Differences between diurnal patterns of salivary cortisol and dehydroepiandrosterone in healthy female adolescents

Abstract

The adrenal hormones cortisol and dehydroepiandrosterone (DHEA) share a common secretagogue: adrenocorticotropic hormone; however, secretion of these hormones can be dissociated suggesting subtle individual regulation at the level of the adrenal gland. We examined differences in the diurnal patterns of cortisol and DHEA secretion in healthy adolescent girls, with the aim of informing the possibility of exploiting these differences to aid interpretation of data from clinical populations in which these patterns can become dysregulated. Fifty-six healthy females aged 10–18 years provided saliva samples at 0 and 30 min (morning samples) and 12 h post-awakening on 2 consecutive weekdays. For morning salivary cortisol in relation to morning DHEA concentrations, correlational analysis revealed only a trend (p = 0.054). Similarly, the association between evening cortisol and DHEA was characterised as a trend (p = 0.084). Mean morning DHEA concentrations showed more day-to-day consistency than equivalent cortisol samples (r = 0.829 for DHEA and 0.468 for cortisol; z = 3.487, p < 0.0005). Unlike the cortisol pattern, characterised by a marked awakening response (cortisol awakening response, CAR), a significant rise in DHEA concentration post-awakening was not evident. Finally, there was a strong association between morning and evening concentrations of DHEA, not found for cortisol. The study shows differences in cortisol and DHEA secretion in the post-awakening period and informs work that seeks to examine correlates of dysregulated hypothalamic–pituitary–adrenal axis function. Parallel examination of both hormones enables enhanced interpretation of aberrant patterns of the CAR, i.e. an exploration of whether dysregulation affects both hormones (reflecting overall steroidogenic capacity) or cortisol alone (CAR-specific mechanisms).

Keywords::

• ACTH
• adolescence
• CAR
• cortisol
• DHEA
• saliva

Introduction

The adrenal steroids cortisol and dehydroepiandrosterone (DHEA) are released in response to the common secretagogue adrenocorticotropic hormone (ACTH) and both steroids exhibit a diurnal decline from a morning acrophase (Granger et al. 1999; Hucklebridge et al. 2005). However, it has been suggested that measurement of salivary DHEA provides a more stable index of adrenal steroidogenic capacity than salivary cortisol: intra-individual between-day and within-day concentrations are more consistent than for salivary cortisol in both adults and adolescents (Hucklebridge et al. 2005; Matchock et al. 2007). Although overall secretion of cortisol and DHEA tends to be correlated, secretion of each hormone has specific regulatory inputs at the level of the adrenal that enables dissociation (Buijs et al. 2003; Hucklebridge et al. 2005; Ulrich-Lai et al. 2006). It was the purpose of this study to contribute to knowledge about the differences in the diurnal pattern of these two hormones in healthy participants. The secondary aim was to highlight the usefulness of studying both hormones in parallel (especially in the immediate post-awakening period), to aid interpretation of clinical investigations in which hypothalamic–pituitary–adrenal (HPA) axis dysregulation is suspected.

Although there is a large literature linking diurnal patterns of cortisol with psychosocial characteristics and health during adolescence (e.g. Van den Bergh et al. 2008; Oskis et al. 2011) there is scant literature on the equivalent and parallel diurnal pattern of DHEA. Indeed, for DHEA, there is more research on DHEA in stress response investigation in studies of adolescents and young adults (Shirtcliff et al. 2007; Wemm et al. 2010). It has been suggested that an understanding of both of these adrenal hormones may be important for elucidating biological correlates of psychopathology (Stroud et al. 2009). Analysis of diurnal patterns of both cortisol and DHEA, studied in parallel, would illuminate underlying regulatory mechanisms implicated in associations with various psychopathologies and health.

