Plasma and Salivary Steroid Hormone Responses of Men to High-Intensity Cycling and Resistance Exercise

Abstract

Hough, JP, Papacosta, E, Wraith, E, and Gleeson, M. Plasma and salivary steroid hormone responses of men to high-intensity cycling and resistance exercise. J Strength Cond Res 25(1): 23-31, 2011-Hormonal responses to exercise could be used as a marker of overreaching. A short exercise protocol that induces robust hormonal elevations in a normal trained state should be able to highlight hormonal changes during overreaching. This study compared plasma and salivary cortisol and testosterone responses to 4 exercise trials; these were (a) continuous cycle to fatigue at 75% peak power output (Ẇ_max) (FAT); (b) 30-minute cycle alternating 1-minute 60% and 1 minute 90% Ẇ_max (60/90); (c) 30-minute cycle alternating 1-minute 55% and 4-minute 80% Ẇ_max (55/80); and (d) Squatting 8 sets of 10 repetitions at 10 repetition maximum (RESIST). Blood and saliva samples were collected pre-exercise and at 0, 10, 20, 30, 40, 50, and 60 minute postexercise. Pre- to postexercise plasma cortisol increased in all exercise trials, except 60/90. Increases in 55/80 remained above pre-exercise levels for the entire postexercise period. Salivary cortisol increased from pre- to postexercise in FAT and 55/80 trials only. Once elevated after 55/80, it remained so for the postexercise period. Plasma testosterone increased from pre- to postexercise in all trials except 55/80. Saliva testosterone increased from pre- to postexercise in all trials with the longest elevation occurring after 55/80. Area under the curve analysis indicated that the exercise response of salivary hormones was greater in all cycle trials (cortisol) and in the 60/90 and 55/80 trials (testosterone) compared with the other trials. This study indicates that the 55/80 cycle protocol induces a prolonged salivary and plasma cortisol and salivary testosterone response compared with the other trials and so may be a useful diagnostic tool of overreaching.

Introduction

Intensified training that results in overreaching is extensively used by athletes in an attempt to improve performance. Overreaching causes a decrement in performance for a limited period, but after sufficient recovery, a super compensatory increase in performance may occur (^9,16). Despite the benefits of overreaching, it is possible to develop the overtraining syndrome (OTS) if insufficient recovery occurs, and full recovery from OTS may take many weeks or months (¹¹). Having a reliable biological marker to monitor the presence of overreaching could therefore reduce the occurrence of OTS.

Previous studies have suggested that resting levels of the hormones cortisol and testosterone could be used as markers of overreaching; however, studies in this area have reported contrasting findings (^8,12,19). This indicates that resting measurements of cortisol and testosterone may not be sensitive enough to be used as a reliable biological marker of overreaching.

Meeusen et al. (^14,15) suggested that measuring the hormonal response to 2 consecutive incremental cycles to fatigue separated by a short recovery period may identify the presence of overreaching. They reported a blunted response of cortisol to exercise in a group of overreached athletes compared with when the athletes were in a normal trained state. Our laboratory used the protocol of Meeusen et al. (^14,15) before and after a 4-day intensified training period but were unable to replicate their findings (unpublished data). It was concluded that an exercise protocol of a similar duration but at a higher intensity may be more appropriate to highlight hormonal changes that occur in an overreached athlete.

First, it is necessary to design an exercise protocol that will elicit a robust increase in cortisol and testosterone when the athlete is in a normal trained state. A protocol like this could highlight any changes in exercise-induced cortisol and testosterone when an athlete increases training loads in an attempt to overreach. The response of cortisol to exercise depends on the intensity, duration, and mode of exercise completed. In general, the exercise intensity needs to be above a moderate level (>60% V̇O₂max) and last for at least 20 minutes. In addition, the responses of cortisol to exercise may not be immediate (¹⁷). Both cycling and resistance exercise bouts have also been shown to elicit acute increases in cortisol and testosterone (^1,13,14).

