Sex Differences in Corticosteroids in Man M. SCHONESHOFER AND G. G. WAGNER Department of Internal Medicine, Klinikum Steglitz, Freie Universitat Berlin, Berlin, West Germany
tion of all steroids studied from one 3 ml serum sample. The mean steroid concentrations (ng/ldl) in FF were found to be as follows: P, 20.3; 17-OHP, 58; DOC, 3.8; B, 241; S, 20.7; 18-OH-DOC, 12.3; Aldo, 8.8; F, 9860. Apart from P, steroid concentrations in FF were significantly lower than in M. Serum 17-OHP, DOC, B and Aldo in FF were significantly lower than in FL. With the exception of P, there were no significant differences of mean steroid concentrations between M and FL. (J Clin Endocrinol Metab 45: 814, 1977)
ABSTRACT. The serum concentrations of progesterone (P), 17-OH-progesterone (17-OHP), 11-deoxycorticosterone (DOC), corticosterone (B), 11-deoxycortisol (S), 18-OH-ll-deoxycorticosterone (18-OHDOC), aldosterone (Aldo) and cortisol (F) were measured in 18 normal men (M), 14 normal women in the follicular phase (FF) and 14 normal women in the luteal phase (FL) of menstrual cycle from blood samples obtained between 0800 and 0900 h. Steroids were estimated by a sensitive and specific radioimmunoassay allowing the simultaneous determina-
T
HE SEX specificity of gestagens, estrogens and androgens in man is well documented. However, up to the present, information about sex differences in serum concentrations of glucoand mineralocorticoids is scanty and contradictor)' (1-5). In comparative studies described hitherto, a possible influence of the menstrual cycle on serum gluco- and mineralocorticosteroid concentrations in females was not taken into account. Such an influence of the menstrual cycle was previously reported for serum aldosterone (1). Accordingly, we have studied serum 11deoxycorticosterone (DOC), corticosterone (B), deoxycortisol (S), 18-OH-ll-deoxycorticosterone (18-OH-DOC), aldosterone (Aldo) and cortisol (F) in a series of men and of women in the follicular and luteal phase of the menstrual cycle.
Materials and Methods Patients All subjects were adult hospital staff members who had negative histories and physical exams. Women were menstruating normally and did not take contraceptive pills. The phase of cycle was established by history and serum progesterone (P) levels. P levels lower than 50 ng/ldl were considered to be evidence of the follicular phase, levels higher than 300 ng/ldl as evidence of the luteal phase. Received November 8, 1976. Reprint requests to: M. Schoneshofer, Dept. of Internal Medicine, Klinikum Steglitz, Freie Universitat Berlin, Hindenburgdamm 30, D-1000 Berlin 45, West Germany.
Serum samples Blood was drawn between 0800 and 0900 h from a cubital vein into plastic tubes. After clotting, the sample was centrifuged and the frozen serum stored at —20 C until assayed. Analytical methods Serum concentrations of P, 17-OH-progesterone (17-OHP), DOC, B, S, 18-OH-DOC, Aldo and F were determined by a simultaneous radioimmunological method (6). In brief, serum samples, to which radioactive tracer amounts of each steroid had been added, were extracted with organic solvents of increasing polarity. Thus, four fractions were obtained, each containing steroids of similar polarity, such as P in fraction I, 17-OHP and DOC in fraction II, B, S, and 18-OH DOC in fraction III, Aldo and F in fraction IV. After overnight paper chromatography of each fraction, steroids were located by radioscanning and eluted with methanol. Final quantitation was achieved by 3Hrecovery measurement and radioimmunological determination of steroid in aliquots of eluate. Mean sensitivity of standard curves (2 standard deviations of zero point) ranged from 2.9 pg for DOC to 7.4 pg for 18-OH-DOC. Within-assay variability ranged from 4.8% for P to 17.9% for 18-OH-DOC, between-assay variability from 7.6% for F to 25.9% for 18-OH-DOC. In order to reduce the influence of between-assay variability, an almost equivalent number of subjects of each group studied was assayed in a particular batch. The radioimmunoassay data were evaluated using a computer program (7) applying the "spline-approximation" technique as standard curve model (8). An IBM1800 computer was used for the operating procedures. Statistical evaluation of the data was performed by Student's t test.
