Acta Physiol Scand 1991, 141, 373-378
ADONIS 000 167729100055W
Carbohydrate ingestion during prolonged running exercise results in an increase of serum cortisol and decrease of gonadotrophins T. J . V A S A N K A R 1 , U. M. K U J A L A , T.T. V I L J A N E N and I. T. H U H T A N I E M I Department of Physiology, University of Turku and Paavo Nurmi Centre, Turku, Finland VASANKARI,T. J., KUJALA,U. M., VILJANEN, T. T. & HUHTANIEMI I. T. 1991. Carbohydrate ingestion during prolonged running exercise results in an increase of serum cortisol and decrease of gonadotrophins. Acta Physiol Scand 141, 373-378. Received 25 March 1990 accepted 29 September 1990. ISSN 0001-6772. Department of Physiology, University of Turku, Finland. We tested the hypothesis that improved availability of energy through carbohydrate ingestion could counteract the documented suppression of pituitary-gonadal function during prolonged exercise. Nine trained males repeated twice a 36 kilometre running exercise with two weeks interval. During the tests the subjects ingested in a randomized single-blind fashion a total of 1050 millilitres of carbohydrate and placebo solutions. The total amount of ingested carbohydrate was 105 grammes. Venous blood samples were taken before the exercise (sample A), immediately after exercise (B), and 2 hours later (C). In the B samples plasma glucose was 14% higher ( P < 0.01) and serum cortisol 13% higher ( P < 0.05) in the carbohydrate than in the control trial. In contrast, the level of LH was 18% (P< 0.05) and that of FSH 11yo lower (P< 0.05) in the carbohydrate than in the placebo trial. Serum testosterone concentration did not differ between the treatment groups. We conclude that carbohydrate ingestion does not counteract the exercise-associated suppression of gonadotrophin secretion, but results in paradoxical increase of serum cortisol and decrease of LH and FSH, in comparison to placebo-treated controls. The higher cortisol level in carbohydrate group may be secondary to higher insulin levels. Key words :cortisol, exercise, FSH, glucose, LH, testosterone
Prolonged exercise can lead to hypoglycaemia (Felig et al. 1982). A 4-day diet low in carbohydrate has been shown to decrease the time to exhaustion in exercise when compared to a diet high in carbohydrate (Galbo et a1 1979). T h e effects of glucose ingestion/infusion on perceived exhaustion and various hormone levels during exercise are greater if fasting precedes the exercise (Galbo et al. 1981). Serum testosterone concentration has been shown to decrease after prolonged physical exercise (Kreuz et al. 1972, Dessypris et al. 1976). T h e exercise induced suppression of serum testosterone is associated Correspondence: Tommi Vasankari, Department of Physiology, University of Turku, SF-20520 Turku, Finland
with two effects : suppressed GnRH stimulation of gonadotrophin release during exercise and decreased testicular capacity to secret testosterone during the recovery period (Kujala et al. 1990). T h e reported gonadotrophin responses during exercise have been variable (Dessypris et al. 1976, Kuoppasalmi et al. 1980, Schmid et al. 1982, Kuusi et al. 1984, Mougin et al. 1988). Long-term fasting has been reported to decrease serum testosterone levels, and after refeeding these levels were normalized (Klibanski et al. 1981). In one study, the decrease of testosterone in prolonged physical strain and sleep deprivation could not be prevented with a high carbohydrate diet (Opstad et al. 1982). In endurance athletes, the repeated exerciseassociated suppression of testosterone may
373
374
T.3. Vasankari et al.
Table 1.Mean (range) of some characteristics of the subjects Age (years) Height (cm) Weight (kg) Body fat (yo) Maximal oxygen uptake (ml kg-' min-') Anaerobic threshold (ml kg-l min-l) Regular training background (years) Training during the preceding three years (1000 km) Training during the preceding year (1000 km) Training during the week before placebo trial (km) Training during the week before carbohydrate trial (km) Training during the day before placebo trial (km) Training during the day before carbohydrate trial (km)
cumulatively lead to androgen deficiency, which may be unfavourable, when training-induced improvement in performance is expected. Although the nutritional status in long-term has pronouilced effects on pituitary-gonadal function (Klibanski et al. 1981), it is not known whether the availability of energy during exercise (e.g. through ingestion of carbohydrates) has effect on the hypothalamus-pituitary-gonadal axis. Therefore, we challenged a group of healthy trained males during prolonged exercise with a high energy carbohydrate solution to study its effects on pituitary-gonadal and adrenal cortical functions during and after exercise. MATERIALS AND METHODS Subjects. Nine trained adult males were studied (Table 1). The subjects signed an informed letter of consent after being told about the purpose, possible risks and stress associated with the study. A few weeks before the period of