Salivary concentrations of cortisol and testosterone and prediction of performance in a professional triathlon competition

Abstract

The aim of this study was to examine salivary cortisol and testosterone concentrations in professional male athletes during a short triathlon competition using non-invasive methods, and to determine whether these hormone concentrations could be accurate predictors of performance. Eight adult male athletes (age, mean ± SEM: 27.8 ± 3.2 years; body mass index: 21.66 ± 0.42) in a professional triathlon team volunteered to participate in this study. Saliva samples were taken on the competition day and 7 days after competition on a rest day. The performance of the athletes was assessed by their rank order in the competition. Salivary cortisol concentrations were greater on the competition day than on the rest day in the early morning, immediately after waking up, 30 min later, immediately before the start of the competition, and later in the evening. Testosterone concentrations were greater on the competition day in the morning and in the evening. The diurnal rhythm of both cortisol and testosterone concentrations was maintained on both days and the testosterone/cortisol ratio (T/C ratio) was similar between days. The performance of the athletes was positively correlated with salivary cortisol concentration in the early morning of the competition day, but was not correlated with testosterone concentrations at any of the time points. In conclusion, early morning salivary cortisol concentration, but not T/C ratio, could be used to predict performance in athletes during a professional triathlon competition.

Keywords::

• Cortisol
• physical performance
• physical stress
• T/C ratio
• testosterone
• triathlon

Introduction

Studies in a wide range of endurance sports (e.g. cycling, marathon running, football, handball, rugby, tennis, and swimming) and force/heavy-resistance exercises (e.g. judo, taekwondo, weightlifting, and power-lifting) have almost all shown increased cortisol concentrations during their performance (Cook et al. 1986; O'Connor and Corrigan 1987; Keizer et al. 1989; Passelergue et al. 1995; Passelergue and Lac 1999; Cormack et al. 2008; Elloumi et al. 2008; Filaire et al. 2009; Le Panse et al. 2010; Pilz-Burstein et al. 2010). Cortisol is well known for its role in psychological and social stress (Pervanidou et al. 2007; Garcia et al. 2008; Osterberg et al. 2009), but it also plays an important role, in physical exercise. Cortisol secretion mobilizes energy stores, and it tends to increase with the intensity and duration of exercise as well as the training level of subjects (Keizer et al. 1989; Snegovskaya and Viru 1993a,b; Passelergue et al. 1995). An anticipatory increase in the level of cortisol has been observed in tennis players on the day of a competition (Filaire et al. 2009), attributable to psychological stress. Some studies (Snegovskaya and Viru 1993a,b; Passelergue et al. 1995; Salvador et al. 1999; Filaire et al. 2001), but not all (Chatard et al. 2002; Le Panse et al. 2010; Pilz-Burstein et al. 2010), have also shown that cortisol responsiveness to exercise is correlated with the performance level of an athlete.

Variations in testosterone secretion have also been described during acute exercise and the performance of various sports (Cook et al. 1986; Passelergue et al. 1995; Passelergue and Lac 1999; Filaire and Lac 2000; Elloumi et al. 2008; Fry and Lohnes 2010; Vingren et al. 2010). The increase in testosterone secretion before a competition has been associated with improved performance in psychomotor (Herrmann and Beach 1976) and mental activities (Klaiber et al. 1971; Herrmann et al. 1976), and may make the athlete more willing to use risk strategies (Daitzman and Zuckerman 1980). Thus, it has been proposed that both cortisol concentration and the testosterone/cortisol (T/C) ratio not only reflect adrenal and testicular activity but also might predict an athlete's performance. For this reason, analysis of salivary cortisol and testosterone may be useful in monitoring an athlete's progress (Snegovskaya and Viru 1993a).

A triathlon consists of three endurance sports, all of which must be completed by the athlete. The most common combination is swimming, cycling, and running, which can range from short to long distances. The short distance version consists of 1.5 km swimming, 42 km cycling, and 10 km running; whereas the long distance version includes 4 km swimming, 180 km cycling, and 42 km running.

