Antioxidant

Enzyme Activity During Prolonged Exercise in Amenorrheic Eumenorrheic Athletes J.A. Kanaley

and

and L.L. Ji

Exogenous 179-estradiol (EJ has been shown to be associated with elevated levels of erythrocyte glutathione peroxidase (GPX) activity. The purpose of this study was to examine the influence of endogenous E,, as defined by menstrual status (amenorrhea Y eumenorrhea), on activity of blood antioxidant enzymes at rest and during prolonged exercise. Six amenorrheic (AMc) and six eumenorrheic (EUc) athletes were subjected to a treadmill running test at 60% VO,max for 90 minutes. Serial blood samples were taken from a forearm vein at rest, 30,60, and 90 minutes during exercise, and 15 minutes into recovery. Resting estrogen levels were significantly lower in AMc athletes at rest and during exercise as compared with EUc athletes, whereas plasma cortisol levels in AMc were significantly higher. GPX activity was significantly higher in AMc than EUc at rest (46.9 f 7.7 v 30.2 2 2.2 nmol/min x mg Hb, P < .05, respectively) and throughout exercise. Glutathione reductase (GR) activity was similar between the two groups at rest and was significantly higher (P < .Ol) in AMc than EUc during exercise. Plasma lipid peroxidation and catalase activity did not change significantly in response to exercise, nor were they different between AMc and EUc athletes. GPX activity was found to be negatively correlated with E2 (r = -.64, P < .Ol) and positively correlated with cortisol (I = .69, P < .Ol). It is tentatively concluded that the alteration of hormonal status in amenorrhea has an influence on the blood antioxidant enzyme system. Copyright 0 1991 by W.B. Saunders Company

I

NTENSE physical training in female athletes has been shown to be associated with an increased occurrence of menstrual dysfunction.’ In particular, amenorrhea (one to two menstrual cycles per year) has been of great concern among well-trained athletes primarily because of the implications on bone health.’ The low 17l3-estradiol (E,) levels of these athletes as compared with eumenorrheic athletes (EUc) represent an altered hormonal milieu and may alter the internal environment sufficiently to cause other disorders. Severe physical exercise is known to induce oxygen-free radical (hydrogen peroxide [H202], hydroxyl radicals [OH’], and superoxide anion [OJ) generation.’ High levels of cytotoxic radicals can result in lipid peroxidation, which may cause considerable damage to cell membrane, mitochondria, and DNA.4 Free radical scavenger enzymes, such as catalase, glutathione peroxidase (GPX), superoxide dismutase (SOD), and glutathione reductase (GR) reduce the susceptibility of the cell to the potentially harmful byproducts of one-electron reduction of molecular oxygen. Increases in lipid peroxidation in erythrocytes and other tissues indicate that the defense system is unable to cope with free radical formation during periods of increased oxygen consumption. A number of animal studies have demonstrated that erythrocyte antioxidant enzyme activity increases in response to chronic aerobic training.5,6 Unfortunately, studies regarding human erythrocyte antioxidant enzymes have been limited. Ohno et al’ reported that 30 minutes of submaximal exercise resulted in increased erythrocytes GR activity in sedentary men. Lovlin et al* showed

From the Department of Kinesiology, Universiry of Illinois, UrbanaChampaign, IL. Supported in part by the Illinois Association for Health, Physical Education, Recreation and Dance and the American Heart Association (Illinois Afilia te). Address reprint requests to J.A. Kanaley, PhD, St. Mary’s Endocrine Research Unit, Mayo Clinic, Rochester, MN 55905. Copyright 0 1991 by W B. Saunders Company 00260495/9Il4001-0016$03.0010 88

that exhaustive exercise in man could induce extensive lipid peroxidation in blood plasma. An early study by Cape1 et al’ demonstrated that women using oral contraceptives had higher GPX activity than either premenopausal or post menopausal non-oral contraceptive users. Moreover, those individuals who used oral contraceptives for a longer duration had a higher GPX than individuals using oral contraceptives for less than 6 months. Increased lipid and lipoperoxides have also been found in oral contraceptive users, especially those who use the high estrogenic steroid.“’ This gives rise to the question as to whether differences in endogenous plasma E, levels can influence lipid peroxidation and subsequently the activity of the antioxidant enzymes. Hence, the purpose of this investigation was to examine the influence of menstrual status (amenorrhea v eumenorrhea) on the activity of free radical scavenging enzymes in erythrocytes during prolonged aerobic exercise. METHODS Subjects Six amenorrheic (AMc) and 6 EUc athletes participated in this study and informed consent was obtained from all subjects. The AMc athletes had only one to two menstrual cycles in the past year and no menstrual cycles in the past 7 months. The EUc athletes were tested in the early follicular phase (day 3 to 5) of their cycle. These women were well-trained athletes, who ran a minimum of 35 miles per week. Exercise

