Resting steroid hormone concentrations in lifetime exercisers and lifetime sedentary males

Abstract

Introduction: Advancing age in men is associated with a progressive decline in serum testosterone (T) and interactions between exercise, aging and androgen status are poorly understood. The primary aim of this study was to establish the influence of lifelong training history on serum T, cortisol (C) and sex hormone binding globulin (SHBG) in aging men. A secondary aim was to determine the agreement between serum and salivary measurement of steroid hormones in ageing men.

Methods: Serum and salivary steroid hormones (serum C, T and SHBG, and salivary measures of C and T) were determined and compared between two distinct groups; lifelong exercising males (LE [n = 20], 60.4 ± 4.7 year) and age matched lifelong sedentary individuals (SED [n = 28], 62.5 ± 5.3 years).

Results: T-test revealed a lack of significant differences for serum C or SHBG between LE and SED, while Mann-Whitney U revealed a lack of differences in total T (TT), bioavailable T (bio-T) or free testosterone (free-T). Further, salivary T (sal-T) did not correlate with serum markers of T in LE, SED, or when pooled (r = 0.040; p > 0.05).

Conclusions: Findings from this investigation suggested that resting levels of serum T and calculated free-T was unable to distinguish between diverse lifelong training histories in aging men. Further, sal-T was not an appropriate indicator of serum T and calculated free-T values in older males and considerable caution should be exercised when interpreting sal-T measurements in aging males.

• Aging
• cortisol
• exercise
• saliva
• sex hormone binding globulin
• testosterone

Introduction

The age associated gradual decline of systemic testosterone (T) levels in men is known to be considerably heterogeneous [1]. This decline results from testicular changes and altered neuroendocrine regulation of luteinising hormone (LH) secretion [2]. The decline in systemic T occurs in the presence of a concomitant increase in serum levels of sex hormone binding globulin (SHBG) [3] resulting in a pronounced decline in free testosterone (free-T) levels [4].

There is some evidence that chronic exercise may assist the maintenance of muscle mass and strength with advancing age [5–7]. Additionally, T supplementation studies have demonstrated that when combined with exercise, significant increases in muscle mass can be achieved [8,9], suggesting an interaction between circulating T concentrations and maintenance of muscle mass [10,11]. Given these muscle–steroid hormone interactions, it should not be unexpected for lifelong exercisers to have more favorable steroid hormone profile than lifelong sedentary counterparts.

T in serum is partly bound to SHBG and albumin and only free-T is considered to be available for uptake by tissues [2]. Salivary T (sal-T) is purported to represent free-T [12], leading a number of authors describe sal-T as “free testosterone” [13,14]. However, since SHBG has been identified in saliva [15] and given the increasing concentrations of SHBG with age, the relationship between sal-T and calculated free-T remains to be determined in aging men.

Therefore, the primary aim of the present study was to examine the influence of lifelong training status on serum T in older males. A secondary aim was to examine the agreement between serum and saliva measurements of steroid hormones using standard analytical techniques. We hypothesized that lifelong exercisers would demonstrate higher concentrations of serum total T (TT) than their sedentary counterparts. A secondary hypothesis was that significantly positive levels of agreement would exist between serum and salivary measures of steroid hormones in ageing men with diverse lifelong training history.

Methods

Participants

Participants were recruited by word of mouth and consisted of n = 20 lifelong exercisers (LE) (aged 60 ± 5 years; height 174 ± 6 cm; body mass 79 ± 12 kg; peak oxygen uptake (VO₂ peak) 40 ± 6 mlkg min⁻¹) and n = 28 sedentary (SED) males (aged 63 ± 5 years; height 175 ± 6 cm; body mass 90 ± 18 kg; (VO₂ peak) 27 ± 5 mlkg min⁻¹). Participant training history was self-reported and peak aerobic capacity was used to confirm participant allocation to either LE or SED.