814
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 04:50 For personal use only. No other uses without permission. . All rights reserved.
COMMENTS
Results Eighteen men (M), 14 women in the follicular phase (FF) and 14 women in the luteal phase (FL) were studied. All subjects were fasting and had been in an upright posture for some hours. During venipuncture the subjects were seated or lying. The means, standard deviations and ranges of serum steroid concentrations as well as statistical values of comparison are summarized in Table 1. P and 17 OH? Mean P and 17 OHP levels exhibited the known significant difference between FF and FL (9). While serum P levels were not significantly different between M and FF, 17 OHP levels were significantly higher in M than in FF. Inverse relations were found for the comparison of P and 17 OHP between M and FL. Corticosteroids Mean serum concentrations of all corticosteroids in FF exhibited the lowest values if compared with M and FL. They were significantly lower than in M. While mean serum concentrations of DOC, B and Aldo in FF were significantly lower than in FL, differences in mean serum S, 18-OH-DOC and F levels between FF and FL were insignificant. No significant differences of mean serum corticosteroid concentrations between M and FL could be demonstrated.
Discussion All steroids studied are secreted, at least in part, by the adrenal cortex as judged by the increase after ACTH stimulation (10,11). The sex dependence of serum P and 17 OHP concentrations and their variation during the menstrual cycle are well explained by the additional sources of these steroids in the ovaries (12) and testes (13). DOC, B, 18-OH-DOC, S, Aldo and F are assumed to be secreted exclusively by the adrenal cortex. So far, no unequivocal differences between the sexes were demonstrable for these steroids. The difficulty of detecting sex differences is attributable to several reasons: 1) Very low serum concentrations of DOC, 18-OH-DOC, S and Aldo, which require sensitive, specific and precise methods of estimation, 2) The variability of serum corticosteroid concentrations caused by the episodic secretion pattern of the adrenal cortex and 3) The
815
influence of menstrual cycle on corticosteroid secretion pattern in women. As described in detail elsewhere (6), the present assay techniques were sufficiently sensitive to allow the precise estimation of normal serum steroid concentrations in the medium range of radioimmunoassay standard curves. Specificity was achieved by nearly complete chromatographic separation and, in addition, by the specificity of antisera used. Serum concentrations of DOC and S determined in the recent study are the lowest; those of P, 17-OHP, B, F and Aldo are in the lower range of normal values reported hitherto. Serum 18-OH-DOC levels are similar to those recently reported by Chandler et al. (14). Sex differences of the serum corticosteroids reported in the present study were significant only if men were compared with women in the follicular phase. An influence of menstrual cycle on serum concentrations of aldosterone has previously been demonstrated by Sundsfjord and Aakvaag (1), as well as by Katz and Romfh (15). They found a significant increase in peripheral Aldo levels during the luteal phase of the cycle which was interpreted as a compensatory response to the natriuretic action of progesterone (15). These findings are confirmed by the present study and by the previous finding of a marked increase in aldosterone excretion following progesterone administration to healthy men (16). When Nabors et al. (2) compared plasma concentrations of B in men and women without considering the phase of the menstrual cycle, higher values were found in women than in men, thus contrasting to the results of the present study. Schwartz and Abraham observed higher plasma levels of B in the luteal phase than in the follicular phase of the menstrual cycle (17), a finding which is supported by the present results. Zumoff et al. (3) reported significantly higher plasma levels of F in men than in women, which is confirmed by our results in M and FF. It must be pointed out that F concentrations in women in the luteal phase, in contrast to the other corticosteroids, are still lower, althought not significantly, than in men. In a previous study, significant sex differences were found in this laboratory for serum DOC using radioimmunoassay (5). Sex differences of S and 18-OH-DOC have not yet been reported. The mechanisms by which sex or the menstrual cycle affect peripheral corticosteroid concentrations have not yet been elucidated, although different mechanisms have already been
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 04:50 For personal use only. No other uses without permission. . All rights reserved.