the experimental trials each subject performed maximal bicycle ergometer test, including electrocardiographic monitoring and direct maximal oxygen uptake measurement. Body fat was estimated by measuring four skin folds (Durning & Womersley 1974). Some characteristics of the subjects are summarized in Table 1. Study design. The subjects repeated twice a total of 36 km running exercise (5 km warming up, mean heart rate about 120 min-'; 27 km subjectively maximal, noncompetitive running, mean heart rate 16G175 min-'; 4 km cooling down, mean heart rate about 120min-l) with an interval of 21 days. The mean running time, which was recorded from the 27 km maximal running, was 1 h 42 min 58 s in the placebo trial and 1 h 41 min 48 s in the carbohydrate trial. T h e total running time in the both trials was
26.4 ( 2 4 3 4 ) 179.5 (173-188) 67.2 (6&73) 8.8 (6.2-11.6) 64.8 (57-70) 50.1 ( 4 6 5 2 ) 8.2 (3-12) 18.5 (1g22.5) 6.3 (3.6-7.5) 53.6 (10&200) 46.4 (1 1&180) 13.7 (6-25) 14.1 (6-28)
approximately 2 h 20 min. The subjects ran 100 to 200 km during the week before each test. The day before the test, each subject repeated individually standardized submaximal training, mean length 13.9 km (5-28 km). During the days preceding the study, the subjects had a normal mixed diet; none of them underwent special carbohydrate loading. In the morning before each test the subjects had a light mixed breakfast without coffee. In the test mornings the air temperature varied from 1.2 "C to +3.3 "C, and the relative humidity from 85% to 99%. In both trials, the subjects ingested in seven 150 ml boluses, a total of 1050 ml of two different solutions in a single-blind randomized order. The carbohydrate solution (carbohydrate trial) contained 5 g glucose and 10 g maltodextrin (glucose polymer consisting of 5-7 glucose units molecule-') in 150 ml water (osmolality 163 mOsm 1-l); the other solution (placebo trial) contained 20 mg saccharin (Hermesetas@)in 150 ml water (osmolality 10 mOsm IF'). In both trials the subjects ingested the 150 ml bolus after warming up, at 5.3 km intervals during the 27 km test, after cooling down and one hour after cooling down. In the CH trial the total amount of ingested carbohydrates was 105 g. The subjects came to the laboratory during both trials at 09.00 h. Before warming up, an antecubital sample (A) of venous blood was taken from sitting subjects. Sample B was taken 2-3 min after the 27 km maximal run, and sample C 2 h after the B sample. Plasma and serum were separated by centrifugation and thereafter stored at -20 "C until analysed. One of the nine subjects had upset stomach during the exercise, and his physical performance was disturbed to such an extent that he was not included in the study.
+
Analytical methods. After both trials, the subjects marked the degree of exhaustion at the end of the exercise on a visual analogue scale (a 10 cm long line,
Cortisol and gonadotrophins in exercise
375
Table 2. Plasma (P)/serum (S) concentrations of glucose, cortisol, LH, FSH and testosterone before exercise (A), immediately after exercise (B), and 2 h after exercise (C). Mean fSE, n = 8 A
B
C
-~
P-Glucose (mmol/-') Placebo Carbohydrate S-Cortisol (nmol 1-') Placebo Carbohydrate S-LH ( I U 1-l) Placebo Carbohydrate S-FSH (IU 1-') Placebo Carbohydrate S-Testosterone (nmol 1 ') Placebo Carbohydrate
~
4.4 f0.3 4.3 f0.2
5.6 f0.3 6.4 f0.3""
4.8fO.l 5.2 f 0.3
605 f67 585 f 4 6
819 f81 924 k 88"
595 f 9 1 576 f82
3.2 k 0.7 3.0 k 0.4
2.8 f0.4 2.3 f0.4'
2.1 f 0 . 6 1.7f0.3
2.7 f0.4 2.9k0.5
3.1 +0.5 2.8f0.5"
2.7 f0.5 2.6 f0.4
19.6f3.4 20.2 f 2.3
25.4f 3.2 23.2 k 2.9
16.6f 1.7 15.8f2.0
Asterisks indicate statistical difference between placebo and carbohydrate trials. Paired t-test : "; 0.01 < P < 0.05, **; 0.001 < P < 0.01. Joyce et al. 1975). The length of the segment of the line was converted into percents of maximal exhaustion. Serum total protein, osmolality, plasma glucose, serum cortisol, LH, FSH and testosterone were analysed. Serum total protein was measured by the Bradford method (Bradford 1976). Osmostat osmometer (Kagaku, Japan) was used for the determination of osmolality (C.V. < 1%). A D I G GA-1120 automatic glucose analyser (Kagaku, Japan) was utilized for determination of plasma glucose. Serum cortisol was measured by RIA kits purchased from Farmos Diagnostica (Turku, Finland). Serum LH and F S H were measured using time-resolved immunofluorometric assay (IFMA) kits (Delfia h L H Spec for LH and Delfia hFSH for FSH) provided by PhamaciaWallac (Turku, Finland). Serum testosterone was analysed after diethylether extraction by RIA (Huhtaniemi et al. 1985). The intra-assay and interassay CVs of the individual RIA/FIA methods were below 4 and 970, respectively. All assays were carried out in duplicate. Statistical difference in perceived exhaustion was tested using the rank sum test on paired data and other statistical differences were tested using the paired Students t-test. The serum concentrations of the hormones are expressed as mean and standard error (SE).