Blood and saliva can be used to measure cortisol and testosterone concentrations. Previous studies have demonstrated positive correlations between blood and salivary cortisol and testosterone (¹). It would be advantageous to be able to measure hormonal levels with a noninvasive method such as saliva sampling. Peak salivary hormone levels are likely to occur later than in plasma because it takes some time for the hormones to diffuse into the saliva. Hence, the timing of peak postexercise hormone levels in saliva needs to be established.

The aim of this study was to find a short-duration (∼30 minutes), high-intensity exercise bout that would cause a robust elevation of plasma and salivary cortisol and testosterone while the athlete is in a normal trained state. The relationship between the plasma and salivary hormonal responses to these exercise tasks was also examined. It was hypothesized that each of the protocols would cause an increase in salivary and plasma cortisol and testosterone, but the magnitude and time course of these increases may not be the same in all protocols. In addition, it was hypothesized that the plasma and salivary measurements would show a positive correlation. Finally, the time at which the postexercise hormonal peaks occurred was examined. The expectation was to find hormonal changes in the saliva to occur slightly later than in plasma.

Methods

Experimental Approach to the Problem

This study assessed if short-duration, high-intensity exercise would induce an increase in the hormones cortisol and testosterone while the athlete is in a normal trained state. This study was a randomized, crossover design where 3 short duration (∼30 minutes), high-intensity cycling trials and a resistance exercise trial were completed. The 3 cycling trials consisted of different combinations of exercise intensities but were designed to elicit similar workloads. The high-intensity resistance trial was completed in a similar duration as the cycling trials. Saliva and blood samples were collected pre- and at 0, 10, 20, 30, 40, 50, and 60 minutes postexercise (with blood collection until 10 minutes postexercise only in the resistance exercise trial). Concentrations of salivary and plasma cortisol and testosterone were determined, and the magnitude and time course of the hormonal responses to each exercise trial were compared.

Subjects

Ten healthy male subjects volunteered to take part in this study. Each subject exercised at least 3 times per week, and they all had some experience in weight training; however, none were competitive athletes. The subject characteristics are outlined in Table 1.

Table 1: Descriptive characteristics of subjects.

The study was approved by the Loughborough University Ethics Committee. After obtaining approval, a full verbal and written explanation of this study and possible risks involved was given to each subject. Written informed consent to participate was obtained before the testing began.

Procedures

All subjects reported to the laboratory on 7 separate occasions. To avoid any circadian rhythm and seasonal variation in the hormones analyzed, all testing sessions took place at the same time of the day and during the UK summer months of May to August. For the main trial visits, each subject consumed a standard breakfast 4 hours before testing began. Subjects remained fasted until the end of each main trial but drank water ad libitum during this time. The participants abstained from exercise, caffeine, and alcohol intake 24 hours before each trial. All subjects were given instructions on measuring, weighing, and recording food intake and were asked to complete a food record diary 24 hours before each main trial and were instructed to consume a diet as similar as possible before each main trial day. Caloric and macronutrient intake was determined by the use of CompEat version 5.8 software (Nutrition Systems, Oxford, United Kingdom). Mean energy intake 24 hours before each trial was 10.7 ± 3.2 MJ with 56 ± 11% from carbohydrate, 28 ± 10% from fat and 16 ± 4% from protein. Body mass was measured with the subjects in shorts and socks before all trials. During exercise, heart rate (HR) and rating of perceived exertion (RPE) using a 6-20 Borg scale were recorded in the last 15 seconds of each minute.