The cortisol awakening response (CAR) is one such example. Whilst the CAR has been widely associated with a range of psychopathologies (see Fries et al. 2009), it remains unclear whether the aberrant underlying mechanism reflects problems with general steroidogenic capacity of the adrenal (which would affect both cortisol and DHEA) or whether more specific mechanisms affecting cortisol and not DHEA secretion are implicated (Clow et al. 2010). A concurrent analysis of both hormones would enable a more insightful interpretation of data. To our knowledge, there is only one study that has documented in detail the pattern of both hormones in healthy participants across the day, relative to awakening time (Hucklebridge et al. 2005). In that study, there were moderate correlations between the concentrations of hormone secretion, which were interpreted as reflecting overall steroidogenic capacity of the adrenal. However, post-awakening cortisol and DHEA could be distinguished, indicating cortisol-specific inputs at the level of the adrenal in the generation of the CAR. If confirmed, these findings could provide a rationale for a methodology to probe more deeply into the mechanisms underlying aberrant patterns of the CAR, i.e. an exploration of whether dysregulation affects both hormones (reflecting overall steroidogenic capacity) or cortisol alone (CAR-specific mechanisms).

The aim of this investigation was to investigate differences between the diurnal patterns of salivary DHEA and cortisol, paying particular attention to the immediate post-awakening period. We chose to study healthy (i.e. medication and psychopathology-free) female, post-menarche adolescents, as the diurnal pattern of cortisol and DHEA is fully established by this developmental phase (Matchock et al. 2007) and most existing research into the relationship between cortisol and DHEA is undertaken in adolescence; a time of vulnerability to future psychopathology, especially for females (Adam et al. 2008). We hypothesised that overall concentrations of cortisol and DHEA would be moderately associated but that there would be differences in patterns of secretion, especially after awakening and in terms of within-day correlations of morning and evening concentrations.

Materials and methods

Participants

Fifty-six healthy adolescent girls, mean ( ± SD) age 14.27 ( ± 2.49) years, ranging from 10 to 18 years, participated in the study. All were post-menarche, non-smokers, without acute or chronic illness or taking prescribed medication (including the oral contraceptive). The mean body mass index (BMI) was 21.06 ( ± 3.67). Most participants lived and attended school in the Greater London area and all were British born. All provided informed written consent (individually or by parent/guardian if the participant was under 16 years old). Ethical approval for the study was obtained from the University of Westminster Ethics Committee.

Protocol and procedure

Participants were mainly recruited by approaching local schools. Recruitment and data collection was conducted by the same researcher (AO: white female aged 23, 24 years). Participants received a study pack, which included a copy of the participant information sheet and consent form, full standardised written instructions and saliva sampling materials. Participants collected saliva samples on awakening, 30 min and 12 h post-awakening on 2 consecutive weekdays. The protocol is based on the paradigm established by Edwards et al. (2001) to capture the morning acrophase and the lowered evening concentrations, relevant to both steroids. Profiles compromising similar three sampling points have been reported to be reliable in previous studies examining cortisol in adolescents (O'Connor et al. 2005).

Saliva was collected by passive drool into an eppendorf tube. For 30 min prior to the collection of each sample, participants were requested to take nil by mouth other than water, avoid vigorous exercise and brushing teeth. Participants froze samples on the same day of collection and insulated packs were used to transfer samples to the laboratory where they were stored at − 20°C until assay. On each study day, participants recorded their awakening time, method of waking up (whether spontaneous or by alarm clock) and the exact times of collection of saliva samples. See Oskis et al. (2009) for further details regarding the protocol and procedure.

Salivary assays

Samples were thawed and centrifuged at 1500g (3000 rpm) for 15 min. Cortisol concentration was determined by the Expanded Range High Sensitivity Enzyme Linked Immuno-Sorbent Assay developed by Salimetrics LLC (State College, PA, USA). Similarly, the Salimetrics Salivary DHEA Enzyme Immunoassay was used to measure DHEA concentrations. Given that freeze–thaw cycles for DHEA are recommended to be kept to a minimum (Schwartz et al. 2005), the DHEA assay was carried out first.

Standard range for cortisol assay: 0.0828–82.80 nmol/l. Correlation of assay with serum cortisol: r = 0.91, p < 0.0001, n = 49 samples. For the DHEA assay, the standard range was 0.03–3.47 nmol/l. Correlation of assay with serum DHEA: r = 0.857, p < 0.0001, n = 39 samples. For both cortisol and DHEA, intra and inter-assay variations were below 5 and 10%, respectively.