RESULTS Seven of the eight subjects estimated the exhaustion to be more severe in placebo trial than in carbohydrate trial. The mean (+SE)
visual analogue scale exhaustion index in the placebo trial was 73 (+6.9)% and in the carbohydrate trial 58 ( f6.2)y0(NS). Plasma glucose level of the B sample was lower in each subject in the trial in which the exhaustion was estimated to be more severe. The groups were similar in all the pretrial (A) measurements. Serum total protein Concentration did not change significantly during the run, indicating that there was no significant change in plasma volume. In serum osmolality, there was a significant increase between A and B sample in the placebo and carbohydrate trials (by 2.1 and 3.9%). Similar decreases were seen between B and C samples in both trials (1.9 and 2.4%). There was no difference in the increase and the decrease between the trials. In the B sample, all subjects, except one, had higher plasma glucose concentrations in the carbohydrate trial than in the placebo trial ( P < 0.02). The levels in the C sample decreased towards starting values, and the difference between the trials disappeared. The increase of glucose concentration from A to B sample was 27% in the placebo trial and 49% in the carbohydrate trial (group difference of the change; P < 0.05). The decrease of glucose values from B to C sample in placebo trial was 15% and in carbohydrate trial 19% (NS) (Table 2). During exercise, the serum cortisol concen-
376
T.3. Vasankari et al.
tration increased from A to B samples in the placebo trial by 36% and in the carbohydrate trial 58 yo (NS), the postexercise levels (B) being significantly higher in the carbohydrate trial ( P < 0.05). After a 2-h rest period the cortisol levels were decreased to or taken below the starting level in all subjects except one who was severely exhausted in both trials (S-cortisol over 1000 nmol 1-' in both trials). In the C sample there was no significant difference between the trials (Table 2). In the B sample, both LH and FSH concentrations were significantly greater in the placebo trial than in the carbohydrate trial ( P < 0.05, Table 2). In the samples taken 2 h later the same tendency was seen; six of the eight subjects had higher LH and FSH values in the placebo than the carbohydrate trial, but the differences were no more significant. The reduction of LH levels from A sample to C sample was 34% ( P < 0.05) in placebo trial and 43% ( P < 0.01) in the carbohydrate trial. In FSH values from A to B sample the difference of change between the trials was statistically significant ( P < 0.05). Serum testosterone increased from A samples to B samples 30% ( P < 0.01) in the placebo and 15% ( P < 0.01) in the carbohydrate trial and decreased from A samples to C samples 15 yo (NS) in the placebo and 22% ( P < 0.01) in the carbohydrate trial. N o significant group differences were seen.
DISCUSSION None of the participants of the placebo trial reached hypoglycaemia, despite the fact that only a light mixed breakfast preceded the trials. On the other hand it has been demonstrated that in exercise to exhaustion only one third of subjects were hypoglycaemic after the exercise (Felig et al. 1982). In the study of Felig et al. (1982) as well as in our study there was a significant difference in blood glucose levels between placebo and carbohydrate trials. The mechanisms of perceived exhaustion is not exactly known. During exercise, the glucose infusion has not eliminated exhaustion symptoms (Galbo et al. 1979). There are differing results on, if deficiency of energy substrates, especially depletion of muscular glycogen stores in exhausted muscle fibres, is decisive for exhaustion. According to the theory of Galbo et al. (1979), the primarily recruited muscle fibres are exhausted first and that after
primarily resting fibres also have to be recruited. According to this theory exhaustion occurs when a fixed intensity of innervation is exceeded, with secondarily to the pheripheral phenomena demands activation of increased area of the motor cortex. Additional exogenous glucose can delay this process even though there is no resultant difference in serum glucose concentration. This fits with subjective exhaustion analysis in our study. Also carbohydrate ingestion during exercise has been shown to be beneficial for prolonged psychophysical performance (Kujala et al. 1989). Most studies have shown that serum cortisol increases during prolonged physical exercise (Dessypris et al. 1976, Kuoppasalmi et al. 1980, MacConnie et al. 1986). Sotsky et al. (1988) observed that serum cortisol increased during exercise more if the subjects had low preexercise blood glucose than when being euglycaemic. Also Galbo et al. (1981) reported that the increase in serum cortisol values was greater after 59 h fasting than in control exercise, but when the fasting subjects repeated the exercise period with infusion of glucose after 30 min rest, their cortisol levels increased further. This association between glucose and cortisol levels agrees with out study. Because serum insulin increased at the same time, it was supposed that the change was secondary. It is also known, that during prolonged exercise, carbohydrate ingestion attenuates the decrease of serum insulin (Kujala et al. 1989), which may partly explain the association between carbohydrate ingestion and cortisol levels also in our study. Surprisingly the concentrations of gonadotrophins in placebo trial were higher, which may be explained by higher cortisol levels in the carbohydrate trial. It has been shown that hypercortisolism has acute suppressive effect on circulating testosterone (Cumming et al. 1983), which effect may be due to direct gonadal suppression (Bambino et al. 1981), or to inhibition of pituitary function (Sakamura et al. 1975). Because of decreased gonadotrophins our study provides evidence at least for the latter mechanism. Comparison between placebo and carbohydrate trials in C sample agrees with the study of Opstad et al. (1982) in which the reduction of testosterone could not be prevented with a high carbohydrate diet. In conclusion, carbohydrate ingestion during exercise results, at least in highly trained athletes,