Maximal Oxygen Uptake Assessment

On the first laboratory visit, a continuous, incremental Maximal Oxygen Uptake (V̇O₂max) test was completed on an electronically braked cycle ergometer (Lode, Groningen, The Netherlands). The test began at 95 W, and the duration of each stage was 3 minutes. The workload in the second stage increased to 165 or 200 W, depending on the known fitness level or V̇O₂max of the subject from a previous test in the last year. The V̇O₂max test was designed to be completed in the recommended time of 9-15 minutes (²). After the second stage, the workload was increased at the end of every stage by 35 W until volitional exhaustion. When the subject's HR reached 140 b·min⁻¹ expired gas was collected for 1 minute into Douglas bags during the final minute of each stage and during the final minute of the exercise test. Expired gas was analyzed using an O₂/CO₂ analyzer (Servomex 1440, Crowborough, United Kingdom) along with a dry gas meter (Harvard Apparatus, Edenbridge, United Kingdom) for the determination of the rates of oxygen consumption (V̇O₂) and carbon dioxide production (V̇CO₂). The HR was recorded continuously using short range radio telemetry (Polar F2, Polar Electro Oy, Kempele, Finland). Maximum power output (Ẇ max) was determined using the equation

where Ẇ_final is the power output during the final stage completed, t is the amount of time (seconds) reached in the final uncompleted stage, T is the duration of each stage (180 seconds), and W_inc is the workload increment (35 W). Power outputs equivalent to 55, 60, 75, 80, and 90% of Ẇ_max for each individual were calculated, and these values were used as the power outputs (Ẇ) during the main trials.

Ten-Repetition Maximum Assessment

On the second visit to the laboratory, the participant completed a 10 repetition maximum (RM) squat test. An adjustable box was placed behind the subject when squatting so that the subject's knee angle did not exceed 90°. Only attempts reaching a knee angle of 90° were considered successful attempts. The knee angle was verified by the same experimenter for all trials. At the beginning of the test, a warm-up set of 10 repetitions with 20-kg resistance was completed. Each set was separated by at least a 3-minute rest. The resistance was increased in appropriate measures to establish the 10RM in at least 7 attempts.

Main Trials

In visits 3-7, the participant undertook 5 main trials. These were completed in a randomized order. The trials were as follows: (a) Continuous cycle to volitional fatigue at 75% Ẇ_max (FAT); (b) 30-minute continuous cycle alternating 1 minute at 60% Ẇ_max and 1 minute at 90% Ẇ_max (60/90); (c) 30-minute continuous cycle alternating 1 minute at 55% Ẇ_max and 4 minutes at 80% Ẇ_max (55/80); (d) Squatting 8 sets of 10 repetitions at their 10RM. There was a 90-second rest between each set (RESIST); and (e) Resting trial. Completed at the same time of the day as the exercise trials (REST).

Blood Collection and Analysis

At the beginning of FAT, 60/90, 55/80, and REST, a cannula (18GA Venflon, Becton, Dickinson and Co., Oxford, United Kingdom) was inserted into an antecubital vein in a supine position. The subject moved to an upright sitting position and rested in this position for 10 minutes before blood collection. Blood samples were collected into 4.5 mL tri-potassium ethylenediaminetetraacetic acid (K₃EDTA) Vacutainers (Becton, Dickinson and Co.) at pre-exercise, 0, 10, 20, 30, 40, 50, and 60 minutes postexercise. During the RESIST trial, blood samples were collected by venepuncture from an antecubital vein at pre-exercise, 0 and 10 minutes postexercise into a K₃EDTA Vacutainer. Venepuncture sampling was used for this trial to allow arm flexibility so that the subjects could complete the squatting exercise without undue arm discomfort. All blood samples were immediately centrifuged at 1,500g for 10 minutes, and the plasma was divided into aliquots and stored at −20°C until the analysis. Plasma cortisol and plasma testosterone concentrations were determined using commercially available enzyme-linked immunosorbent assay (ELISA) kits (DRG Instruments, Marburg, Germany). The sensitivity of the kits was 6.9 nmol·L⁻¹ (plasma cortisol) and 0.29 nmol·L⁻¹ (plasma testosterone). The mean intraassay coefficients of variation were 2.8 and 2.2% for cortisol and testosterone, respectively. The mean interassay coefficients of variation were 2.7 and 3.4% for cortisol and testosterone, respectively.