Statistical analysis

Raw cortisol and DHEA concentrations at each sampling point were moderately positively skewed. A square-root transformation was carried out on these data, which reduced skewness statistics. Mean concentrations represented in figures are original values.

Cortisol and DHEA data comprised three samples taken at 0, 30 min and 12 h post-awakening on 2 consecutive days. Separate within-subjects analyses of variances (ANOVAs) were used for each hormone, with factors “sampling day” (first and second) and “sampling time” (0, 30 min and 12 h post-awakening). To determine the effects of other variables measured in this study, developmental markers (age and BMI) and situational variables (awakening time and awakening method) were entered as separate covariates in these ANOVAs.

Composite measures were also used to investigate similarities and differences between cortisol and DHEA. Pearson's tests of correlation were undertaken for this purpose. In these analyses, the post-awakening dynamic was computed as the mean 30 min concentration minus the mean concentration at 0 min post-awakening. We also computed mean concentrations of cortisol and DHEA from 0 to 30 min post-awakening and refer to this as mean morning cortisol and DHEA.

Results

Patterns of cortisol and DHEA secretory activity

Separate ANOVAs revealed the characteristic diurnal patterns of each hormone. The main effect of sample was found for both DHEA (F_(2,110) = 84.012, p < 0.0005, partial η² = 0.604; Figure 1, left) and cortisol (F_(1.8,99.1) = 197.93, p < 0.0005, partial η² = 0.783; Figure 1, right). In both cases, salivary concentrations were the highest at 0–30 min post-awakening and were the lowest by the evening sampling point. The mean ( ± SD) evening concentration of DHEA was 0.55 ( ± 0.46) nmol/l, significantly lower than morning concentrations of 1.47 nmol/l ( ± 0.85) (t = 11.287, p>0.0005). Participants displayed a robust CAR; the mean ( ± SD) change in cortisol concentration from the awakening value to 30 min across the 2 days was 6.68 nmol/l ( ± 6.04), and this declined to a mean concentration of 1.91 nmol/l ( ± 1.94) at 12 h post-awakening. The main effect of sampling day was not present for either DHEA (F_(1,55) = 1.877, p = 0.176, partial η² = 0.033) or cortisol (F_(1,55) = 0.067, p = 0.797, partial η² = 0.001), illustrating that concentrations for both steroids were consistent across days. No significant interactions were present between the factors “sampling day” and “time” in either ANOVA (DHEA: F_(1.8,96.5) = 0.985, p = 0.368, partial η² = 0.018; cortisol: F_(2,110) = 0.091, p = 0.913, partial η² = 0.002).

Figure 1.  Salivary DHEA (Left) and cortisol (Right) concentrations at 0, 30 min and 12 h post-awakening in nanomoles per litre (nmol/l) (n = 56). Values are mean ± SEM. Separate mixed ANOVAs revealed the main effect of sample for both DHEA (p < 0.0005; Left) and cortisol (p < 0.0005; Right).

Cortisol and DHEA secretory activity: Comparisons and relationships

The group mean morning salivary DHEA concentrations were highly positively correlated across the 2 sampling days (r = 0.829, p < 0.0005), as were 12 h concentrations (r = 0.661, p < 0.0005). Cortisol secretion was also correlated over days, for the CAR (r = 0.653, p < 0.0005), mean salivary concentrations of cortisol from 0 to 30 min post-awakening (r = 0.468, p < 0.0005) and for evening concentrations (r = 0.771, p < 0.0005). Across-day correlations for morning DHEA (i.e. mean concentrations 0–30 min post-awakening) were significantly greater than for the equivalent measure of cortisol, as evidenced by the significant difference between these two correlations (z = 3.487, p < 0.0005). However, there was no statistically significant difference in across-day correlations of 12 h post-awakening concentrations (z = 1.175, p = 0.240).

There was a trend for mean morning concentrations of DHEA (0+30 min post-awakening) to be correlated with mean concentrations of cortisol during the same time frame (r = 0.259, p = 0.054) and with the increase in cortisol concentrations in the first 30 min post-awakening (r = 0.255, p = 0.057). Mean morning salivary DHEA concentration was not associated with the cortisol concentration in the first sample (r = 0.067, p = 0.623). Similarly, there was only a trend towards a positive correlation between DHEA and cortisol in the evening saliva sample collected 12 h after awakening (r = 0.233, p = 0.084).

