. In 828 American boys of different race/ethnicity, blood TE, LH and inhibin B concentrations were positively associated with age and pubertal progression and significantly increased across PH stages [15]. Furthermore, we have been able to ascertain that there was a progressive increase of TE, DHEAS/DHEA, AE and 17-OHP in serum and saliva with age in our boys and the hormone concentrations significantly increased across pubertal development considering PH stages and TV.

Testis volume and pubertal stages

FSH increases testicular size due to the growth of seminiferous tubules. LH stimulates the testes to produce TE by testicular Leydig cells, thus additionally promoting testicular enlargement [1]. Therefore, the volume of the testis is directly related to testicular function [16] and precise measurement of TV is of importance for the assessment of pubertal development and testicular or pubertal disorders. Ultrasound scan is the most exact method for measuring TV and is currently regarded as standard [17], but also shows an accurate correlation with measurements by the Prader orchidometer (R²=0.956) [18]. In the literature, measurements of TV are often related to the age of children and only occasionally to pubertal stages [17], [18]. With consideration of Tanner stages for PH in a Dutch study [17], the distribution of sonographically measured TV within PH stages 1–5 was reasonably comparable with our orchidometric values. However, the Dutch orchidometric measurements were higher than ours, particularly with regard to PH stages 3–5.

Relationship of salivary steroid levels with serum total or free levels

Saliva contains free unbound steroids which are lipid soluble and can be transferred from blood into saliva by passive diffusion along a concentration gradient. Therefore, salivary steroid levels reflect the serum free steroid concentration of the total circulating level. Studies comparing salivary and serum total or free steroid levels found significant correlations with coefficients (r) higher than 0.8 [7]. The best correlation between salivary and total serum concentration in our participants was observed for TE followed by DHEA, AE and 17-OHP. When comparing results of various studies, possible differences should be considered that may be based on different sampling and determination methods of steroids in saliva and blood [8]. Furthermore, differences that must be noted are related to age and sex of study participants or to the metabolism of steroids and proteins [6], [7]. In human saliva, a number of different proteins have been discovered reflecting those found in plasma and thus also steroid binding proteins, for example, sex hormone binding globulin, albumin, corticosteroid binding protein or proline rich protein which may influence salivary steroid concentrations [19], [20], [21].

Testosterone as a monitor of male pubertal development

During pubertal development in males, TE will increasingly be provided by maturing testes and is associated with pubertal progression. Regarding our normal boys, TE in serum and saliva showed the highest pubertal increase of all hormones analyzed and increased progressively during the transition from prepuberty to postpuberty. In addition, TE revealed the strongest relation of serum and saliva levels with PH stages and TV. The continuous increase of serum TE throughout puberty in healthy boys has been described in the literature [22], [23], [24]. Furthermore, our data in terms of the correlation between serum TE levels and pubertal maturation are in accordance with the findings of other authors [25], [26]. But also data concerning salivary TE are of great interest for assessing pubertal maturation. A steady annual increase of TE throughout puberty was described for salivary TE in 1047 boys aged 9–17 years, while the increase of TE values was particularly pronounced between 12 and 14 years of age [27]. This corresponds to our TE levels both in saliva and serum and thus conforms to the transition from prepuberty to puberty.

Although immunoassay techniques have proved to be suitable for the use in pediatric clinical routine, certain restrictions should be considered regarding cross reactivity, validation or the assessment of low steroid concentrations [8]. Liquid chromatography in combination with tandem mass spectrometry (LC-MS/MS) is also used for pediatric steroid analysis, improves sensitivity and specificity and also offers the opportunity for the simultaneous measurement of several analytes [28]. Furthermore, it was possible to demonstrate that salivary TE can be accurately measured by LC-MS/MS in female and male adults [29] and reference values for salivary TE analyzed using isotope-dilution liquid-chromatography tandem mass spectrometry (ID-LC-MS/MS) have recently been published for adolescent boys and girls during puberty [30]. However, comparing the results of seven published LC-MS/MS methods when measuring TE, AE and DHEA in serum, it was noted that assays may show differences in standardization or high variation [31]. A comparison between the use of enzyme immunoassays and LC-MS/MS for the measurement of salivary TE showed only moderate correspondence of the results of the two methods [32]. The previously published data show the difficulty in comparing the TE results produced with different immunoassays and LC-MS/MS methods and improved ways of assessing TE and other steroids should be developed. For example, endocrinological research might benefit from the combination of these methods with gas chromatography-MS [33].

Taken together, due to the demonstrated significant correlation between saliva and serum of all four sex steroids, it should be possible that salivary steroid levels can be used as a noninvasive measurement tool to assess and compare associations with the pubertal development from pre- to postpuberty. The steroid levels in serum and saliva progressively increased with age. The highest increases were observed in particular to TE. Therefore, salivary TE permits a good characterization of pubertal maturation stages. The consideration of further salivary sex steroids did not improve diagnostic accuracy.