Saliva Collection and Analysis

The participants drank water ad libitum during the main trials; however, to avoid the possibility of diluting the saliva sample, they were not permitted to drink during the 10 minutes before saliva sampling. Saliva samples were collected pre-exercise and 0, 10, 20, 30, 40, 50, and 60 minutes postexercise in all trials. Subjects were seated throughout and gave an unstimulated saliva sample by dribbling into a sterile vial (Sterilin, Caerphily, United Kingdom) with eyes open, head tilted slightly forward, and making minimal orofacial movement. Minimum collection time was 2 minutes for each subject to allow for collection of a sufficient sample volume. All saliva samples were immediately divided into aliquots and stored at −20°C until further analysis. The salivary cortisol and salivary testosterone concentrations were determined using commercially available ELISA kits (Salimetrics, State College, PA, USA). The sensitivity of the kits were 0.08 nmol·L⁻¹ (salivary cortisol) and <3.46 pmol·L⁻¹ (salivary testosterone). The mean intraassay coefficients of variation were 3.8 and 2.7% for cortisol and testosterone, respectively. The mean interassay coefficients of variation were 3.3 and 2.7% for cortisol and testosterone, respectively.

Statistical Analyses

All data in the text and tables are presented as mean values and SDs. All data in the figures are presented as mean values and SEM to avoid clutter. Data were checked for normality, homogeneity of variance, and sphericity before statistical analysis. If a data set was not normally distributed, logarithmic transformation was performed on the data. Logarithmic transformation data sets were used for the analysis of the responses of salivary and plasma cortisol, salivary testosterone, and plasma and salivary C/T ratio in the RESIST trial. In addition logarithmic data sets were also used in the analysis of the response of the salivary C/T ratio in the FAT trial. The HR, RPE, and area under the curve (AUC) responses during exercise and the hormonal response to the individual exercise and resting trials were analyzed. When the data sets were parametric, a 1-way repeated-measures analysis of variance (ANOVA) was completed. Significant differences were assessed using Student's paired samples t-tests with Holm-Bonferroni adjustments for multiple comparisons. Nonparametric data were compared using Friedmans ANOVA, and post hoc analysis was completed using Wilcoxon signed-rank tests. Statistical significance was set at p ≤ 0.05.

Results

Heart Rate, Ratings of Perceived Exertion, and Time to Exhaustion

The mean HR response to RESIST was lower than all other exercise trials (p < 0.05). The mean RPE scores were not different in all exercise trials (Table 2). The mean time to exhaustion in the FAT trial was 32.7 ± 12.8 minutes.

Table 2: Mean HR, RPE, and total work completed during FAT, 60/90, 55/80, and RESIST trials.*†

Hormonal Responses during REST

Salivary and plasma cortisol and testosterone concentration did not change during the resting trial (Figure 1A, B).

Figure 1: A) Plasma cortisol response to FAT, 60/90, 55/80, RESIST and REST (mean ± SEM). Different from Pre-Ex values: a. FAT; b. 60/90; c. 55/80; d. RESIST; and e. REST. B) Saliva cortisol response to FAT, 60/90, 55/80, RESIST and REST (mean ± SEM). Different from Pre-Ex values: a. FAT; b. 60/90; c. 55/80; d. RESIST; and e. REST. (RESIST = squatting 8 sets of 10 repetitions at 10 repetition maximum; FAT = continuous cycle to volitional fatigue at 75% Ẇ max; REST = resting trial.)