Cortisol and gonadotrophins in exercise
377
KLIBANSKI, A,, BEITINS,I.Z., BADGER, T., LITTLE,R. & MCARTHUR, J.W. 1981. Reproductive function during fasting in men. 3 Clin Endocrrnol Metab 53, 258-263. O.J., KVIST,M., KARKUJALA,U.M., HEINONEN, KKAINEN, 0-P., MARNIEMI, J., NIITTYMAKI, K. & HAVAS,E. 1989. Orienteering performance and ingestion of glucose and glucose polymers. Br 3 The authors thank The Finnish Olympic Committee Sports Med 23, 105-108. for financial support of this study. Our acknowledgeU.M., ALEN, M. & HUHTANIEMI, I.T. 1990. ments are also due to Pertti Koskinen, M D and Jukka KUJALA, Gonadotrophin-releasing hormone and human Kapanen, MSc, for their expertise help in this study. chorionic gonadotrophin tests reveal that both hypothalamic and testicular endocrine functions are suppressed during acute prolonged physical exREFERENCES ercise. Clin Endocrinol 33, 219-225. M. & K., NAVERI,H., HARKONEN, BAMBINO,T.H. & HSUEH,A.J.W. 1981. Direct KUOPPASALMI, ADLERCREUTZ, H. 1980. Plasma cortisol, androinhibitory effect of glucocorticoids upon testicular stenedione, testosterone and luteinizing hormone in luteinizing hormone receptor and steroidegenesis in running exercise of different intensities. Scand 3 vivo and in vitro. Endocrinology 108, 2142-2148. Clin Lab Invest 40, 403409. BRADFORD, M.M. 1976. A rapid and sensitive method E., for the quantitation of microgram quantities of KUUSI, T., KOSTIAINEN,E., VARTIAINEN, PITKANEN, L., EHNHOLM, C., KORHONEN, H.J., protein utilizing the principle of protein-dye NISSINEN,A. & PUSKA,P. 1984. Acute effects of binding. Anal Biochem 72, 248-254. marathon running on levels of serum lipoproteins CUMMING, D.C., QUIGLEY, M.E. &YEN,S.S.C. 1983. and androgenic hormones in healthy males. MetabAcute suppression of circulating testosterone levels olism 33, 527-531. by cortisol in men. 3 Clin Endocrinol Metab 57, KREUZ,L.E., ROSE,R.M. & JENNINGS, J.R. 1972. 67 1-673. Suppression of plasma testosterone levels and DESSYPRIS, A,, KUOPPASALMI, K. & ADLERCREUTZ, H. psychological stress. Arch Gen Psychiat 26,479482. 1976. Plasma cortisol, testosterone, androstenedione MACCONNIE, S.E., BARKAN, A., LAMPMAN,R.M., and luteinizing hormone (LH) in a non-competitive SCHORK,M.A. & BEITINS,I.Z. 1986. Decreased marathon run. 3 Steroid Biochem 7, 33-37. hypothalamic gonadotrophin-releasing hormone seDURNING, J.V.G.A. WOMERSLEY, J. 1974. Body fat cretion in male marathon runners. N E n g l 3 Med assessed from total body density and its estimation 315, 411417. from skinfold thickness : measurements on 481 men MOUGIN,C., HENRIET, M.T., BAULAY, A., HATON, and women aged from 16 to 72 years. B r 3 Nutr 32, S. & GAILLARD, R.C. 1988. Plasma D., BERTHELAY, 77-97. levels of beta-endorphin, prolactin and gonadoA. & WAHREN, J. FELIG,P., CHERIF,A,, MINAGAWA, trophins in male athletes after an international 1982. Hypoglycemia during prolonged exercise in nordic ski race. Eur 3 Appl Physiol57, 425429. normal men. N E n g l 3 Med 306, 895-900. A. 1982. Decreased serum OPSTAD, P.K. & AAKVAAG, N.J. 1979. GALBO,H., HOLST,J.J. & CHRISTENSEN, levels of oestradiol, testosterone and prolactin The effect of different diets and insulin on the during prolonged physical strain and sleep deprihormonal response to prolonged exercise. Acta vation, and the influence of a high calory diet. Eur Physzol Scand 107, 19-32. 3 Appl Physiol49, 343-348. N.J., MIKINES, K.J., SONNE, SAKAMURA, GALBO, H., CHRISTENSEN, M., TAKEBE, K. & NAKAGAWA, S. 1975. J., HAGEN, C. & FAHRENKRUG, J. 1981. B., HILSTED, Inhibition of luteinizing hormone secretion induced The effect of fasting on the hormonal response to by synthetic LRH by long-term treatment with graded exercise. J Cltn Endocrinol Metab 52, glucocorticoids in human subjects. 3Clin Endocrinot 1106-1112. Metab 40, 774-779. HUHTANIEMI, I.T., NIKULA, H. & RANNIKKO, S. 1985. SCHMID,P., PUSCH,H.H., WOLF, W., PILGER,E., Treatment of prostatic cancer with a gonadoPESSENHOFER, H., SCHWABERGER, G., PRISTAUTZ, H. trophin-releasing hormone agonist analog : acute & PURSTNER, P. 1982. Serum FSH, LH, and and long term effects on endocrine functions of the testosterone in humans after physical exercise. Znt 3 testis tissue. 3 Clin Endocrinol Metab 61, 698-704. Sports Med 3, 84-89. JOYCE, C.R.B., ZUTSHI, D.W., HRUBES, V. & MASON, SOTSKY,M.J., SHMUEL,S. & SHAMOON, H. 1989. R.M. 1975. Comparison of fixed interval and visual Regulation of counterregulatory hormone secretion in man during exercise and hypoglycemia 3 Clin analogue scales for rating chronic pain. Eur 3 Clin Pharmacol 8, 4 15420. Endocrinol Metab 68, 9-16. in higher postexercise cortisol and lower gonadotrophin levels. Carbohydrate ingestion during exercise does not seem to be beneficial in preventing the postexercise decrease of gonadotrophin secretion.
