0021-972X/92/7406-1432$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright 0 1992by The Endocrine Society
Vol.
Printed
An Abnormality of the Growth Hormone/Insulin-Like Growth Factor-I Axis in Women with Polycystic Syndrome Due to Coexistent Obesity* J. SLOWINSKA-SRZEDNICKA, W. ZGLICZYNSKI, A. MAKOWSKA, A. BRZEZINSKA, P. SOSZYNSKI, AND S. ZGLICZYNSKI Department
of Endocrinology
Medical
Center for Postgraduate
Education,
ABSTRACT. In order to evaluate the GH/insulin-like growth factor-I (IGF-I) axis in the polycystic ovary syndrome (PCO), 21 women aged 18-38 yr were studied. The GH responses to the GH-releasing hormone (GHRH), and plasma concentrations of IGF-I were measured in seven obese women with PCO, seven obese healthv controls without PCO, and in seven nonobese subjects. ” Total GH secretion, as expressed by the integrated GH resnonse to GHRH. in PC0 obese women (617.4 f 150 uelL.min) and in obese women without PC0 (32i.l f 161.4 hgjL.minj were lower than that in nonobese healthy controls (3181.4 f 644.3 pg/L. min, P < 0.001 and P < 0.001, respectively). Plasma
T
HE pituitary GH secretion is under hypothalamic control via the stimulatory action of the GH-releasing hormone (GHRH) and the inhibitory action of somatostatin (1). The GH promotes the generation of the insulin-like growth factor-I (IGF-I) which exerts a negative feedback effect on GH release (2). Besides the GH, insulin, sex hormones, and nutrition are important regulators of serum IGF-I levels (3). The polycystic ovary syndrome (PCO) is characterized by oligomenorrhea or amenorrhea, anovulation, hirsutism, and bilateral cystic ovaries (4). The main feature of this syndrome lies in excessive production of androgens, insulin resistance, and hyperinsulinemia (5, 6). A very common finding in PC0 women is obesity (4,6). Hyperinsulinemia and obesity may have an influence on the GH/IGF-I axis. Recently, Kazer et al. (7) have demonstrated that serum GH levels were lower in women with PC0 than in healthy subjects. The aim of this study was to determine the GH response to GHRH stimulation and its ReceivedJuly 29, 1991. Address requests for reprints to: J. Slowinska-Srzednicka, M.D., Department of Endocrinology, Bielanski Hospital, Ceglowska 80, Ol609 Warsaw, Poland. * The work was supported by CMKP 16/91 Grant.
74, No. 6
in U.S.A.
Ovary
W. JESKE,
Warsaw, Poland
concentrations of IGF-I in obese PC0 women (199.5 + 39.1 ag/ L), and in obese women without PC0 (192.4 f 36.8 lg/L) were similar to the IGF-I levels in nonobese controls (224.3 + 33.2 /dJ. In obese women with and without PCO, a negative correlation was found between the body mass index and the peak GH responses to GHRH (r = -0.639, P < 0.02) and between age and IGF-I levels (r = -0.520, P c 0.05). These findings suggest that an abnormality of the GH/IGF-I axis in PC0 women may be due to coexistent obesity. (J Clin Endocrinol Metab 74: 1432-1435,1992)
relationship to plasma with PC0 and obesity. Material
IGF-I
concentration
in women
and Methods
Patients The study was approved by the ethical committee of the Medical Center for Postgraduate Education. Women were recruited from the population of the Outpatient Clinic of Endocrinology, and gave fully informed consent for the investigation.
Seven obese women [body mass index (BMI) between 29.741.2 kg/m*] aged 18-38 yr, with clinical and biochemical features of PC0 were studied. All had oligomenorrhea,
hirsutism,
elevated plasma levels of androstenedione and testosterone, the LH:FSH
ratio of 2 or more, and enlarged
ultrasonogaphy.