Hormonal Responses to Exercise

Cortisol

Plasma cortisol increased from pre-exercise to 0 minutes postexercise in response to FAT, 60/90, and 55/80 and at 10 minutes postexercise only in the RESIST trial (samples were only collected to 10 minutes postexercise in this trial) (p < 0.01). Plasma cortisol remained elevated throughout the 60 minutes postexercise period in the 55/80 trial. The plasma cortisol returned to pre-exercise levels at 30 and 50 minutes in the 60/90 and FAT trials, respectively (Figure 1A). Salivary cortisol increased from pre-exercise values to postexercise at 0 minutes (FAT) and 10 minutes (55/80) only (p < 0.01). Salivary cortisol remained elevated throughout the remainder of the 60 minutes postexercise period in 55/80 and returned to pre-exercise values at 40 minutes postexercise in the FAT trial (Figure 1B).

Peak concentrations were observed at 10-20 minutes postexercise for plasma cortisol and ∼30 minutes postexercise for salivary cortisol. The AUC analysis indicated that the 3 cycling trials induced a larger salivary cortisol response compared with the RESIST trial (p < 0.01). No differences between trials were found for plasma cortisol AUC.

Testosterone

Plasma testosterone increased at 0 minutes postexercise from pre-exercise values in all exercise trials excluding 55/80 (p < 0.01). The values remained elevated throughout the 10 minutes postexercise period in the RESIST trial and until a return to pre-exercise levels at 30 minutes, and 10 minutes postexercise in the FAT and 60/90 trials, respectively (Figure 2A). Salivary testosterone increased postexercise in all cycle trials at 10 minutes postexercise and returned to pre-exercise values at 30 minutes postexercise in the FAT and 60/90 trials and at 40 minutes postexercise in the 55/80 trial. The RESIST trial showed an increase from pre-exercise at 0 minutes until a return to pre-exercise values at 30 minutes postexercise (Figure 2B) (p < 0.01).

Figure 2: A) Plasma testosterone response to FAT, 60/90, 55/80, RESIST and REST (mean ± SEM). Different from Pre-Ex: a. FAT; b. 60/90; c. 55/80; d. RESIST; and e. REST. B). Salivary testosterone response to FAT, 60/90, 55/80, RESIST and REST (mean ± SEM). Different from Pre-Ex: a. FAT; b. 60/90; c. 55/80; d. RESIST; and e. REST. (RESIST = squatting 8 sets of 10 repetitions at 10 repetition maximum; FAT = continuous cycle to volitional fatigue at 75% Ẇ max; REST = resting trial.)

Peak plasma and salivary testosterone values occurred at 0 and ∼10 minutes postexercise, respectively. The AUC analysis indicated that the salivary testosterone response to the exercise was higher in the 60/90 and 55/80 trials compared with the FAT and RESIST trials (p < 0.05). The AUC analysis indicated that the plasma testosterone response to exercise did not differ between trials.

Cortisol/Testosterone Ratio

The Plasma C/T ratio increased at 10 minutes postexercise (FAT) and at 20 minutes postexercise (55/80) compared with pre-exercise and remained above the pre-exercise levels until a return at 60 minutes postexercise (FAT) and at 40 minutes postexercise (55/80). No change was found for the plasma C/T ratio for 60/90 and RESIST trials (Figure 3A).

Figure 3: A) Plasma C/T response to FAT, 60/90, 55/80, RESIST and REST (mean ± SEM). Different from Pre-Ex: a. FAT; b. 60/90; c. 55/80; d. RESIST; and e. REST. B) Salivary C/T response to FAT, 60/90, 55/80, RESIST and REST. Different from Pre-Ex: a. FAT; b. 60/90; c. 55/80; d. RESIST; e. REST. (RESIST = squatting 8 sets of 10 repetitions at 10 repetition maximum; FAT = continuous cycle to volitional fatigue at 75% Ẇ max; REST = resting trial.)

The salivary C/T ratio increased from pre-exercise at 20 and 30 minutes postexercise in FAT and 55/80 trials, respectively (p < 0.05). The salivary C/T remained above the pre-exercise levels for the remainder of the 60 minutes postexercise period in both trials. The salivary C/T ratio remained unchanged in the 60/90 and RESIST trials (Figure 3B).

The peak C/T ratio was observed in plasma at ∼10 minutes postexercise and in saliva at ∼30-40-minutes postexercise.

