The traditional hypothesis for the ergogenic action of glucocorticoids is that this class of drug induces central nervous system excitation and euphoria at rest, and increases blood glucose and energy mobilisation.1 2 Consequently systemic administration of this pharmacological class is banned by the World Antidoping Agency (WADA), but currently the ban is only on use during competition––that is, permitting detection of acute abuse. However, the repercussions of short-term glucocorticoid administration during exercise in men have not been well documented, and the performance results are inconsistent. Indeed, of the published studies on short-term administration of glucocorticoids, only one study has focused on the effects of short-term dexamethasone intake3 during maximum exercise, which did not show any ergogenic effect of the treatment. In contrast, we e recently showed that short-term therapeutic prednisolone intake4 significantly improves performance in healthy men during submaximum exercise. The concomitant alterations in the hormonal and metabolic exercise parameters analysed showed that short-term administration of this drug had both central and peripheral effects. Indeed, we found a decrease in adrenocorticotrophic hormone (ACTH) and dehydroepiandrosterone (DHEA) concentrations, as expected, with a parallel decrease in both growth hormone (GH) and prolactin secretion and an increase in blood glucose concentrations. However, this previous study was conducted on recreationally trained men, and the effects of glucocorticoid treatment associated with strenuous physical training were not investigated.

Therefore, in a first attempt to determine whether these preliminary results could be extrapolated to elite athletes, the purpose of the present study was to investigate the influence of short-term prednisolone administration (7 days) delivered orally at therapeutic dosage (60 mg/day) combined with a standardised training regimen (2 hours/day) on performance during a submaximum exercise (70–75% VO_2max) in healthy recreationally trained volunteers. Furthermore, in order to contribute to a wider knowledge of glucocorticoid action mechanisms during exercise, a larger number of pituitary (ACTH, prolactin, thyroid-stimulating hormone (TSH), growth hormone (GH), luteinising hormone (LH)) and peripheral hormones (DHEA, insulin, testosterone) and metabolic parameters (pH, bicarbonates, glucose, lactate) were monitored in the present study.

METHODS

The protocol was approved by the Ethics Committee of Tours Hospital, and all subjects gave their informed consent after experimental procedures and possible risks had been explained both verbally and in writing.

Subjects

Eight recreational male athletes, actively cycling and/or running 2–3 times/week for at least 3 years, were chosen for this experiment. The mean (SE) age of the group was 21.3 (0.5) years, and body mass was 68.6 (3.1) kg. None of the subjects was taking any medication or had a family history of any endocrine disorder.

Procedure

All subjects had previously participated in physical exercise experiments in the laboratory. In the month before the study, an incremental test for maximum O₂ uptake (VO_2peak) was conducted on a Monark cycle ergometer (model 918E; Monark-Crescent AB, Varberg, Sweden) to select a power output in watts eliciting 70–75% VO_2peak (W70–75), following a standard laboratory procedure. Mean VO₂ peak was 54.0 (1.3) ml/kg/min. Subjects were asked to abstain from caffeine and alcohol 24 hours before each trial.

Treatment

This double-blind, randomised crossover study consisted of two 1-week treatments for each subject separated by a 3-week drug-free washout period, using lactose (placebo) and prednisolone. The placebo and prednisolone (Hydrocortancyl 5 mg tablet; Aventis Laboratory, Paris, France) were packaged in identical capsules. During the experimental periods (7 days), the subjects took either prednisolone (60 mg) or placebo (lactose) at 07:00–08:00 daily at home. When questioned after the completion of the study if they knew which the two treatments they received first, two mentioned some overexcitement after prednisolone treatment, whereas the six others had not noticed any difference. No significant changes in body weight were measured at the end of experiment.

Trials to exhaustion were performed just before each treatment (placebo 1 and prednisolone 1) without ingestion of capsules and on the last day of each weekly treatment (placebo 2 and prednisolone 2), 3 hours after ingestion of the final capsule.

Training

To standardise training effects, each participant maintained a training diary of duration, mode and intensity of activity. Each day during the 2-week study, participants also completed treatment-specific training sessions consisting of a 60 minutes of ergometer cycling at 50% VO_2max followed by interval training, 2×15 min at 70% VO_2max, with a 5-minute recovery in between, and 6×1 minute at 90% VO_2max with a 1-minute recovery in between.

Protocol

Trials were held at the same time of day (10:00–11:00) for each subject in order to prevent diurnal variations in hormonal responses. On the testing days, subjects reported to the laboratory at 09:00–10:00, without ingesting capsules (placebo 1 and prednisolone 1) or two hours after ingesting capsules containing either placebo or prednisolone (placebo 2 and prednisolone 2) and 1 hour after ingesting a small meal, which was identical for each trial. Dietary consistency (about 500 kcal) was confirmed through self-reported diet records and questioning before each trial.