Few studies have investigated adrenocortical hormones and other parameters in a triathlon competition. Rogers et al. (1986) showed that circulating cortisol concentration was increased three fold after a long distance competition and was negatively correlated with the athlete's aerobic capacity. Urhausen and Kindermann (1987) showed that free plasma testosterone concentration was reduced on the 2nd, 3rd, and 4th days after the competition and that cortisol concentration was increased 4.5-fold until the end of the competition and then gradually returned to the normal on subsequent days. This suggests an anabolic deficit lasting for several days as a result of prolonged physical strain. Odagiri et al. (1996) showed that serum cortisol concentration did not differ between athletes with or without psychological exhaustion after a triathlon competition, independent of their performance in the competition. However, there are no studies to date that investigate the relationship between cortisol and testosterone concentrations and athletic performance or how these hormones could serve as predictors of athletic performance during a triathlon competition. Moreover, most of these studies have assayed hormones in plasma or serum, which involves invasive blood collection procedures that may interfere with the results. For instance, blood collection was shown to be an additional stressor in tennis players (Kirschbaum and Hellhammer 1994). Given the positive correlations between these hormones in the blood and saliva both at rest and during exercise (O'Connor and Corrigan 1987; Port 1991; Del Corral et al. 1994; Paccotti et al. 2005; Cadore et al. 2008), the influence of physical exercise on hypothalamic–pituitary–adrenal axis activity has been investigated in a number of studies using the less invasive salivary cortisol method (Cook et al. 1986; Lopez Calbet et al. 1993; Passelergue and Lac 1999; Filaire and Lac 2000; Chatard et al. 2002; Cormack et al. 2008; Elloumi et al. 2008; Filaire et al. 2009). The only study to date that determined concentrations of cortisol and testosterone in saliva samples of triathlon athletes was conducted by Boehncke et al. (2009) for the purpose of comparing metabolic and hormonal parameters and the endurance capacities of diabetic and non-diabetic athletes.

Although there are conflicting results in the literature, it has been shown that for some sports, cortisol and testosterone concentrations might be related to physical performance, probably due to their influence on energy metabolism and/or behavior. Therefore, the hypothesis guiding this study was that the salivary concentration of these two hormones, as well as their ratio, would predict the performance of athletes during a triathlon competition. Thus, the aim of this study was to examine salivary cortisol and testosterone concentrations in professional male athletes during short triathlon competitions, taking place within a competitive season, and to determine whether the concentrations of these hormones could be accurate predictors of performance for this sport.

Material and methods

Subjects

Eight adult male athletes (age, mean ± SEM: 27.8 ± 3.2 years; body mass index: 21.66 ± 0.42) in a professional triathlon team (national level) volunteered to participate in this study. The study was approved by the Committee of Ethics in Research of the Federal University of São Paulo (UNIFESP), Santos, São Paulo, Brazil (certificate number: 0191/07) and was conducted according to the principles of the Helsinki Declaration. Formal permission to work with the team was given by the team's head coach. Prior to data collection, athletes consented not to use any medications and signed an informed consent form that explained the nature of the study.

Protocol

Written and verbal instructions on how to collect saliva samples were given to the athletes. They were strongly encouraged to follow the instructions closely and were required to refrain from eating, drinking, chewing gum, or brushing teeth for at least 30 min before collecting any samples.

Six saliva samples were taken on a competition day as well as approximately 7 days after competition on a rest day according to the following schedule: one sample immediately after awakening at 06:30 h, and then again at 07:00 h (30 min after awakening), 09:00, 12:00, 17:00 and 22:00 h. The competitions were performed at 09:00 h and lasted, in full, about 2–4 h (1.5 km swim, 42 km cycling, and 10 km running).

The decision to collect samples 7 days after the competition, here called rest days, was based on the training schedule, which required that the training load during this period was of low intensity and low volume, whereas the training load during the 7 days before the competition was based on a protocol of high intensity and low volume.

Performance

The athletes' performance was determined by their rank order in the analyzed competition, defined by the elapsed time to complete that competition, including all three modalities of the triathlon (running, swimming, and cycling).