Plasma and Salivary Hormone Correlations

Positive correlations were found between the peak postexercise salivary and plasma cortisol values for each exercise protocol (p < 0.05) (Figures 4A-D). Peak postexercise salivary and plasma testosterone values did not correlate (Figures 5A-D).

Figure 4: Relationship between peak postexercise saliva and plasma cortisol concentrations in (A) FAT, (B) 60/90, (C) 55/80 and (D) RESIST. (RESIST = squatting 8 sets of 10 repetitions at 10 repetition maximum; FAT = continuous cycle to volitional fatigue at 75% Ẇ max.)

Figure 5: Relationship between peak postexercise saliva and plasma testosterone concentrations in (A) FAT, (B) 60/90, (C) 55/80 and (D) RESIST. (RESIST = squatting 8 sets of 10 repetitions at 10 repetition maximum; FAT = continuous cycle to volitional fatigue at 75% Ẇ max.)

Discussion

The main aim of this study was to find a short duration, high-intensity exercise protocol that would cause a robust elevation of plasma and salivary cortisol and testosterone. All exercise trials tested caused an increase in plasma cortisol with the 55/80 trial showing a prolonged postexercise elevation compared with the other trials. Salivary cortisol only increased in the postexercise period in the FAT and 55/80 trials. The 55/80 trial also elicited a prolonged salivary cortisol elevation compared with the FAT trial. Salivary testosterone increased above pre-exercise values in all the exercise trials but AUC analysis indicated that the 60/90 and 55/80 trials showed a larger increase compared with the other trials. These results suggest that the 55/80 trial causes robust elevations of plasma and salivary cortisol and salivary testosterone.

The pre- to postexercise increases in the plasma and salivary cortisol concentrations reported in this study correspond to previous research on hormonal responses to high-intensity exercise (^4,9,18,20). In this study, the levels of both salivary and plasma cortisol once elevated remained higher than pre-exercise levels during the 60 minutes postexercise period in the 55/80 trial. This delayed return of cortisol to basal values after exercise has been reported by others. Elloumi et al. (⁵) reported that salivary cortisol increased from the beginning to immediately after a rugby match, and it remained higher than pre-exercise levels up to 4 hours after the match. Initially, it was hypothesized that the FAT and 60/90 trials would induce an elevation of salivary and plasma cortisol for a similar duration as in the 55/80 trial. An unexpectedly large variation in exercise completion time was seen in the FAT trial. This was probably the reason why the FAT trial did not show consistent elevations of salivary and plasma cortisol. The 60/90 trial required the subject to complete the same work as completed in the 55/80 trial. Despite completing the same total amount of work, it is likely that the 60/90 trial was not exercising the subjects at a high enough intensity for a long enough period to cause changes in the salivary and plasma cortisol similar to the 55/80 trial. Exercising above 60% V̇O₂max for at least 20 minutes is suggested as the threshold for changes in cortisol concentrations to occur. The 60/90 trial exercised the subjects for 15 minutes only above the 60% V̇O₂max threshold level.

This study found increases in salivary and plasma testosterone from pre- to postexercise in all exercise protocols except that no changes were found in plasma testosterone in the 55/80 trial. In agreement with our findings, Coolomp et al. (³) also reported increases in plasma-free testosterone from pre-exercise to 30 minutes of a cycle at 75% V̇O₂max. It is possible that no change in the plasma testosterone was seen after the 55/80 because of missed blood sample collections. The immediately postexercise blood sample collections for 4 subjects were missed because of their veins shutting down as a result of the exercise. On all occasions, the vein returned to normal at the 10 minutes postexercise time point. Because the peak plasma testosterone concentration values occurred at 0 minutes postexercise, it means that the peak values were missed in 4 of 10 subjects.