polycystic
ovaries on
The late-onset of 21-hydroxylase deficiency
was excluded because baseline plasma 17-hydroxyprogesterone level was found to be less than than 5 nmol/L and less than 10 nmol/L 60 min after 250 pg Synacthen given im. The Cushing’s syndrome and the androgen secretion tumor were excluded by measuring the circadian rhythm of plasma cortisol the excretion of glucocorticoid metabolites in urine, and plasma concentrations of cortisol, testosterone, androstenedione, and DHEAs before and after the dexamethasone suppresion test. The control group comprized of 14 normally ovulating women without PCO, aged 18-38 yr: 7 with obesity (BMI
between 30.0-44.0 kg/m’) and 7 nonobese subjects (BMI between
21.0-23.3
kg/m’)
(see Table
1). None had any clinical
1432
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
GH/IGF-I
AXIS
IN PC0 AND
evidence of hyperandrogenemia and all had normal plasma levels of testosterone and androstenedione, and normal ovaries on ultrasonography. Their glucose tolerance test was normal, according to the criteria of the National Diabetes Data Group. In women with known menstrual cycles the studies were performed in the midfollicular phase on days 6-9 of the cycle.
OBESITY
1433
identified as the highest single GH measurement for each subject after GHRH injection. Group mean values were compared using the analysis of variance. When F ratios showed significant differences, the means between groups were further analyzed using Duncan’s multiple range test. Correlations between age, BMI, and hormone concentrations were examined using a simple regression analysis.
Study design All women were admitted to the Department of Endocrinology, Medical Center for Postgraduate Education. The GHRH stimulation test was done in the morning, after an overnight fast of 14 h. A catheter was inserted into an antecubital vein 30 min before the first sampling. Each woman received an iv injection of a bolus of 100 pg synthetic GHRH (GRF l-29NH Serono, Geref 100). Blood samples were obtained for GH determination at -15, 0, 15, 30, 45, 60, 90, and 120 min. Blood samples for IGF-I, testosterone (T), androstenedione (A), dehydroepiandrosterone-sulfate (DHEA-S), sex hormone binding globulin (SHBG), glucose and insulin determinations were taken before GHRH injection. In all subjects, the oral glucose tolerance test (OGTT) was done on the second day after the GHRH test, as well. Blood samples were collected before and 30, 60, 120, 150, and 180 min after a standard glucose load (75 d. Assays Serum glucose concentrations were determined by the glucose oxidase method (Abbott, Spectrum). Commercial radioimmunological kits were used for plasma measurements of insulin, GH (Institute of Atomic Energy, Poland), and androstenedione (Wien Laboratories). Plasma levels of 17-hydroxyprogesterone were measured by RIA using Medgenix kits (Belgium). Neither extraction nor chromatography were required because of high specificity of the coated antibodies. The plasma concentrations of DHEA-S were estimated by RIA using Immunotech S.A. kits (France) with monoclonal antibodycoated tubes. Without previous extraction or chromatography plasma concentrations of testosterone were determined by RIA with Farmos Diagnostica kits (Finland). The antibodies used were specific, the cross-reactions with 5-cw-dihydrotestosterone and with 5-fl-dihydrotestosterone were 17.7% and 10.2%, respectively. Plasma IGF-I and SHBG concentrations were measured by an immunoradiometric method using kits obtained from BykSangtec (Germany) and Farmos Diagnostica (Finland), respectively. The IGF-I levels were estimated after preliminary acid extraction of plasma. The intraassay coefficients of variation were: insulin < 7.0%; GH < 4.2%; T < 6.6%; A < 10.9%; DHEA-S < 5.0%; IGF-I < 7.5% and SHBG < 4.9%. The interassay coefficients of variation were: insulin C 11.8%; GH < 10.8%; T < 6.4%; A C 12.5%; DHEA-S < 10.6%; IGF-I < 12% and SHBG < 10%. Statistical methods The integrated 2-h GH response to GHRH, and 3-h insulin response to glucose load were calculated as the areas under the curve by the trapezoid rule. The GH peak responses were