Saliva sampling and analysis

Saliva samples were collected using Salivette^® devices (Sarstedt, Numbrecht, Germany), and were promptly stored in refrigerated boxes (competition day) or the home freezer (resting day) before transportation to the laboratory, where they were stored and kept frozen at − 20°C. Samples were thawed and centrifuged at least once to allow precipitation of mucins before the analysis of salivary cortisol and testosterone concentrations using enzyme immunoassay kits (Diagnostic Systems Laboratories, Inc., Webster, NY, USA and Salimetrics LLC, State College, PA, USA). The analytical sensitivities were 0.011 μg/dl and 1.0 pg/ml and the coefficients of variation (inter- and intra-assay) were always < 10%.

Statistical analysis

Data are presented as means ± SEM and analyzed by two-way analysis of variance (ANOVA) followed by Newman–Keuls multiple comparison tests for pair-wise comparisons between data on the competition day and the rest day. Correlations between salivary hormone values and performance were calculated using Pearson's product-moment correlation test. The null hypothesis was rejected when p < 0.05.

Results

Salivary cortisol

Cortisol concentration was higher on the competition day than on the rest day immediately after waking up (ANOVA, F_1,14 = 4.61, p < 0.05), 30 min later (ANOVA, F_1,14 = 5.75, p < 0.05), immediately before the start of the competition at 09:00 h (ANOVA, F_1,14 = 6.90, p < 0.05), and in the evening at 22:00 h (ANOVA, F_1,14 = 8.97, p < 0.05). Salivary cortisol concentration was similar on competition day and rest day in the samples collected immediately after the end of the competition and at 17:00 h (Figure 1). The area under the curve for cortisol was also greater on the competition day than on the rest day (7.08 ± 0.55 μg/dl vs. 6.51 ± 0.32 μg/dl; respectively; ANOVA, F_1,14 = 6.60, p < 0.05). The diurnal rhythm of cortisol was maintained on competition day and rest day with high values in the morning and low values in the evening (Figure 1). The cortisol awakening response was not significantly altered on the competition day compared with the rest day (0.54 ± 0.53 μg/dl vs. 0.24 ± 0.12 μg/dl; respectively); however, it is important to emphasize that the variability in cortisol awakening response was greater on the competition day, as shown by the SE of the mean.

Figure 1.  Mean ( ± SEM) salivary cortisol concentrations in male athletes on the day of a short professional triathlon competition and on a rest day (about 7 days after the competition); n = 8; pairs of identical letters show statistical difference between the respective pairs of vertical bars (ANOVA, followed by the Newman–Keuls test, p < 0.05).

Salivary testosterone

The diurnal rhythm of testosterone concentration was also preserved on both days, with high concentration in the morning and low concentration in the evening, although testosterone concentration was higher on the competition day than on the rest day in both the morning (ANOVA, F_1,14 = 4.68, p < 0.05) and evening at 22:00 h (ANOVA, F_1,14 = 15.71, p < 0.05) (Figure 2).

Figure 2.  Mean ( ± SEM) salivary testosterone concentrations (morning: at 07:00 h; evening: at 22:00 h) in male athletes on the day of a short professional triathlon competition and on a rest day (about 7 days after the competition); n = 8; pairs of identical letters show statistical difference between the respective pairs of vertical bars (ANOVA, followed by the Newman–Keuls test, p < 0.05).

Testosterone/cortisol ratio

As seen in Figure 3, T/C ratio was similar on the competition day and the rest day. The T/C ratio was higher in the evening than in the morning due to a more pronounced decline in cortisol concentration in the evening.

Figure 3.  Mean ( ± SEM) salivary T/C concentration ratio (morning: at 07:00 h; nocturnal: at 22:00 h) in male athletes on the day of a short professional triathlon competition and on a rest day (about 7 days after the competition). There was no significant difference between the competition day and the rest day (n = 8; ANOVA, followed by the Newman–Keuls test, p>0.05).

Correlation between salivary hormone concentration and performance

The time taken for each athlete to complete the triathlon was used as an indicator of performance. The winning athlete completed the triathlon in 2h 05min 29s (in the “2010 Triathlon World Championship” the best time was 1h 42min 26s; www.triathlon.org).