ADONIS 000 167729100055W
Carbohydrate ingestion during prolonged running exercise results in an increase of serum cortisol and decrease of gonadotrophins T. J . V A S A N K A R 1 , U. M. K U J A L A , T.T. V I L J A N E N and I. T. H U H T A N I E M I Department of Physiology, University of Turku and Paavo Nurmi Centre, Turku, Finland VASANKARI,T. J., KUJALA,U. M., VILJANEN, T. T. & HUHTANIEMI I. T. 1991. Carbohydrate ingestion during prolonged running exercise results in an increase of serum cortisol and decrease of gonadotrophins. Acta Physiol Scand 141, 373-378. Received 25 March 1990 accepted 29 September 1990. ISSN 0001-6772. Department of Physiology, University of Turku, Finland. We tested the hypothesis that improved availability of energy through carbohydrate ingestion could counteract the documented suppression of pituitary-gonadal function during prolonged exercise. Nine trained males repeated twice a 36 kilometre running exercise with two weeks interval. During the tests the subjects ingested in a randomized single-blind fashion a total of 1050 millilitres of carbohydrate and placebo solutions. The total amount of ingested carbohydrate was 105 grammes. Venous blood samples were taken before the exercise (sample A), immediately after exercise (B), and 2 hours later (C). In the B samples plasma glucose was 14% higher ( P < 0.01) and serum cortisol 13% higher ( P < 0.05) in the carbohydrate than in the control trial. In contrast, the level of LH was 18% (P< 0.05) and that of FSH 11yo lower (P< 0.05) in the carbohydrate than in the placebo trial. Serum testosterone concentration did not differ between the treatment groups. We conclude that carbohydrate ingestion does not counteract the exercise-associated suppression of gonadotrophin secretion, but results in paradoxical increase of serum cortisol and decrease of LH and FSH, in comparison to placebo-treated controls. The higher cortisol level in carbohydrate group may be secondary to higher insulin levels. Key words :cortisol, exercise, FSH, glucose, LH, testosterone
Prolonged exercise can lead to hypoglycaemia (Felig et al. 1982). A 4-day diet low in carbohydrate has been shown to decrease the time to exhaustion in exercise when compared to a diet high in carbohydrate (Galbo et a1 1979). T h e effects of glucose ingestion/infusion on perceived exhaustion and various hormone levels during exercise are greater if fasting precedes the exercise (Galbo et al. 1981). Serum testosterone concentration has been shown to decrease after prolonged physical exercise (Kreuz et al. 1972, Dessypris et al. 1976). T h e exercise induced suppression of serum testosterone is associated Correspondence: Tommi Vasankari, Department of Physiology, University of Turku, SF-20520 Turku, Finland
with two effects : suppressed GnRH stimulation of gonadotrophin release during exercise and decreased testicular capacity to secret testosterone during the recovery period (Kujala et al. 1990). T h e reported gonadotrophin responses during exercise have been variable (Dessypris et al. 1976, Kuoppasalmi et al. 1980, Schmid et al. 1982, Kuusi et al. 1984, Mougin et al. 1988). Long-term fasting has been reported to decrease serum testosterone levels, and after refeeding these levels were normalized (Klibanski et al. 1981). In one study, the decrease of testosterone in prolonged physical strain and sleep deprivation could not be prevented with a high carbohydrate diet (Opstad et al. 1982). In endurance athletes, the repeated exerciseassociated suppression of testosterone may
373
374
T.3. Vasankari et al.
Table 1.Mean (range) of some characteristics of the subjects Age (years) Height (cm) Weight (kg) Body fat (yo) Maximal oxygen uptake (ml kg-' min-') Anaerobic threshold (ml kg-l min-l) Regular training background (years) Training during the preceding three years (1000 km) Training during the preceding year (1000 km) Training during the week before placebo trial (km) Training during the week before carbohydrate trial (km) Training during the day before placebo trial (km) Training during the day before carbohydrate trial (km)
cumulatively lead to androgen deficiency, which may be unfavourable, when training-induced improvement in performance is expected. Although the nutritional status in long-term has pronouilced effects on pituitary-gonadal function (Klibanski et al. 1981), it is not known whether the availability of energy during exercise (e.g. through ingestion of carbohydrates) has effect on the hypothalamus-pituitary-gonadal axis. Therefore, we challenged a group of healthy trained males during prolonged exercise with a high energy carbohydrate solution to study its effects on pituitary-gonadal and adrenal cortical functions during and after exercise. MATERIALS AND METHODS Subjects. Nine trained adult males were studied (Table 1). The subjects signed an informed letter of consent after being told about the purpose, possible risks and stress associated with the study. A few weeks before the period of the experimental