Salivary collection and analysis

Duplicate salivary samples of approximately 1.8 ml were collected via passive drool into graduated 2 ml cryovials (Salimetrics, State College, PA). To prevent potential contamination, participants were advised to avoid drinking hot fluids or brushing their teeth 2 h prior to testing. Salivary samples were collected and stored at −80 °C until assay. Samples were assayed in duplicate (without separation or extraction) for sal-C and sal-T using commercially available immunoassay protocols (Salimetrics, State College, PA). Intra- and inter-assay coefficients of variation were 4% and 9% for sal-C and 7% and 10% for sal-T.

Blood collection and analysis

Venous blood and saliva samples were obtained between 07:00 and 09:00 h following a 12 h overnight fast. Blood was sampled from an antecubital forearm vein following 20 min of supine rest using a 20-gauge disposable needle and Vacutainer tube holder (Becton Dickinson, Oxford, UK). Approximately, 14 ml of blood was withdrawn into two 10 ml serum separator (SST) tubes, prior to being centrifuged at 6000 rpm at 15 °C for 15 min. The resultant serum was divided into appropriate aliquots and stored at −80 °C until subsequent analysis.

Serum concentrations of TT, SHBG and C were determined by electrochemiluminescent immunoassay on the E601 module of the Roche Cobas 6000 (Burgess Hill, West Sussex, UK). Inter-assay CVs over a six months period were 4.5%, 2.4% and 4.2% for TT, SHBG and C, respectively. Analyses were carried out in the Clinical Biochemistry Laboratory at Royal Glamorgan Hospital (Wales, UK). Free-T and bioavailable T (bio-T) was calculated using the Vermueulen formula [16], which has been validated against equilibrium dialysis [17]. TT adjustment for body mass index (BMI) was calculated by dividing TT by BMI and multiplying by 100. TT adjustment for fat mass was calculated by dividing TT by fat mass and multiplying by 10.

Body composition

Height was measured to the nearest 0.1 cm using a stadiometer (Seca, Birmingham, UK). A multi frequency bioelectric impedance analyzer (BIA), (Tanita MC-180MA Body Composition Analyzer, Tanita UK Ltd.) was used to measure body composition as previously described [18]. GMON software (v1.7.0, Tanita UK Ltd.) generated values for fat mass, fat free mass (FFM) and body fat percentage. BMI was calculated by dividing subject weight in kilograms by the square of the participant’s height in meters.

Determination of peak aerobic capacity (VO₂ peak)

Methods have been previously described elsewhere [18]. Briefly, a ramped protocol was preceded by a 3 min warm up on an air-braked cycle ergometer (Wattbike Ltd., Nottingham, UK). Work-rate was increased every minute by raising the dampener setting by one until volitional exhaustion. Oxygen uptake was measured throughout the exercise test using a rapid-response gas analyzer (Metamax II, Cortex, Leipzig, Germany) and gas exchange variables were calculated and displayed breath-by-breath after accounting for the delay between the volume and concentration signals. The volume transducer was calibrated prior to each test with a 3 L calibration syringe and analyzers were calibrated with precision-analysed gases that spanned the expected range of expired oxygen (O₂) and carbon dioxide (CO₂) concentrations. Heart rate (HR) was recorded every 5 s using short-range telemetry. A fingertip blood sample was collected into a portable automated lactate analyzer (Lactate Pro, Arkray, Inc., Kyoto, Japan) within 45 s of the termination of exercise. VO₂ peak was calculated as the highest 20 s value achieved during the test.

Determination of peak power output

The Quebec 10 s cycling test consisted of 10 s maximal sprint against constant resistance on an air-braked cycle ergometer (Wattbike Ltd., Nottingham, UK) and has been detailed elsewhere [18]. Participants completed a standardized 3 min warm-up involving pedaling at 60 rpm interspersed with three 2–3 s sprints. Tests commenced from a standing start. Participants were verbally encouraged throughout the test to avoid pacing and to sustain their supramaximal effort throughout the test. A recovery period of 5 min was permitted between the warm-up and the test. Power output was calculated each second for the duration of the test and peak power over 1 s was recorded.