As demonstrated in Figure 4, athletic performance was positively correlated with salivary cortisol concentration in the early morning of the competition day (at waking, Panel A, r = 0.79, p < 0.05, and 30 min after awakening, Panel B, r = 0.76, p < 0.05). Figure 5 shows that the athletic performance was not correlated with salivary testosterone concentration at any of the time points on the competition day.

Figure 4.  Correlations between salivary cortisol concentrations and performance of male athletes on the day of a professional short triathlon competition. Panel A: waking cortisol concentrations; panel B: cortisol concentrations 30 min after awakening. The total time to complete the competition for each athlete is placed above each point in panel A, in h:min:s. Pearson's correlation tests were used to determine r and p values shown on graphs.

Figure 5.  Correlations between salivary testosterone concentrations and performance of male athletes on the day of a professional short triathlon competition. Panel A: morning testosterone concentrations; panel B: evening testosterone concentrations. The total time to complete the competition for each athlete is placed above each point in panel A, in h:min:s. Pearson's tests showed no significant correlations.

Discussion

The main finding of this study was that early morning salivary cortisol concentration can predict athletic performance in a triathlon competition, which supports the hypothesis for this study. The data also show that athletes have higher concentrations of salivary cortisol and testosterone during the early morning hours of a competition day than on the rest day, indicating an anticipatory response to the stress of the competition similar to that described by Filaire et al. (2009) for tennis players.

In addition, the athletes exhibited a delayed cortisol response to the stress of the competition at 22:00 h. This finding is in agreement with most studies in the field, which have demonstrated an increase in cortisol secretion after intense exercise (Cook et al. 1986; O'Connor and Corrigan 1987; Keizer et al. 1989; Passelergue and Lac 1999; Cormack et al. 2008; Elloumi et al. 2008; Filaire et al. 2009). However, the salivary cortisol concentration immediately after the competition was similar to that observed on the rest day at the same time. This immediate return to basal cortisol concentration may reflect the high level of physical fitness of the athletes, and indicates that the physical stress of an Olympic triathlon competition had a lesser impact on cortisol secretion than the psychological or anticipatory stress on the morning of the competition. However, the positive correlation between the anticipatory response to competition and the athletes' performance indicates a physiological mechanism for this phenomenon. Previous studies have linked the increase in cortisol levels with the improvement of motor coordination (Herrmann and Beach 1976), alertness, and attention (Klaiber et al. 1971; Herrmann et al. 1976), as well as the mobilization of energy substrates (e.g. mobilization of free fatty acids that are essential for the maintenance of long-term physical activity, a characteristic of a triathlon; Keizer et al. 1989; Snegovskaya and Viru 1993a,b; Passelergue et al. 1995). All these mechanisms can contribute to the improvement of physical performance through enhancement of the ability of the body to cope with the upcoming physical challenge. Therefore, the anticipatory response to exercise can be considered as a specific pattern of reaction to “pre-competition” stress.

According to a classification of stress agents (Pacàk and Palkovits 2001), a triathlon can be considered a stressor with physical and psychological components. Furthermore, the exact magnitude of the response may depend on the intensity and duration of the exercise, as well as on the training level of the subjects and their ability to mobilize energy reserves (Keizer et al. 1989; Snegovskaya and Viru 1993a,b; Passelergue et al. 1995).

Cortisol is secreted according to a typical daily rhythm that peaks shortly before waking and falls throughout the day toward a nadir at midnight. An additional increase in cortisol secretion occurs 20–45 min after waking (i.e. the cortisol awakening response; Kirschbaum and Hellhammer 1994; Clow et al. 2004). The cortisol rhythm can be disrupted and cortisol concentration is altered by stress (McEwen 1998; Bella et al. 2010), psychological and environmental influences such as low socioeconomic status (Steptoe et al. 2003; Garcia et al. 2008), or chronic inflammatory diseases (Petrelluzzi et al. 2008). The data presented here have shown that for triathlon athletes, the cortisol rhythm is maintained on the rest day and the competition day, although in the latter the concentration is higher during the morning and late at night. As a result of this adrenal response to the competition, the area under the cortisol curve was greater on the competition day, hence athletes are exposed to higher concentrations of cortisol on these days. However, the cortisol awakening response on the competition day was not significantly different from the rest day but instead was more variable (as indicated by a larger SEM value). This may reflect that the cortisol awakening response seems not to be an accurate biomarker of stressful situations, indicating different mechanisms regulating the cortisol awakening and stress responses (Clow et al. 2010a,b).