trials each subject performed maximal bicycle ergometer test, including electrocardiographic monitoring and direct maximal oxygen uptake measurement. Body fat was estimated by measuring four skin folds (Durning & Womersley 1974). Some characteristics of the subjects are summarized in Table 1. Study design. The subjects repeated twice a total of 36 km running exercise (5 km warming up, mean heart rate about 120 min-'; 27 km subjectively maximal, noncompetitive running, mean heart rate 16G175 min-'; 4 km cooling down, mean heart rate about 120min-l) with an interval of 21 days. The mean running time, which was recorded from the 27 km maximal running, was 1 h 42 min 58 s in the placebo trial and 1 h 41 min 48 s in the carbohydrate trial. T h e total running time in the both trials was
26.4 ( 2 4 3 4 ) 179.5 (173-188) 67.2 (6&73) 8.8 (6.2-11.6) 64.8 (57-70) 50.1 ( 4 6 5 2 ) 8.2 (3-12) 18.5 (1g22.5) 6.3 (3.6-7.5) 53.6 (10&200) 46.4 (1 1&180) 13.7 (6-25) 14.1 (6-28)
approximately 2 h 20 min. The subjects ran 100 to 200 km during the week before each test. The day before the test, each subject repeated individually standardized submaximal training, mean length 13.9 km (5-28 km). During the days preceding the study, the subjects had a normal mixed diet; none of them underwent special carbohydrate loading. In the morning before each test the subjects had a light mixed breakfast without coffee. In the test mornings the air temperature varied from 1.2 "C to +3.3 "C, and the relative humidity from 85% to 99%. In both trials, the subjects ingested in seven 150 ml boluses, a total of 1050 ml of two different solutions in a single-blind randomized order. The carbohydrate solution (carbohydrate trial) contained 5 g glucose and 10 g maltodextrin (glucose polymer consisting of 5-7 glucose units molecule-') in 150 ml water (osmolality 163 mOsm 1-l); the other solution (placebo trial) contained 20 mg saccharin (Hermesetas@)in 150 ml water (osmolality 10 mOsm IF'). In both trials the subjects ingested the 150 ml bolus after warming up, at 5.3 km intervals during the 27 km test, after cooling down and one hour after cooling down. In the CH trial the total amount of ingested carbohydrates was 105 g. The subjects came to the laboratory during both trials at 09.00 h. Before warming up, an antecubital sample (A) of venous blood was taken from sitting subjects. Sample B was taken 2-3 min after the 27 km maximal run, and sample C 2 h after the B sample. Plasma and serum were separated by centrifugation and thereafter stored at -20 "C until analysed. One of the nine subjects had upset stomach during the exercise, and his physical performance was disturbed to such an extent that he was not included in the study.
+
Analytical methods. After both trials, the subjects marked the degree of exhaustion at the end of the exercise on a visual analogue scale (a 10 cm long line,
Cortisol and gonadotrophins in exercise
375
Table 2. Plasma (P)/serum (S) concentrations of glucose, cortisol, LH, FSH and testosterone before exercise (A), immediately after exercise (B), and 2 h after exercise (C). Mean fSE, n = 8 A
B
C
-~
P-Glucose (mmol/-') Placebo Carbohydrate S-Cortisol (nmol 1-') Placebo Carbohydrate S-LH ( I U 1-l) Placebo Carbohydrate S-FSH (IU 1-') Placebo Carbohydrate S-Testosterone (nmol 1 ') Placebo Carbohydrate
~
4.4 f0.3 4.3 f0.2
5.6 f0.3 6.4 f0.3""
4.8fO.l 5.2 f 0.3
605 f67 585 f 4 6
819 f81 924 k 88"
595 f 9 1 576 f82
3.2 k 0.7 3.0 k 0.4
2.8 f0.4 2.3 f0.4'
2.1 f 0 . 6 1.7f0.3
2.7 f0.4 2.9k0.5
3.1 +0.5 2.8f0.5"
2.7 f0.5 2.6 f0.4
19.6f3.4 20.2 f 2.3
25.4f 3.2 23.2 k 2.9
16.6f 1.7 15.8f2.0
Asterisks indicate statistical difference between placebo and carbohydrate trials. Paired t-test : "; 0.01 < P < 0.05, **; 0.001 < P < 0.01. Joyce et al. 1975). The length of the segment of the line was converted into percents of maximal exhaustion. Serum total protein, osmolality, plasma glucose, serum cortisol, LH, FSH and testosterone were analysed. Serum total protein was measured by the Bradford method (Bradford 1976). Osmostat osmometer (Kagaku, Japan) was used for the determination of osmolality (C.V. < 1%). A D I G GA-1120 automatic glucose analyser (Kagaku, Japan) was utilized for determination of plasma glucose. Serum cortisol was measured by RIA kits purchased from Farmos Diagnostica (Turku, Finland). Serum LH and F S H were measured using time-resolved immunofluorometric assay (IFMA) kits (Delfia h L H Spec for LH and Delfia hFSH for FSH) provided by PhamaciaWallac (Turku, Finland). Serum testosterone was analysed after diethylether extraction by RIA (Huhtaniemi et al. 1985). The intra-assay and interassay CVs of the individual RIA/FIA methods were below 4 and 970, respectively. All assays were carried out in duplicate. Statistical difference in perceived exhaustion was tested using the rank sum test on paired data and other statistical differences were tested using the paired Students t-test. The serum concentrations of the hormones are expressed as mean and standard error (SE).