816
JCE & M • 1977 Vol 45 • No 4
COMMENTS TABLE 1. Serum concentrations of adrenal steroids in normal males and females Males (M) mean ± SD range (ng/dl) comparison
Steroid
Females (FF) mean ± SD range (ng/dl) comparison
Females (FL) mean ± SD range (ng/dl) comparison
Progesterone
17.9 ± 10.2 4.8-47.9 M-FF (NS)
20.3 ± 8.6 7.7-42.1 M-FL(P< 0.001)
1,184 ± 696 365-3,000 F F - F L (P < 0.001)
17-OH-progesterone
179 ± 61 73-317 M-FF ( P < 0.001)
58 ±26 20-96 M-FL (NS)
196 ± 75 83-340 F F - F L (P < 0.001)
Deoxycorticosterone
6.65 ± 2.6 3.1-13.7 M - F F ( P < 0.002)
3.8 ± 1.7 2.4-8.4 M-FL (NS)
6.6 ± 2.2 4.1-11.5 F F - F L ( P < 0.002)
Corticosterone
421 ± 217 85-915 M-FF ( P < 0.01)
241 ± 148 57-540 M-FL(NS)
439 ± 232 175-1927 F F - F L (P < 0.01)
Deoxycortisol
49.9 ± 27.3 15.2-103.6 M-FF ( P < 0.002)
20.7 ± 15.2 4.6-50.4 M-FL (NS)
33.5 ± 18.4 13.7-73.4 F F - F L (NS)
18-0H-D0C
20.9 ± 11.0 9.3-51.5 M-FF (P
tion of all steroids studied from one 3 ml serum sample. The mean steroid concentrations (ng/ldl) in FF were found to be as follows: P, 20.3; 17-OHP, 58; DOC, 3.8; B, 241; S, 20.7; 18-OH-DOC, 12.3; Aldo, 8.8; F, 9860. Apart from P, steroid concentrations in FF were significantly lower than in M. Serum 17-OHP, DOC, B and Aldo in FF were significantly lower than in FL. With the exception of P, there were no significant differences of mean steroid concentrations between M and FL. (J Clin Endocrinol Metab 45: 814, 1977)
ABSTRACT. The serum concentrations of progesterone (P), 17-OH-progesterone (17-OHP), 11-deoxycorticosterone (DOC), corticosterone (B), 11-deoxycortisol (S), 18-OH-ll-deoxycorticosterone (18-OHDOC), aldosterone (Aldo) and cortisol (F) were measured in 18 normal men (M), 14 normal women in the follicular phase (FF) and 14 normal women in the luteal phase (FL) of menstrual cycle from blood samples obtained between 0800 and 0900 h. Steroids were estimated by a sensitive and specific radioimmunoassay allowing the simultaneous determina-
T
HE SEX specificity of gestagens, estrogens and androgens in man is well documented. However, up to the present, information about sex differences in serum concentrations of glucoand mineralocorticoids is scanty and contradictor)' (1-5). In comparative studies described hitherto, a possible influence of the menstrual cycle on serum gluco- and mineralocorticosteroid concentrations in females was not taken into account. Such an influence of the menstrual cycle was previously reported for serum aldosterone (1). Accordingly, we have studied serum 11deoxycorticosterone (DOC), corticosterone (B), deoxycortisol (S), 18-OH-ll-deoxycorticosterone (18-OH-DOC), aldosterone (Aldo) and cortisol (F) in a series of men and of women in the follicular and luteal phase of the menstrual cycle.