Nevertheless, it has to be considered that training might affect basal cortisol concentration as data from our laboratory have shown that morning salivary cortisol concentration on the rest day was similar in professional basketball players (0.91 ± 0.11 μg/dl; unpublished data) and triathlon athletes (0.88 ± 0.10 μg/dl), but higher than those of non-athlete men (0.68 ± 0.07 μg/dl) (Garcia et al. 2008). At other analyzed time-points, this difference was not clear.

Testosterone secretion also shows a daily rhythm that is similar to the cortisol rhythm, with higher concentrations in the morning (Guignard et al. 1980; Touitou and Haus 2000; Lac and Chamoux 2006) but no waking response. Because the effects of testosterone are mainly anabolic and those of cortisol are predominantly catabolic, the ratio between the concentrations of T/C is considered an index of the balance between catabolism and anabolism (Passelergue and Lac 1999; Elloumi et al. 2008). The data presented here show that salivary testosterone concentration was higher in the morning and evening of the competition day and that the T/C ratio was not altered. Therefore, although athletes are exposed to high concentrations of both catabolic and anabolic hormones, the balance between catabolism and anabolism seems to be preserved. This finding differs from that in studies of wrestlers and rugby players, which demonstrate that the T/C ratio decreases during the high intensity (catabolic) phase (Passelergue and Lac 1999; Elloumi et al. 2008). Since we did not observe changes in the T/C ratio induced by the competition, our results contradict previous studies which suggest that this parameter may be a useful instrument to predict physical performance (Snegovskaya and Viru 1993a).

In agreement with the present data, Iurimiaé et al. (1989) have shown that circulating testosterone concentration increases in response to an Olympic triathlon competition. Although testosterone concentration may be positively associated with psychomotor performance and psychological motivation (Klaiber et al. 1971; Herrmann et al. 1976), our results showed no correlation between salivary testosterone and performance in the triathlon competition. In contrast, the present data showed a strong correlation between athletic performance and morning salivary cortisol concentration. This finding agrees with previous studies of rowers, weightlifters, and judo athletes which found a positive correlation between physical performance and elevated cortisol concentration (Snegovskaya and Viru 1993a,b; Passelergue et al. 1995; Salvador et al. 1999; Filaire et al. 2001). Indeed, Snegovskaya and Viru (1993a) reported that an improvement in the performance capacity of rowers was associated with elevated cortisol concentration during supramaximal exercise. However, studies of swimmers, weightlifters, and taekwondo performers showed no correlation between physical performance and salivary or circulating cortisol concentration (Chatard et al. 2002; Le Panse et al. 2010; Pilz-Burstein et al. 2010). It is noteworthy that conflicting results between studies can be attributed to the diverse stressors related to each sport, which fits well with the theory of specificity of the stress response (Weiner 1991; Chrousos and Gold 1992; Chrousos 1998; Pacàk and Palkovits 2001). According to this theory, the stress response depends on the nature, intensity, frequency, and other parameters of the stressor which could explain the large variability in the hormonal profiles of athletes in different sports.

It is also important to emphasize that this study, which demonstrated a correlation between pre-competition hormonal levels with performance, stands out from most studies conducted to date, which showed correlation between hormonal concentrations and performance during and/or immediately after competitions.

In conclusion, the increase in salivary cortisol concentration during a professional triathlon, particularly in the morning, could be used to predict performance in athletes. In addition, the T/C ratio was unable to predict the performance of athletes in an Olympic triathlon competition. Further work is needed to determine how the analysis of not only cortisol and testosterone but also of other indicators, such as salivary alpha-amylase and dehydroepiandrosterone, might be useful as tools in planning and monitoring the intraindividual progress of a triathlete.

Declaration of interest: This study was carried out with the financial support of CAPES, CNPq, and FAPESP. The authors wish to thank the athletes for their dedicated performance and collaboration as well as to Rosana Merino, the coach, for her expert assistance. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.