Results The subjects were of similar age (Table 1). The BMI values were higher in obese women than those in nonobese controls. No difference was observed in BMI between these two obese groups. In PC0 obese women the insulin response to glucose load and the plasma levels of testosterone and androstenedione were higher, but the plasma concentrations of SHBG were lower than those in obese women without PC0 and in nonobese controls. No differences in DHEA-S concentrations were found between groups. In PC0 obese women a positive correlation between fasting insulin and testosterone levels (r = 0.890, P < 0.001) and a negative correlation between fasting insulin and SHBG concentrations (r = -0.820, P < 0.01) were found. In all obese women a negative correlation between the integrated insulin response to glucose load and SHBG concentrations (r = 0.618, P < 0.02) was noted. The GH response to GHRH is shown in Fig. 1 and Table 2. No significant differences were found in the basal GH levels between groups. Total GH secretion, as expressed by the integrated GH response to GHRH, in PC0 women and in obese women without PC0 were lower than that in nonobese subjects. In all obese women, negative correlations were found between BMI and the peak GH responses (r = -0.639, P < 0.02) and between age and IGF-I levels (r = -0.520, p < 0.02). No significant correlations were found between IGF-I levels and BMI, insulin, T, A, and the GH responses to GHRH. Despite decreased GH secretion in subjects with obesity, plasma concentrations of IGF-I in obese PC0 women and in obese women without PC0 were similar to the IGF-I levels in nonobese controls. Discussion In our study obese women with PC0 showed a significantly lower GH response to GHRH than nonobese healthy controls. However, no statistically significant difference was observed in the GH response to GHRH between obese women with PC0 and those without PCO, both groups of obese subjects being comparable with respect to age and degree of obesity. Recently, Kazer et al. (7) determined GH concentrations over a period of 14 h by collecting blood samples
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
SLOWINSKA-SRZEDNICKA
1434 TABLE
JCE&M*1992 Vo174*No6
1. Characteristics and plasma hormone levels in obese women (with and without PCO) and in nonobese healthy controls Obese PC0 women n = 7
31.0 f 36.1 f BMI (kg/m*) 4.47 f Fasting glucose (mmol/L) 195.9 f Fasting insulin (pmol/L) Integrated insulin area 101856.3 f (pmol/L . min) 9.16 f Testosterone (nmol/L) 5.22 -t DHEA-S (hmol/L) SHBG (nmol/L) 15.4 f 8.94 f Androstenedione (nmol/L) Values are mean f SEM. a Determined by analysis of variance. *P < 0.001 us. nonobese healthy controls. c P < 0.01 us. obese women and VS. nonobese d P < 0.05 vs. obese women.
Age (yr)
3.2 1.5* 0.17 32.3’ 16093.5b~d
Obese women n = 7 28.2 34.4 4.22 179.4 70432.6
1.3’ 0.9 3.2b*d 0.95’
3.01 3.26 29.1 5.67
f f f f k
2.0 2.2” 0.16 32.3* 4079.7*
f 0.27 rf: 0.3 f 4.4b zk 0.73
Nonobese healthy controls n = 7
P value”
NS
28.0 21.0 3.91 70.7 28750.2
f + f f f
2.2 0.56 0.16 5.7 3171.4
0.0001 NS 0.01 0.002
3.15 3.47 63.0 4.73
f f r f
0.3 0.4 9.4 0.93
0.002 NS 0.0001 0.002
healthy controls.
GE h/l1
time rmini hc. 1. The GH response to 100 pg GHRH in PC0 obese women (x- -x), in obese women without PC0 (*- - -*), in nonobese healthy controls (+- - -+).
between 0900 and 2300 h, and showed that the GH secretion was decreased in PC0 women. In our study, we have demonstrated that the reduced GH secretion may be due to coexistent obesity present in over 50% of PC0 women (5), rather than to the PCO. In accordance with previous observations plasma levels of IGF-I in obese women with (7-10) and without PC0 (11, 12) were also not significantly different from the TABLE
ET AL.