Statistical analysis

Data were analyzed using SPSS (version 20) (IBM North America, New York, NY). To determine parametricity, Levene’s tests (homogeneity of variance) and Shapiro-Wilk (normal distribution) were employed. Where parametric assumptions were met, data were analyzed using t-test to compare differences between groups. Where parametric assumptions were breached, Mann-Whitney U was employed. Pearson’s product moment correlation was used to determine agreement between measures. Significance was set a priori at p ≤ 0.05 and observed statistical power was 0.97. Data are presented as mean ± SD.

Results

T-test revealed a lack of significant differences in age, or serum C (p > 0.05) between groups. Body fat percentage was significantly lower and power, adjusted TT (to both BMI and absolute fat mass), and peak aerobic capacity as determined by VO₂ peak were significantly higher in the LE group (Table 1). LE had significantly lower blood pressure and BMI than SED counterparts. TT, SHBG, free-T and bio-T were not significantly different between groups.

Table 1. Mean exercise capacity, hormonal and anthropometric values between groups consisting aging lifelong exercisers (LE) and age matched lifelong sedentary (SED) men.

Variable	LE	SED
VO2 peak (ml kg min^−1)	40.0 ± 5.7**	27.6 ± 4.6
FFM (kg)	62.8 ± 5.9	63.76 ± 8.69
Body fat percentage (%)	19.7 ± 6.6**	28.73 ± 5.00
BMI (kg/m^2)	25.8 ± 3.1**	29.75 ± 4.75
Peak power output (W)	783.6 ± 171.9**	672.8 ± 156.0
TT (nmol L^−1)	14.76 ± 3.7	13.25 ± 6.15
C (nmol L^−1)	302.8 ± 134.4	318.2 ± 136.6
TT adjusted for BMI ( × 100)	599 ± 156*	467 ± 212
TT Adjusted for fat mass ( × 10)	10.98 ± 7.69**	5.57 ± 2.70
SHBG (nmol L^−1)	48.6 ± 13.8	42.99 ± 23.22
Bio-T (nmol L^−1)	5.83 ± 0.67	5.27 ± 1.62
Free-T (pmol L^−1)	248 ± 28	225 ± 69
Systolic blood pressure (mmHg)	127.1 ± 12.4**	137.2 ± 14.1
Diastolic blood pressure (mmHg)	80.4 ± 6.6**	84.9 ± 6.8

VO₂ peak = peak aerobic capacity; FFM = fat free mass; BMI = body mass index; TT = total testosterone; SHBG = sex hormone binding globulin; bio-T = bioavailable testosterone; free-T = free testosterone. Values are mean ± SD.

*Difference is significant at the 0.05 level.

**Difference is significant at the 0.01 level.

Sal-T did not correlate with serum T (r = 0.040, p = 0.813), SHBG (r = −0.106, p = 0.531), bio-T (r = 0.133, p = 0.432) or free-T (r = 0.134, p = 0.431) in either LE or SED. Sal-C correlated significantly with serum C in both LE (r = 0.540, p < 0.05) and SED (r = 0.612, p < 0.05). TT did not correlate with BMI (r = −0.175, p = 0.31) or absolute fat mass (r = −0.147, p = 0.37).

Discussion

The main finding of the present study is that lifelong exercisers, despite having more favourable body composition and cardiorespiratory fitness, have resting steroid hormone concentrations similar to their lifelong sedentary counterparts. However, these similarities did not remain when TT was adjusted for BMI or absolute fat mass. Further, although sal-C correlated well with concentrations of C in serum, sal-T did not significantly correlate with serum values of TT, bio-T or free-T, indicating that sal-T is a poor determinant of serum T in ageing men, irrespective of exercise training history.