After insertion of a catheter into a superficial forearm vein, subjects rested (30 minutes), and a resting blood sample was taken just before the start of exercising. At 10:00–11:00, they exercised at W70–75 until exhaustion. Blood samples were taken every 10 minute during the first 30 minutes of exercise and the first 20 minutes of recovery. No samples were taken between 30 minutes and exhaustion, so that subjects could not count samples as a crude time device. Exhaustion was determined by the investigators when cadence could no longer be maintained at a rate of 90% of the subject’s set rate, i.e., 80 (3) rpm. Water was given ad libitum during exercise. Subjects did not have access to any indication of time after the initial 30 minute sampling period during the exercise, and results were disclosed only on completion of the entire study.

Analysis

Blood samples (8 ml) were immediately transferred to different tubes. A 2-ml aliquot was placed in a chilled sodium heparinised tube for insulin and GH determination, 3 ml were transferred to a nontreated tube for prolactin, TSH, free testosterone and LH analyses, and the remaining 3 ml were placed in a chilled EDTA–aprotinin tube for ACTH and DHEA analysis. All tubes were promptly spun in a centrifuge held at 4°C for 10 minutes at 2000 g, and stored at −72°C until assays.

Haematocrit, bicarbonate, pH, blood glucose and lactate were immediately measured (OMNI, Neuilly, France). Haemoconcentration occurred in all the exercise samples, but there was no difference between trials.

ELISA was used for most of the analyses: kits from Biomerica, Newport Beach, USA (ACTH); DSL, Marburg, Germany (GH) and Bioadvance, Paris, France (DHEA, TSH, LH, prolactin, testosterone and insulin) were used. All assays were made in duplicate. Coefficients of variation (interassay and intra-assay) for all parameters were always <10%.

Statistics

Data are presented as means (SEM). A specific test for crossover trials was used to determine whether there were any significant differences in performance between before and the end of each treatment. Differences in all the measured hormonal and metabolic variables were statistically analysed for time and treatment effects using a two-way analysis of variance. When a significant F ratio was observed, a Newman–Keuls multiple comparison test was performed to determine the location of the differences. Statistical significance was set at p<0.05.

RESULTS

Performance responses

No rank effect was seen (fig 1). There was no significant difference in times to exhaustion between placebo 1, placebo 2 and prednisolone 1 (mean (SEM) 50.4 (6.2), 64.0 (9.1) and 56.1 (9.1) minutes, respectively). Time of cycling to exhaustion was significantly (p<0.05) increased after the prednisolone treatment (prednisolone 2 was 107.0 (20.7) minutes).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Individual performance times of cycling to exhaustion before (placebo 1, prednisolone 1) and after (placebo 2, prednisolone 2) treatment.

Hormonal and metabolic parameters

ACTH, PRL, GH, TSH, LH, DHEA, testosterone, insulin

There was no significant difference in any of these measured parameters after the week of training with placebo 2 versus placebo 1 or prednisolone 1 (figs 2, 3). Exercise induced a significant and similar increase in basal ACTH, PRL, GH, DHEA and testosterone without any significant change in TSH and in LH, before each treatment (placebo 1, prednisolone 1) and after placebo treatment (placebo 2). In parallel, exercise induced a significant decrease in insulin basal values after 10 minutes until the end of the experiment in the placebo 1 and 2 and the prednisolone 1 trial.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 (A) Adrenocorticotrophic hormone, (B) luteinising hormone (mIU/l), (C) dehydroepiandrosterone (ng/ml) and (D) free testosterone (pg/ml) responses. Data are mean (SE) at rest (0), during cycling to exhaustion (10, 20, 30, End) and recovery (R10, R20) before (placebo 1, prednisolone 1) and after (placebo 2, prednisolone 2) treatment. *Significant difference between placebo and prednisolone (p<0.05). Start of significant difference between rest and exercise †before and ‡after placebo treatment and §before (p<0.05) and ¶after prednisolone treatment (all p<0.05).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 (A) Growth hormone (ng/ml), (B) prolactin (ng/ml), (C) thyroid-stimulating hormone (mIU/ml), and (D) insulin (mIU/ml) responses. Data are mean (SE) at rest (0), during cycling to exhaustion (10, 20, 30, End) and recovery (R10, R20) before (placebo 1, prednisolone 1) and after (placebo 2, prednisolone 2) treatment. *Significant difference between placebo and prednisolone (p<0.05). Start of significant difference between rest and exercise †before and ‡after placebo treatment and §before (p<0.05) and ¶after prednisolone treatment (all p<0.05).

ACTH, DHEA and testosterone values were significantly decreased with prednisolone treatment compared with placebo (p<0.05) at rest, while exercising and in recovery. For testosterone but not for ACTH and DHEA, exercise induced a significant increase in basal concentrations at exhaustion (p<0.05) with the prednisolone treatment.