RESULTS Seven of the eight subjects estimated the exhaustion to be more severe in placebo trial than in carbohydrate trial. The mean (+SE)
visual analogue scale exhaustion index in the placebo trial was 73 (+6.9)% and in the carbohydrate trial 58 ( f6.2)y0(NS). Plasma glucose level of the B sample was lower in each subject in the trial in which the exhaustion was estimated to be more severe. The groups were similar in all the pretrial (A) measurements. Serum total protein Concentration did not change significantly during the run, indicating that there was no significant change in plasma volume. In serum osmolality, there was a significant increase between A and B sample in the placebo and carbohydrate trials (by 2.1 and 3.9%). Similar decreases were seen between B and C samples in both trials (1.9 and 2.4%). There was no difference in the increase and the decrease between the trials. In the B sample, all subjects, except one, had higher plasma glucose concentrations in the carbohydrate trial than in the placebo trial ( P < 0.02). The levels in the C sample decreased towards starting values, and the difference between the trials disappeared. The increase of glucose concentration from A to B sample was 27% in the placebo trial and 49% in the carbohydrate trial (group difference of the change; P < 0.05). The decrease of glucose values from B to C sample in placebo trial was 15% and in carbohydrate trial 19% (NS) (Table 2). During exercise, the serum cortisol concen-
376
T.3. Vasankari et al.
tration increased from A to B samples in the placebo trial by 36% and in the carbohydrate trial 58 yo (NS), the postexercise levels (B) being significantly higher in the carbohydrate trial ( P < 0.05). After a 2-h rest period the cortisol levels were decreased to or taken below the starting level in all subjects except one who was severely exhausted in both trials (S-cortisol over 1000 nmol 1-' in both trials). In the C sample there was no significant difference between the trials (Table 2). In the B sample, both LH and FSH concentrations were significantly greater in the placebo trial than in the carbohydrate trial ( P < 0.05, Table 2). In the samples taken 2 h later the same tendency was seen; six of the eight subjects had higher LH and FSH values in the placebo than the carbohydrate trial, but the differences were no more significant. The reduction of LH levels from A sample to C sample was 34% ( P < 0.05) in placebo trial and 43% ( P < 0.01) in the carbohydrate trial. In FSH values from A to B sample the difference of change between the trials was statistically significant ( P < 0.05). Serum testosterone increased from A samples to B samples 30% ( P < 0.01) in the placebo and 15% ( P < 0.01) in the carbohydrate trial and decreased from A samples to C samples 15 yo (NS) in the placebo and 22% ( P < 0.01) in the carbohydrate trial. N o significant group differences were seen.
DISCUSSION None of the participants of the placebo trial reached hypoglycaemia, despite the fact that only a light mixed breakfast preceded the trials. On the other hand it has been demonstrated that in exercise to exhaustion only one third of subjects were hypoglycaemic after the exercise (Felig et al. 1982). In the study of Felig et al. (1982) as well as in our study there was a significant difference in blood glucose levels between placebo and carbohydrate trials. The mechanisms of perceived exhaustion is not exactly known. During exercise, the glucose infusion has not eliminated exhaustion symptoms (Galbo et al. 1979). There are differing results on, if deficiency of energy substrates, especially depletion of muscular glycogen stores in exhausted muscle fibres, is decisive for exhaustion. According to the theory of Galbo et al. (1979), the primarily recruited muscle fibres are exhausted first and that after
primarily resting fibres also have to be recruited. According to this theory exhaustion occurs when a fixed intensity of innervation is exceeded, with secondarily to the pheripheral phenomena demands activation of increased area of the motor cortex. Additional exogenous glucose can delay this process even though there is no resultant difference in serum glucose concentration. This fits with subjective exhaustion analysis in our study. Also carbohydrate ingestion during exercise has been shown to be beneficial for prolonged psychophysical performance (Kujala et al. 1989). Most studies have shown that serum cortisol increases during prolonged physical exercise (Dessypris et al. 1976, Kuoppasalmi et al. 1980, MacConnie et al. 1986). Sotsky et al. (1988) observed that serum cortisol increased during exercise more if the subjects had low preexercise blood glucose than when being euglycaemic. Also Galbo et al. (1981) reported that the increase in serum cortisol values was greater after 59 h fasting than in control exercise, but when the fasting subjects repeated the exercise period with infusion of glucose after 30 min rest, their cortisol levels increased further. This association between glucose and cortisol levels agrees with out study. Because serum insulin increased at the same time, it was supposed that the change was secondary. It is also known, that during prolonged exercise, carbohydrate ingestion attenuates the decrease of serum insulin (Kujala et al. 1989), which may partly explain the association between carbohydrate ingestion and cortisol levels also in our study. Surprisingly the concentrations of gonadotrophins in placebo trial were higher, which may be explained by higher cortisol levels in the carbohydrate trial. It has been shown that hypercortisolism has acute suppressive effect on circulating testosterone (Cumming et al. 1983), which effect may be due to direct gonadal suppression (Bambino et al. 1981), or to inhibition of pituitary function (Sakamura et al. 1975). Because of decreased gonadotrophins our study provides evidence at least for the latter mechanism. Comparison between placebo and carbohydrate trials in C sample agrees with the study of Opstad et al. (1982) in which the reduction of testosterone could not be prevented with a high carbohydrate diet. In conclusion, carbohydrate ingestion during exercise results, at least in highly trained athletes,