Materials and Methods Patients All subjects were adult hospital staff members who had negative histories and physical exams. Women were menstruating normally and did not take contraceptive pills. The phase of cycle was established by history and serum progesterone (P) levels. P levels lower than 50 ng/ldl were considered to be evidence of the follicular phase, levels higher than 300 ng/ldl as evidence of the luteal phase. Received November 8, 1976. Reprint requests to: M. Schoneshofer, Dept. of Internal Medicine, Klinikum Steglitz, Freie Universitat Berlin, Hindenburgdamm 30, D-1000 Berlin 45, West Germany.
Serum samples Blood was drawn between 0800 and 0900 h from a cubital vein into plastic tubes. After clotting, the sample was centrifuged and the frozen serum stored at —20 C until assayed. Analytical methods Serum concentrations of P, 17-OH-progesterone (17-OHP), DOC, B, S, 18-OH-DOC, Aldo and F were determined by a simultaneous radioimmunological method (6). In brief, serum samples, to which radioactive tracer amounts of each steroid had been added, were extracted with organic solvents of increasing polarity. Thus, four fractions were obtained, each containing steroids of similar polarity, such as P in fraction I, 17-OHP and DOC in fraction II, B, S, and 18-OH DOC in fraction III, Aldo and F in fraction IV. After overnight paper chromatography of each fraction, steroids were located by radioscanning and eluted with methanol. Final quantitation was achieved by 3Hrecovery measurement and radioimmunological determination of steroid in aliquots of eluate. Mean sensitivity of standard curves (2 standard deviations of zero point) ranged from 2.9 pg for DOC to 7.4 pg for 18-OH-DOC. Within-assay variability ranged from 4.8% for P to 17.9% for 18-OH-DOC, between-assay variability from 7.6% for F to 25.9% for 18-OH-DOC. In order to reduce the influence of between-assay variability, an almost equivalent number of subjects of each group studied was assayed in a particular batch. The radioimmunoassay data were evaluated using a computer program (7) applying the "spline-approximation" technique as standard curve model (8). An IBM1800 computer was used for the operating procedures. Statistical evaluation of the data was performed by Student's t test.
814
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 04:50 For personal use only. No other uses without permission. . All rights reserved.
COMMENTS
Results Eighteen men (M), 14 women in the follicular phase (FF) and 14 women in the luteal phase (FL) were studied. All subjects were fasting and had been in an upright posture for some hours. During venipuncture the subjects were seated or lying. The means, standard deviations and ranges of serum steroid concentrations as well as statistical values of comparison are summarized in Table 1. P and 17 OH? Mean P and 17 OHP levels exhibited the known significant difference between FF and FL (9). While serum P levels were not significantly different between M and FF, 17 OHP levels were significantly higher in M than in FF. Inverse relations were found for the comparison of P and 17 OHP between M and FL. Corticosteroids Mean serum concentrations of all corticosteroids in FF exhibited the lowest values if compared with M and FL. They were significantly lower than in M. While mean serum concentrations of DOC, B and Aldo in FF were significantly lower than in FL, differences in mean serum S, 18-OH-DOC and F levels between FF and FL were insignificant. No significant differences of mean serum corticosteroid concentrations between M and FL could be demonstrated.