values in nonobese controls. Moreover, no correlation was observed between IGF-I levels and GH response to GHRH. It is well known that the plasma concentration of IGFI depends on age, GH, nutrition, insulin, and sex hormones (13). At puberty, the IGF-I level increases steeply as a result of increased secretion of GH and sex hormones (14). With aging IGF-I concentration decreases with decreasing GH secretion (3). During childhood, puberty, and in adult life there is a strong relationship between plasma insulin and IGF-I (14). In addition, in vitro studies suggest that insulin enhances IGF-I production (15) by either direct regulation of the GH receptor (16) or by a permissive effect on post GH receptor events (17). We have demonstrated that insulin secretion in obese women after an oral glucose load is significantly higher than in nonobese women. Hyperinsulinism observed in obesity, particularly evident in women with ovarian androgenization in PCO, may influence IGF-I concentration. The biological action of IGF-I is thought to be partly determined by the IGF small binding protein (IGFBPI). It is also known that insulin affects the production and circulating levels of IGFBP-I (18) and SHBG (19). In this study, in agreement with previous observations
2. Changes in plasma GH and IGF-I levels after GHRH administration
in obese women (with and withou PCO) and in nonobese healthy
controls Obese PC0 women n = 7 Baseline GH (pg/L) GH peak post GHRH (fig/L) Integrated GH area (pg/L . min) IGF-I (air/L)
1.0 13.5 617.4 199.5
f f f +
0.4 3.9” 150* 39.1
Obese women n = 7 0.88 10.4 327.1 1924
f f + f
0.2 3* 161.4* 36.8
Nonobsse healthy controls n = 7 3.2 47.8 3181.4 224.3
f f & k
1.4 9.7 644.3 33.2
VdueS are means f SEM. 0 Determined by analysis of variance. * P < 0.001 us. healthy controls.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
P value” NS 0.001 0.0001 NS
GH/IGF-I
AXIS IN PC0 AND OBESITY
(12, 20), obese women showed lower plasma levels of SHBG than the nonobese controls, and, in particular, low concentrations of SHBG were demonstrated in obese women with PCO. It has been reported elsewhere that in extreme obesity, IGFBP-I and SHBG concentrations are decreased and the levels of these proteins are negatively correlated with insulin concentrations (12). In obese PC0 women, the levels of IGFBP-I have been found to decrease (9,21). On a low-calorie diet in women with PCO, increased concentrations of IGFBP-I were associated with concomitant lower insulin concentrations (8). It may be that decreased IGFBP-I levels in hyperinsulinemia in women with PC0 lead to an increase of the biological action of IGF-I which, in turn, inhibits GH secretion by the pituitary. In conclusion: in obese women with and without PCO, the GH response to GHRH is lower that in nonobese healthy controls. No difference, however, was observed in GH secretion in obese PC0 women as compared with obese women without PCO. Despite decreased GH secretion plasma levels of IGF-I were similar in obese and nonobese patients. No relationship was found between the GH response and the IGF-I levels. These findings suggest that the abnormality in the physiological negative feedback between GH and IGF-I in PC0 women is related to the coexistent obesity. Acknowledgment The authors are indebted to Oskar A. Chomicki M.Sc. for his valuable suggestions and improvement of the English text.
References 1. Frohman LA, Jansson JO. Growth hormone-releasing hormone. Endocr Rev. 1986$223-53. 2. Berelowitz M, Szabo M, Frohman LA, Firestone S, Chu L, Hintz RL. Somatomedin C mediates growth hormone negative feed-back bv effects on both the h.vpothalamus and the pituitary. Science. 1981;212:1279-81. -3. Hall K. Tallv M. The somatomedin-insulin-like aowth factors. J Internal Med. 1989;225:47-54. 4. Franks S. Polycystic ovary syndrome: a changing perspective. Clin Endocrinol (Oxf). 1989;31:87-120. 5. Nestler JE, Clore JN, Blackard WG. The central role of obesity (hyperinsulinemia) in the pathogenesis of polycystic ovary syn-
1435