Testing

Two testing sessions were conducted in this study. In the first session, subjects were given a maximal aerobic power test (VO,max) on a treadmill. A work intensity corresponding to 60% VO,max was determined and used as the workload for the submaximal exercise test given 1 week later. On the morning of testing, subjects arrived in the lab at 6:oO AM in at least an &hour postabsorptive state. A catheter was inserted into a forearm vein and subjects rested for 2 hours. The subjects then ran at 60% VO,max for 90 minutes, followed by 15 minutes of recovery. Blood samples were drawn at rest, 10 minutes before the onset of exercise, and at minutes 30,60, and 90 during exercise and 15 minutes of recovery.

Metabolism,

Vol40, No 1 (January), 1991: pp 88-92

ANTIOXIDANT

69

ENZYME ACTIVITY DURING EXERCISE

Hematocrits were taken on each sample plasma volume shifts could be made.

so that

correction

for 110

Blood Preparation and Assays The blood samples were drawn After completion of the test, the rpm for 20 minutes, and the separated and frozen at -80°C performed.

and immediately placed on ice. blood was centrifuged at 2,500 erythrocytes and plasma were until the enzyme assays were

The hormone assays were performed using assay kits from Diagnostic Products (Los Angeles, CA). For the EZ assay, prior to using the kit, the samples were extracted with a procedure described by Bahr et al.” Samples were extracted first with anesthesia grade ether. After evaporation of the ether, the residue was extracted again with 1 mL of 75% methanol and 1 mL hexane. After removing the hexane phase, the methanol was evaporated and then assayed in duplicate according to the procedures outlined by Diagnostic Products, with a double-antibody procedure. The intraassay and interassay coefficients of variation were 6.7% and 7.5 %, respectively. A coated tube procedure (Diagnostic Products, 19b9) was used for the assay of cortisol (intraassay and interassay coefficient of variation, 4.2% and 4.8%, respectively). The etythrocytes were hemolyzed and diluted with distilled waler at a 1:2ratio before the enzyme assay. Catalase was assayed at 22°C by methods previously described by Ji et al.” Lipid peroxidation in plasma was determined by measuring malondialdehyde (MDA) in butanol extracts according to Uchiyama and Mihara.” GPX was assayed at 37°C outlined by Flohe and Gunzler.” using H,O, as substrate. The method of Carlberg and Mannervik” was used to measure GR activity at 30°C. Lactate was measured using the techniques of Bergmeyer” and glucose levels were determined using glucose oxidase-peroxidase method accordmg to Kneer et al.” Statistics

90

70

0

30 Time

90

110

t

tt Rest

60 (min)

1 Exercise begins

11 repeated measures ANOVA was used to analyze the data. Multiple regression was used to establish the relationships between the hormonal levels and the enzyme activity. The alpha level was set at 0.05. RESULTS

‘The anthropometric characteristics are presented in Table 1. No significant differences were found between the AMc and EUc groups. Figure la shows the plasma hormonal data at rest and during prolonged exercise. Ez levels were significantly lower in the AMc than EUc athletes at rest (43.2 +- 5.7~ 64.3 f 6.5 pg/mL) and remained lower throughout exercise. In EUc subjects, El levels were significantly elevated above resting level by 30 minutes of exercise. Submaximal exercise did not alter plasma E2 levels in AMc athletes. Basal cortisol levels Table 1. Anthropometric

Characteristics

of the Subjects

(mean k SD)

Parameter Age (vr) Height (cm) Weight (kg) Maximal aerobic capacity (mL/kg . min’) Percent fat

EUc

AMc

(n = 6)

(n = 6)

21.9 2 3.8 161.3 ? 7.3 50.3 f 5.2

23.5 2 2.6 160.1 f 5.5 48.5 f 6.0

58.1 + 5.1 16.2 ? 2.8

58.2 + 5.5 12.0 + 4.6

Fig 1. (a) Plasma E, and (b) cortisol responses to prolonged submaximal exercise at 60% VO,max. Bars indicate SE. --El-, EUc; -_*--, Afvlc. *Significant difference between groups at various time points (P < .05); tsignificant increasefrom rest (P i .05); @significant group effect (P < .Ol).