There was no dynamic increase in salivary DHEA concentration following awakening. In the period from 0 to 30 min post-awakening, DHEA concentration increased by 0.006 nmol/l ( ± 0.68) compared to a mean rise of 6.68 nmol/l ( ± 6.04) for cortisol. Secondly, a strong correlation between morning (0+30 min post-awakening) and evening concentrations was present for DHEA (r = 0.559, p < 0.0005) but not for cortisol (r = 0.036, p = 0.794). The difference between these two correlations was significant (z = 3.06, p = 0.001).

All results remained as above when including developmental markers (age and BMI) and situational variables (awakening time and awakening method) within the analyses.

Discussion

This study contributes to the limited literature examining the relationships between salivary cortisol and DHEA concentrations and highlights the usefulness of studying both hormones in parallel. Within this group of healthy female adolescents, cortisol and DHEA secretory activities showed marked differences in terms of patterns and relationships. Despite their common secretagogue (ACTH), there was only a trend for correlation between cortisol and DHEA concentrations, and post-awakening DHEA secretion was found to have more day-to-day consistency than cortisol. In line with previous findings from both adults and adolescents (Hucklebridge et al. 2005; Matchock et al. 2007), DHEA activity in the period from 0 to 30 min post-awakening was related to DHEA concentrations in the evening, but for cortisol this was not the case. Finally, DHEA could be distinguished from cortisol by the lack of a DHEA CAR, in contrast to the robust dynamic exhibited by cortisol in the period up to 30 min post-awakening.

The aim of this study was to highlight dissociation in the patterns of salivary cortisol and DHEA secretion in healthy young females in order that these differences could be exploited to facilitate interpretation of frequently observed dysregulated patterns of HPA axis activity from various clinical populations. Taken together, the findings clearly distinguished healthy post-awakening cortisol and DHEA secretion and are consistent with additional cortisol-specific inputs in the generation of the CAR, at the level of the adrenal gland. Recent work suggests that the CAR is more than a simple measure of HPA axis function; the dynamic of the morning CAR is enhanced by light-sensitive, extra-pituitary mechanisms, involving the suprachiasmatic nuclei via a direct neural projection to the adrenal zona fasciculata, the layer of adrenal cortex from which cortisol is secreted (Scheer and Buijs 1999; Buijs et al. 2003; Clow et al. 2010). Data presented here are consistent with a lack of extra-pituitary input to the regulation of DHEA secretion. This is entirely plausible as there is no direct innervation to the adrenal zona reticularis, the layer of the adrenal cortex from which DHEA is secreted, to enable this to occur (Charlton et al. 1992). Hence, DHEA secretion in the morning may provide a “clean” measure of ACTH–DHEA status (i.e. status of the HPA axis). In contrast, the CAR reflects a combination of both HPA axis activation and “fine-tuning” by extra-pituitary neural mechanisms (Buijs et al. 2003). It would be most interesting to test the hypothesis that peri-awakening DHEA secretion more closely parallels ACTH concentrations than cortisol by examining patterns of all three hormones; it is possible that other hormone-specific differences in regulation account for these findings.

A limitation of this study is that Tanner stages of puberty were not measured, which would be a recommendation for future studies. Another possible limitation is that we did not control for menstrual phase, although this does not seem to have an effect on the CAR (Kirschbaum et al. 1999). However, as the focus of analysis was within- rather than between-participants, it seems unlikely that these factors could substantially have affected the thrust of our findings.

The strategy adopted by this study of healthy participants contributes to fundamental knowledge about the differences in the diurnal patterns of cortisol and DHEA in healthy participants and highlights the usefulness of studying both hormones in parallel. Knowledge of differences in the patterns of secretion of these hormones in healthy participants, particularly in the immediate post-awakening period, could be exploited to probe potential underlying mechanisms for aberrant neuroendocrine function in a range of psychopathologies. Thus, these findings highlight a rationale for studies seeking to explore associations with HPA axis dysfunction, especially the CAR, to examine the diurnal patterns of both cortisol and DHEA.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.