An increase in plasma and salivary cortisol and testosterone was also expected in the resistance exercise trial. This study found an increase in salivary and plasma testosterone and plasma cortisol in response to the RESIST trial. No changes were found in salivary cortisol after this trial. Previous studies have reported findings corresponding to the findings in the blood in this study. McCaulley et al. (¹³) studied serum cortisol and testosterone responses to 3 different resistance bouts. In response to their hypertrophy scheme (4 sets, 10 repetitions at 75% 1 RM), which is the closest to our resistance protocol, they reported a 32 and 12% increase from pre- to postexercise in serum testosterone and cortisol, respectively. Häkkinen and Pakarinen (⁶) also reported a 21 and 148% increase from pre- to postexercise in serum-free testosterone and serum cortisol, respectively, in response to a squatting protocol of 10 sets of 10 repetitions of 70% 1RM.

The robust and prolonged elevation of salivary and plasma cortisol and salivary testosterone after the 55/80 trial suggests that this exercise bout could be a useful diagnostic tool when measuring a possible dysfunction of cortisol and testosterone response to exercise in athletes suspected to be overreached. This study suggests that consistent increases in cortisol and testosterone will occur after the 55/80 trial when in a normal trained state. It has been shown by previous studies that hormone levels can change when the athlete is in a state of overreaching. Therefore, any dysfunction of the hormones cortisol and testosterone should be highlighted by completing the 55/80 trial.

It was hypothesized that the saliva hormonal measurements would correlate with plasma measurements. This study reported positive correlations of peak postexercise salivary and plasma cortisol. This finding is in agreement with Port (¹⁷) who found positive correlations between serum and saliva cortisol in response to submaximal loads in an incremental cycle test. Cadore et al. (¹) also found a positive correlation between serum and salivary cortisol before and after a resistance training session. The positive correlations found in our study may be related to the hormone sampling time postexercise. Only peak postexercise saliva and plasma cortisol samples showed positive correlations. These peaks occurred at different time points postexercise in plasma and saliva. Postexercise cortisol peaks occurred at 10-20 minutes for plasma and ∼30 minutes for saliva. The peaks of plasma and salivary testosterone occurred at 0 minutes and between 0-10 minutes postexercise, respectively. The slight time delay would allow for plasma cortisol changes to be better reflected in the saliva samples. This study found no correlations between plasma and salivary testosterone. This agrees with the findings of Cadore et al. (¹) who reported no correlation between pre- and postexercise salivary and serum free testosterone. However, it is possible the peak values did not correlate because both the salivary and plasma peak values were missed. The salivary and plasma testosterone peaks occur quicker than cortisol as the peaks occur within the first 10 minutes postexercise. Correlations might have been found if more samples had been collected in the first 10-minute period.

In conclusion, the FAT trial range of times to exhaustion was so large that a great interindividual variation in both cortisol and testosterone concentration in response to exercise was observed. The 55/80 and 60/90 trials induced similar increases in both hormones, but the 55/80 postexercise values remained above pre-exercise values for longer than 60/90. The RESIST trial induced a smaller increase in postexercise salivary cortisol and testosterone compared with the 55/80 and 60/90 trials. These findings indicate that the 55/80 cycling protocol causes robust elevations in the hormones cortisol and testosterone when in a period of normal training, whereas the other trials tested did not. This exercise protocol could potentially be used to highlight any changes in cortisol and testosterone that may occur with overreaching.

Practical Applications

The aim of this study was to find a relatively short duration, high-intensity exercise bout that would induce a consistent and significant increase in the hormones cortisol and testosterone. Our data indicate that the 55/80 cycle protocol could be a useful indicator of the hormonal response to exercise in athletes. It has been suggested that the response of the hormones cortisol and testosterone will change in response to exercise when an athlete enters into a stage of overreaching. Therefore, the 55/80 cycle protocol could be used as a test to highlight any changes in cortisol and testosterone response to exercise. However, further study is needed on the hormonal response to this protocol before and after an intensified period of training. This will give a better indication of the usefulness of this protocol as a tool in the diagnosis of overreaching.