Cortisol and gonadotrophins in exercise
377
KLIBANSKI, A,, BEITINS,I.Z., BADGER, T., LITTLE,R. & MCARTHUR, J.W. 1981. Reproductive function during fasting in men. 3 Clin Endocrrnol Metab 53, 258-263. O.J., KVIST,M., KARKUJALA,U.M., HEINONEN, KKAINEN, 0-P., MARNIEMI, J., NIITTYMAKI, K. & HAVAS,E. 1989. Orienteering performance and ingestion of glucose and glucose polymers. Br 3 The authors thank The Finnish Olympic Committee Sports Med 23, 105-108. for financial support of this study. Our acknowledgeU.M., ALEN, M. & HUHTANIEMI, I.T. 1990. ments are also due to Pertti Koskinen, M D and Jukka KUJALA, Gonadotrophin-releasing hormone and human Kapanen, MSc, for their expertise help in this study. chorionic gonadotrophin tests reveal that both hypothalamic and testicular endocrine functions are suppressed during acute prolonged physical exREFERENCES ercise. Clin Endocrinol 33, 219-225. M. & K., NAVERI,H., HARKONEN, BAMBINO,T.H. & HSUEH,A.J.W. 1981. Direct KUOPPASALMI, ADLERCREUTZ, H. 1980. Plasma cortisol, androinhibitory effect of glucocorticoids upon testicular stenedione, testosterone and luteinizing hormone in luteinizing hormone receptor and steroidegenesis in running exercise of different intensities. Scand 3 vivo and in vitro. Endocrinology 108, 2142-2148. Clin Lab Invest 40, 403409. BRADFORD, M.M. 1976. A rapid and sensitive method E., for the quantitation of microgram quantities of KUUSI, T., KOSTIAINEN,E., VARTIAINEN, PITKANEN, L., EHNHOLM, C., KORHONEN, H.J., protein utilizing the principle of protein-dye NISSINEN,A. & PUSKA,P. 1984. Acute effects of binding. Anal Biochem 72, 248-254. marathon running on levels of serum lipoproteins CUMMING, D.C., QUIGLEY, M.E. &YEN,S.S.C. 1983. and androgenic hormones in healthy males. MetabAcute suppression of circulating testosterone levels olism 33, 527-531. by cortisol in men. 3 Clin Endocrinol Metab 57, KREUZ,L.E., ROSE,R.M. & JENNINGS, J.R. 1972. 67 1-673. Suppression of plasma testosterone levels and DESSYPRIS, A,, KUOPPASALMI, K. & ADLERCREUTZ, H. psychological stress. Arch Gen Psychiat 26,479482. 1976. Plasma cortisol, testosterone, androstenedione MACCONNIE, S.E., BARKAN, A., LAMPMAN,R.M., and luteinizing hormone (LH) in a non-competitive SCHORK,M.A. & BEITINS,I.Z. 1986. Decreased marathon run. 3 Steroid Biochem 7, 33-37. hypothalamic gonadotrophin-releasing hormone seDURNING, J.V.G.A. WOMERSLEY, J. 1974. Body fat cretion in male marathon runners. N E n g l 3 Med assessed from total body density and its estimation 315, 411417. from skinfold thickness : measurements on 481 men MOUGIN,C., HENRIET, M.T., BAULAY, A., HATON, and women aged from 16 to 72 years. B r 3 Nutr 32, S. & GAILLARD, R.C. 1988. Plasma D., BERTHELAY, 77-97. levels of beta-endorphin, prolactin and gonadoA. & WAHREN, J. FELIG,P., CHERIF,A,, MINAGAWA, trophins in male athletes after an international 1982. Hypoglycemia during prolonged exercise in nordic ski race. Eur 3 Appl Physiol57, 425429. normal men. N E n g l 3 Med 306, 895-900. A. 1982. Decreased serum OPSTAD, P.K. & AAKVAAG, N.J. 1979. GALBO,H., HOLST,J.J. & CHRISTENSEN, levels of oestradiol, testosterone and prolactin The effect of different diets and insulin on the during prolonged physical strain and sleep deprihormonal response to prolonged exercise. Acta vation, and the influence of a high calory diet. Eur Physzol Scand 107, 19-32. 3 Appl Physiol49, 343-348. N.J., MIKINES, K.J., SONNE, SAKAMURA, GALBO, H., CHRISTENSEN, M., TAKEBE, K. & NAKAGAWA, S. 1975. J., HAGEN, C. & FAHRENKRUG, J. 1981. B., HILSTED, Inhibition of luteinizing hormone secretion induced The effect of fasting on the hormonal response to by synthetic LRH by long-term treatment with graded exercise. J Cltn Endocrinol Metab 52, glucocorticoids in human subjects. 3Clin Endocrinot 1106-1112. Metab 40, 774-779. HUHTANIEMI, I.T., NIKULA, H. & RANNIKKO, S. 1985. SCHMID,P., PUSCH,H.H., WOLF, W., PILGER,E., Treatment of prostatic cancer with a gonadoPESSENHOFER, H., SCHWABERGER, G., PRISTAUTZ, H. trophin-releasing hormone agonist analog : acute & PURSTNER, P. 1982. Serum FSH, LH, and and long term effects on endocrine functions of the testosterone in humans after physical exercise. Znt 3 testis tissue. 3 Clin Endocrinol Metab 61, 698-704. Sports Med 3, 84-89. JOYCE, C.R.B., ZUTSHI, D.W., HRUBES, V. & MASON, SOTSKY,M.J., SHMUEL,S. & SHAMOON, H. 1989. R.M. 1975. Comparison of fixed interval and visual Regulation of counterregulatory hormone secretion in man during exercise and hypoglycemia 3 Clin analogue scales for rating chronic pain. Eur 3 Clin Pharmacol 8, 4 15420. Endocrinol Metab 68, 9-16. in higher postexercise cortisol and lower gonadotrophin levels. Carbohydrate ingestion during exercise does not seem to be beneficial in preventing the postexercise decrease of gonadotrophin secretion.