No change in PRL, GH, TSH, LH and insulin resting values was found after the prednisolone treatment compared with the other experimental conditions. However, a decrease in PRL, GH and TSH was found in prednisolone 2 respectively after 20 minute, 30 minute of exercise and at exhaustion (p<0.05). Exercise in prednisolone 2 significantly increased PRL and GH concentrations and decreased insulin concentrations without any significant change in TSH and in LH.

Bicarbonates, pH, glucose, lactose

There was no significant difference in any of these parameters during rest, exercise and recovery between placebo 1 and 2 and prednisolone 1 (fig 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4 (A) Blood glucose, (B) lactate, (C) bicarbonates and (D) pH responses. Data are mean (SE) at rest (0), during cycling to exhaustion (10, 20, 30, End) and recovery (R10, R20) before (placebo 1, prednisolone 1) and after (placebo 2, prednisolone 2) treatment. *Significant difference between placebo and prednisolone (p<0.05). Start of significant difference between rest and exercise †before and ‡after placebo treatment and §before (p<0.05) and ¶after prednisolone treatment (all p<0.05).

ANOVA did not reveal any significant treatment effect in pH and bicarbonates. Exercise significantly decreased bicarbonates after 10 minutes (p<0.05). Blood glucose concentration was always significantly increased in prednisolone 2 versus placebo 1, placebo 2 and prednisolone 1 (p<0.05). During exercise, glucose level remained constant before and after both placebo and prednisolone treatments.

Basal lactate concentrations were quite similar in all the trials. Exercise induced a significant increase in each experimental condition (p<0.05), without any significant treatment effect.

DISCUSSION

The present study is unique in its examination of the influence of short-term glucocorticoid treatment combined with a strenuous physical activity on endurance and on hormonal and metabolic responses in healthy volunteers. Under the placebo condition, a week of training did not alter the basic responses in any measured parameters. Training associated with glucocorticoid treatment resulted in a marked improvement in endurance performance and prednisolone-induced changes in ACTH, DHEA, PRL, GH, TSH, free testosterone and blood glucose.

A review of the literature found almost no studies on this subject, and inconsistency in the ergogenic effect of glucocorticoid administration in humans that may be attributable to the mode of administration (acute or short-term), the dose used and the intensity of the exercise chosen. Soetens et al5 did not find any significant increase in maximum performance with a 1-mg ACTH injection in professional cyclists. Similarly, we showed previously6 7 that an acute therapeutic administration of oral prednisolone (20 mg) does not improve the time of cycling until exhaustion during submaximum exercise in healthy, moderately trained, male volunteers, despite a probable increase in lipid oxidation and a decrease in carbohydrate oxidation.8 Of the published studies on short-term administration of glucocorticoids, only one focused on the effects of short-term dexamethasone intake3 during maximum exercise, and found no ergogenic effect due to the treatment. However, we recently showed that short-term therapeutic prednisolone intake (60 mg per day for 7 days), significantly improves performance in healthy men during submaximum exercise (70–75% VO_2max), with marked hormonal changes, results that contrast with those achieved after acute intake.4 Associating short-term glucocorticoid treatment with strenuous training has not, to our knowledge, been investigated previously, and a study on the repercussions of glucocorticoid intake in trained athletes is warranted. In the present work, 1 week of strenuous training for 2 hours per day did not significantly modify endurance performance, as shown by the similar times to exhaustion obtained before and after the placebo treatment. However, the longer exhaustion times in the glucocorticoid trials in all except one subject indicates that the addition of strenuous training in conjunction with the prednisolone treatment did not have any opposing action on the glucocorticoid–ergogenic effect. Moreover, the increase in endurance cycling can be positively related to the physical status of the subjects. Indeed, we obtained an average increase in cycling time of about 80% in the present study with corticoid versus placebo, compared with an average increase of 54% in our previous study without training.4 Moreover, as shown in fig 1, the greatest increase in time to exhaustion with prednisolone was obtained in the subject performing the best trial with placebo. Further studies are of course needed to verify this hypothesis and to determine whether elite male athletes are more sensitive to the ergogenic effect of glucocorticoids during endurance exercise. If the ergogenic effects of short-term glucocorticoid oral intake are confirmed in male and female athletes, it would be necessary to prohibit systemic use of this class of drugs at all times (during and outside competitions) and not just in during competition as in the current WADA legislation.