Discussion All steroids studied are secreted, at least in part, by the adrenal cortex as judged by the increase after ACTH stimulation (10,11). The sex dependence of serum P and 17 OHP concentrations and their variation during the menstrual cycle are well explained by the additional sources of these steroids in the ovaries (12) and testes (13). DOC, B, 18-OH-DOC, S, Aldo and F are assumed to be secreted exclusively by the adrenal cortex. So far, no unequivocal differences between the sexes were demonstrable for these steroids. The difficulty of detecting sex differences is attributable to several reasons: 1) Very low serum concentrations of DOC, 18-OH-DOC, S and Aldo, which require sensitive, specific and precise methods of estimation, 2) The variability of serum corticosteroid concentrations caused by the episodic secretion pattern of the adrenal cortex and 3) The
815
influence of menstrual cycle on corticosteroid secretion pattern in women. As described in detail elsewhere (6), the present assay techniques were sufficiently sensitive to allow the precise estimation of normal serum steroid concentrations in the medium range of radioimmunoassay standard curves. Specificity was achieved by nearly complete chromatographic separation and, in addition, by the specificity of antisera used. Serum concentrations of DOC and S determined in the recent study are the lowest; those of P, 17-OHP, B, F and Aldo are in the lower range of normal values reported hitherto. Serum 18-OH-DOC levels are similar to those recently reported by Chandler et al. (14). Sex differences of the serum corticosteroids reported in the present study were significant only if men were compared with women in the follicular phase. An influence of menstrual cycle on serum concentrations of aldosterone has previously been demonstrated by Sundsfjord and Aakvaag (1), as well as by Katz and Romfh (15). They found a significant increase in peripheral Aldo levels during the luteal phase of the cycle which was interpreted as a compensatory response to the natriuretic action of progesterone (15). These findings are confirmed by the present study and by the previous finding of a marked increase in aldosterone excretion following progesterone administration to healthy men (16). When Nabors et al. (2) compared plasma concentrations of B in men and women without considering the phase of the menstrual cycle, higher values were found in women than in men, thus contrasting to the results of the present study. Schwartz and Abraham observed higher plasma levels of B in the luteal phase than in the follicular phase of the menstrual cycle (17), a finding which is supported by the present results. Zumoff et al. (3) reported significantly higher plasma levels of F in men than in women, which is confirmed by our results in M and FF. It must be pointed out that F concentrations in women in the luteal phase, in contrast to the other corticosteroids, are still lower, althought not significantly, than in men. In a previous study, significant sex differences were found in this laboratory for serum DOC using radioimmunoassay (5). Sex differences of S and 18-OH-DOC have not yet been reported. The mechanisms by which sex or the menstrual cycle affect peripheral corticosteroid concentrations have not yet been elucidated, although different mechanisms have already been
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 04 July 2015. at 04:50 For personal use only. No other uses without permission. . All rights reserved.
816
JCE & M • 1977 Vol 45 • No 4
COMMENTS TABLE 1. Serum concentrations of adrenal steroids in normal males and females Males (M) mean ± SD range (ng/dl) comparison
Steroid
Females (FF) mean ± SD range (ng/dl) comparison
Females (FL) mean ± SD range (ng/dl) comparison
Progesterone
17.9 ± 10.2 4.8-47.9 M-FF (NS)
20.3 ± 8.6 7.7-42.1 M-FL(P< 0.001)
1,184 ± 696 365-3,000 F F - F L (P < 0.001)
17-OH-progesterone
179 ± 61 73-317 M-FF ( P < 0.001)
58 ±26 20-96 M-FL (NS)
196 ± 75 83-340 F F - F L (P < 0.001)
Deoxycorticosterone
6.65 ± 2.6 3.1-13.7 M - F F ( P < 0.002)
3.8 ± 1.7 2.4-8.4 M-FL (NS)
6.6 ± 2.2 4.1-11.5 F F - F L ( P < 0.002)
Corticosterone
421 ± 217 85-915 M-FF ( P < 0.01)
241 ± 148 57-540 M-FL(NS)
439 ± 232 175-1927 F F - F L (P < 0.01)
Deoxycortisol
49.9 ± 27.3 15.2-103.6 M-FF ( P < 0.002)
20.7 ± 15.2 4.6-50.4 M-FL (NS)
33.5 ± 18.4 13.7-73.4 F F - F L (NS)
18-0H-D0C
20.9 ± 11.0 9.3-51.5 M-FF (P