were significantly greater in the AMc than in the EUc athletes (Fig lb). These values were 19.5 2 3.1 and 11.1 2 1.0 pg/dL, respectively. Further, the pattern of cortisol response to exercise was significantly different (P < .Ol) between groups. Catalase activity was not significantly different between the two groups at rest and during the early stages of exercise (Fig 2a). However, after 90 minutes of treadmill running, AMc subjects has significantly lower erythrocyte catalase activity than the EUc athletes (P < .OS). As depicted in Fig 2b, AMc athletes had significantly higher erythrocyte GPX activity (P < .0.5) at rest, during exercise, and during recovery than the EUc athletes. Further, GPX activity in AMc subjects decreased in the first 60 minutes of exercise, but increased sharply at minute 90. This delayed response of GPX in AMc athletes coincided with a drop of erythrocyte catalase activity at the same stage of prolonged exercise. GR activity was identical between AMc and EUc groups at rest (Fig 2~). During exercise and recovery, GR activity in AMc subjects was significantly higher (P < .Ol) than EUc subjects.

KANALEY AND JI

a 20 -

T IS -

10 -

increase during 90 minutes of submaximal exercise (Fig 3~). Further, AMc athletes had slightly higher MDA levels in plasma as compared with the EUc, although the differences did not reach statistical significance. The relationship between erythrocyte antioxidant activities and hormonal levels was examined using linear correlation method. Table 2 shows that a significant negative a

S-

O-

55 -

b

T

1.4 -

45 -

1

35

0-I

-

25 -

95 -

b

90 85 -

O-

80 75 70 65 60 55 01

0.20

ot Rest

C

-

0.180

T

Exercise begins

30

60 Time (min)

90

110

T

0.16

-

Exercise

ends

Fig 2. Antioxidant enzyme responses to prolonged submaximal exercise at 60% VO,max. Bars indicate SE. -0~ EUc; -6, AMc. *Significant difference between groups at various time points (P < .05); @significant group effect [P < 31).

.,

0,

0

t

.,

,.,

t

.,

60

30 Time

Rest

Plasma lactate and glucose concentrations were not significantly different between groups or over time (Fig 3a and b). However, amenorrheics had a consistently lower glucose level as compared with those of the EUc group. Lipid peroxidation in the plasma showed a progressive

’

.,,,

90

110

(min) t

Exercise begins Fig 3. Plasma metabolites in response to prolonged submaximal exercise at 60% VO,max. Bars indicate SE. -i3-, EUc; -_*-, AMc.

ANTIOXIDANT

ENZYME ACTIVITY

91

DURING EXERCISE

Table 2. Correlations Between Resting Hormonal Variables and Antioxidant Enzyme Activity

GPX Estrogen -

-.64t

Cortisol

.69t

GR .I8 -.26

Catalase .37 -.48*

MDA Content -.23 .45

*P < .05, t/J < .Ol.

correlation exists between resting levels of E2 and GPX (r = -.64, P < .Ol). Additionally, plasma cortisol levels were significantly correlated with erythrocyte GPX activity (r = .69, P < .Ol), and negatively correlated with catalase activity (r = -.48, P < .05). DISCUSSION

The lower resting E2 levels observed in AMc as compared with EUc athletes were expected and are consistent with earlier findings.“,” Further, the present study demonstrated that EUc subjects had a significant E, response at the early stage of exercise, whereas AMc subjects showed no appreciable alteration either during or after exercise. There are conflicting data in the literature regarding the plasma Ez levels to prolonged submaximal exercise. Many investigators have shown elevated E, levels either during or post-exercise both in the AMc and EUc.18,19Cumming and Rebar” observed an exercise-induced E, response only in EUc women. Furthermore, Jurkowski et al*’ reported that women in the early follicular phase increased their blood E, only at workloads greater than 80% VO,max. In contrast, Loucks and Horvath” found that significant differences in blood EL levels between AMc and EUc existed only at rest, but not post-exercise. Differences between the AMc and EUc athletes were also observed in the basal cortisol levels. Elevated resting cortisol levels in AMc in the present study are in agreement with several previous reports.2”‘4 Loucks et a124 noted increased urinary cortisol excretion in AMc subjects and inferred that the elevated levels were due to increased secretion. These investigators implied that there was a resetting of the hypothalamic-pituitary axis. The secretion of adrenocorticotropin was normal because of the reciprocal effects of increased negative feedback of cortisol and increased corticotropin-releasing hormone. Differences in the pattern of cortisol response were observed between the AMc and EUc athletes. AMc athletes had an initial increase followed by a gradual return to resting levels, whereas EUc athletes showed a slight decrease throughout exercise. Research has shown that plasma cortisol concentration decreases during light physical exercise (