Decreases in levels of ACTH and DHEA after exercise are expected and are generally found after both acute and short-term systemic administration.3 9^–11 Similarly, we found a marked decrease in basal ACTH and DHEA concentration after our short-term prednisolone treatment combined with training compared with values obtained before the treatments and after placebo administration. Under these last three conditions, we found a similar gradual increase in ACTH and in DHEA during exercise and recovery compared with rest values as described in the literature, but these ACTH and DHEA increases seemed to be completely inhibited by prednisolone administration. These results confirmed our previous data,4 and it seems likely that short-term treatment of glucocorticoid even at the therapeutic level induces complete inactivation of the hypothalamic–pituitary–adrenal axis during exercise, irrespective of the subjects’ training status.

Whereas the data after acute intake seem to conflict with either an increase or no modification,11^–13 short-term systemic glucocorticoid administration has been shown to decrease exercise GH concentrations.4 We did not find any significant difference in GH concentrations between the four conditions at rest with, according to the literature, a significant increase in the basal concentrations during exercise. However, as previously reported, we found significantly lower values of exercise GH concentrations after prednisolone treatment in the present study, starting after 30 minutes of exercise, probably due to a steroid-mediated increase in hypothalamic somatostatin tone.14 It seems that the addition of strenuous training did not affect the blunted hormonal response linked to the drug administration. In parallel, we measured prolactin in the peripheral blood circulation, and found, as in our previous studies,4 significantly lower prolactin concentrations after the short-term intake of prednisolone. The mechanism(s) by which glucocorticoids decrease(s) exercise prolactin levels remain(s) unknown, but it may be suggested that short-term administration of prednisolone induced alterations in either central brain serotonin or dopaminergic activity, possibly delaying the onset of fatigue,15^–18 but this was once again regardless of the subjects’ training status. Further studies using an animal model are necessary to clarify the mechanism(s) implicated and to verify the significance of this prolactin change in the improvement in performance.

As indicated by the alterations in exercise GH and prolactin concentrations, it seems that the effects of glucocorticoids on the hormone system are not restricted to the hypothalamic–pituitary–adrenal axis.19 However, to our knowledge, there are no data on the repercussions of glucocorticoids during submaximum exercise on other pituitary hormones such as TSH and LH, and on their subordinate hormonal responses (thyroxine and testosterone). There are only few data on the effects of glucocorticoids on TSH levels, but these suggest that physiological concentrations of glucocorticoids modulate the pituitary–thyroid axis.20 Indeed, it was shown that glucocorticoid administration is associated with reduced basal TSH levels and a blunted TSH response to TRH, despite thyroid hormone levels within the normal range,21 with a probable involvement of enhanced hypothalamic somatostatinergic and dopaminergic inhibitory activities.21 In the present study, we unfortunately did not measure thyroid hormone but we found a significant decrease in TSH concentration at exhaustion after the prednisolone treatment compared with the other conditions.

However, LH concentrations seemed to be unaltered by prednisolone treatment at any time point during the study, but we found a significant decrease in free testosterone concentrations with prednisolone throughout the study, which seems to be in accordance with the literature. Indeed, decrease in testosterone by glucocorticoids has never been described during exercise but it was previously shown in vitro that dexamethasone and other synthetic glucocorticoids may exert a direct inhibitory effect on testosterone production by purified porcine immature Leydig cells.22 Other in vitro studies have shown a direct inhibitory effect of glucocorticoids on testicular LH receptor content and steroidogenesis, suggesting that adrenal glucocorticoids may regulate testis functions.23 Moreover, glucocorticoid treatment seems to aggravate hypogonadism in men with chronic obstructive pulmonary disease,24 with a negative correlation between corticosteroid dosage and serum testosterone level.25 In view of the significant decrease in free testosterone induced by prednisolone intake in the present study, it seems necessary to investigate the repercussions of short-term glucocorticoid intake on the other anabolic hormones and metabolites analysed during a doping control.

Most of the previous studies5 11 26 using therapeutic glucocorticoid administration reported hyperglycaemia at rest after both acute and short-term intake, generally coupled with hyperinsulinaemia after chronic administration.4 27 In agreement with the literature, we obtained significant basal hyperglycaemia after the prednisolone treatment, which persisted throughout the study. However, contrary to our previous study4 on short-term glucocorticoid intake without training, we found in the current study no significant increase in either basal or exercise insulinaemia or lactatemia. It may be hypothesised that the lack of significance could be linked to the stronger power of the statistical test used here. Similarly, even though moderate glucocorticoid alkalosis has already been described,28 pH and bicarbonate concentrations in prednisolone 2 were not significantly higher compared with other conditions.

CONCLUSION

The complexity of the hormonal and metabolic responses to short-term glucocorticoid administration during exercise makes dissociation of the possible causal effects on performance difficult. Subsequent research is therefore needed to determine: (1) whether peripheral versus centrally mediated effects are significant in the glucocorticoid-mediated ergogenic response; (2) whether these results obtained in male subjects are gender-dependent and applicable to elite athletes, and if so, then by inference, whether current WADA legislation needs to be changed.