Typically, resistance trained younger men are known to display higher basal levels of T compared with sedentary counterparts [19]. Although differences in exercise training mode may provide some explanation for differences between studies, there is some support for a lack of impact of exercise training on systemic steroid hormone profiles in the literature. For instance, our findings are in agreement with those of Lovell and colleagues [20] who suggest that training status (sedentary, resistance trained and endurance trained) does not influence absolute resting TT or SHBG concentrations. Lovell and co-workers reported this phenomenon over a 16 weeks training intervention in a cohort of ageing men. Our data extend this and suggest the same phenomenon is manifest following lifelong exercise training. Cogitating the findings of Tremblay et al. [19] and Lovell et al. [20] in conjunction with the present findings, it is possible that training status influences resting anabolic hormones in younger males but advancing age and body composition abrogates this effect. However, Ari et al. [21] reported that a group of master athletes demonstrated higher TT than sedentary controls while Arazi et al. [22] reported that middle aged men (∼50 years) demonstrate higher basal TT concentrations following eight weeks of resistance training. Therefore, it is unclear whether training status influences basal steroid hormone concentration. This may be an area for further exploration with focus pertaining to training modality (resistance or endurance). Yeap and colleagues [23] suggested that healthy behaviours (including exercise) correlated with both TT and SHBG, but not free-T in a similar but larger cohort. In total, Yeap et al. [23] assessed eight behaviours in a lifestyle score that correlated with TT. As these additional factors were not measured in the present study, we cannot discount the possibility that differences in adjusted TT may be due to these factors as the present investigation grouped participants solely on exercise history and confirmed differences with VO₂ peak. Indeed it is likely that participants engaged in lifelong exercise would also engage in at least some of those additional healthy behaviours identified by Yeap et al. [23].

The measurement of sal-C concentrations of older males using a commercial ELISA demonstrated a strong relationship with serum C concentration, and is in agreement with previous work [24–26]. Reid et al. [27] have previously recognized the potential for of salivary determination of C to be an effective method for determining systemic concentrations in aging males. Furthermore, C has been reported to increase with age in saliva [28] and serum [29]. In the present investigation, sal-T was not significantly related to serum values of T, SHBG, calculated free-T or bio-T in older males. The reason why sal-C, but not sal-T, correlated with serum values may be related to known differences in absolute concentrations of the respective hormones and the assay specifications (e.g. sensitivity and specificity) for each hormone, as well as hormone pharmacokinetics (e.g. metabolism and diffusion rate) within the salivary glands [30]. Sal-C is known to have a rapid transmission across the salivary membrane [31]. Another possible factor influencing these relationships may be the low absolute concentrations and narrow data range seen in older individuals. Data from large multicentre trials indicate that systemic T decreases, while SHBG increases linearly with age [1]. Consequently, inter- and intra-individual differences in SHBG levels may also further limit the validity of sal-T measures as SHBG bound T is considered to be unavailable for transport into saliva [32] and known to increase with advancing age [33]. The present study provides some support for this, as SHBG values in this investigation are higher than those previously reported in young males [34].

There are some important limitations in the present study that should be noted. The first relates to the limited sample size, distribution and range of serum values. However, this was in line with previous observational [21] and experimental [20] studies. A further potential confounding factor in this and previous work may relate to androgen receptor (AR) regulation [35]. Changes in AR regulation remain unaccounted for in this and previous work and little is known about the interactions between aging, serum T and AR regulation. However, given that in the present study, LE had a greater peak power output than SED, this may be a possible avenue for further exploration.

In conclusion, the primary finding of the present study is that, despite potentiating a more favourable body composition and , lifelong exercise training does not influence resting systemic concentrations of TT, C, SHBG or calculated free-T or bio-T. A secondary finding was that although sal-C is a significant predictor of serum C, sal-T is poorly predictive of serum T in older males. We recommend that considerable caution should be exercised when using sal-T as a determinant of androgenic status in aging men.

Declaration of interest

The authors report no declarations of interest.