drome. Am J Obstet Gynecol. 1989;161:1095-7. 6. Slowinska-Srzednicka J, Zgliczynski S, Wierzbicki M, et al. The role of hyperinsulinemia in the development of lipid disturbances in non obese and obese women with the poiycystic ovary syndrome. J Endocrinol Invest. 1991;14:569-72. 7. Kazer RR, Unterman TG, Glick RP. An abnormality of the growth hormone/insulin-like growth factor-I axis in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 1990;71:958-62. 8. Kiddy DS, Hamilton-Fairley D, Seppala M, et al. Diet-induced changes in sex hormone binding globulin and free testosterone in women with normal or polycystic ovaries: correlation with serum insulin and insulin-like growth factor-I. Clin Endocrinol (Oxf’). 1989;31:757-63. 9. Conway GS, Jacobs HS, Holly JMP, Wass JAH. Effects of luteinizing hormone, insulin, insulin-like growth factor-I and insulinlike-growth factor small binding protein I in the polycystic ovary svndrome. Clin Endocrinol (Oxn. 1990:33:593-603. 10. Lanzone A, Fulghesu AM, Pappalardo’ S, et al. Growth hormone and somatomedin C secretion-in patients with polycystic ovarian disease. Their relationships with hyperinsulinism and hyperandrogenism. Gynecol Obstet Invest. 1990;29:149-53. 11. Kelliman M. Frohman LA. Enhanced erowth hormone (GH) responsiveness to GH-releasing hormone after dietary manipulation in obese and non obese subjects. J Clin Endocrinol Metab. 1988;66:489-94. 12. Weaver JU, Holly JM, Kopelman PG, et al. Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IGFBP-I) in extreme obesity. Clin Endocrinol (Oxf). 1990;33:415-22. 13. Froesch ER, Zapf J. Insulin-like growth factors and insulin: comparative aspects. Diabetologia. 1985;28:485-93. 14. Smith CP, Dunger D, Williams A, et al. Relationship between insulin, insulin-like growth factor I, and dehydroepiandrosterone sulfate concentrations during childhood, puberty, and adult life. J Clin Endocrinol Metab. 1989;68:932-37. 15. Daughaday WH, Phillips LS, Mueller MS. The effects of insulin and growth hormone on the release of somatomedin by the isolated rat liver. Endocrinology. 1976;98:1214-9. 16. Baxter RC, Bryson JM, Turtle JR. Somatogenic receptors of rat liver: regulation by insulin. Endocrinology. 1980;107:1i76-81. 17. Maes M. Underwood LE. Ketelslegers JM. Low serum somatomedin-C in insulindependent diabetes: evidence for a postreceptor mechanism. Endocrinology. 1986;118:377-82. 18. Suikkari AM, Koivisto A, Koistinen R, Seppala M, Yki-Jarvinen H. Dose-response characteristics for suppresion of low molecular weight plasma insulin-like growth factor-binding protein by insulin. J Clin Endocrinol Metab. 1989;68:135-140. 19. Plymate SR, Matej LA, Jones RE, Friedl KE. Inhibition of sex hormone-binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. J Clin Endocrinol Metab. 1988,67:460-4. 20. Pasquali R, Casimirri F, Plate L, Capelli M. Characterization of obese women with reduced sex hormone-binding globulin concentrations. Horm Metab Res. 1990;22:303-6. 21. Suikkari AM, Ruutiainen K, Erkkola R, Seppala M. Low levels of low molecular weight insulin-like growth factor binding protein in patients with polycystic ovary disease. Hum Reprod. 1989;4:136-9.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
Vol.
Printed
An Abnormality of the Growth Hormone/Insulin-Like Growth Factor-I Axis in Women with Polycystic Syndrome Due to Coexistent Obesity* J. SLOWINSKA-SRZEDNICKA, W. ZGLICZYNSKI, A. MAKOWSKA, A. BRZEZINSKA, P. SOSZYNSKI, AND S. ZGLICZYNSKI Department
of Endocrinology
Medical
Center for Postgraduate
Education,
ABSTRACT. In order to evaluate the GH/insulin-like growth factor-I (IGF-I) axis in the polycystic ovary syndrome (PCO), 21 women aged 18-38 yr were studied. The GH responses to the GH-releasing hormone (GHRH), and plasma concentrations of IGF-I were measured in seven obese women with PCO, seven obese healthv controls without PCO, and in seven nonobese subjects. ” Total GH secretion, as expressed by the integrated GH resnonse to GHRH. in PC0 obese women (617.4 f 150 uelL.min) and in obese women without PC0 (32i.l f 161.4 hgjL.minj were lower than that in nonobese healthy controls (3181.4 f 644.3 pg/L. min, P < 0.001 and P < 0.001, respectively). Plasma
T
HE pituitary GH secretion is under hypothalamic control via the stimulatory action of the GH-releasing hormone (GHRH) and the inhibitory action of somatostatin (1). The GH promotes the generation of the insulin-like growth factor-I (IGF-I) which exerts a negative feedback effect on GH release (2). Besides the GH, insulin, sex hormones, and nutrition are important regulators of serum IGF-I levels (3). The polycystic ovary syndrome (PCO) is characterized by oligomenorrhea or amenorrhea, anovulation, hirsutism, and bilateral cystic ovaries (4). The main feature of this syndrome lies in excessive production of androgens, insulin resistance, and hyperinsulinemia (5, 6). A very common finding in PC0 women is obesity (4,6). Hyperinsulinemia and obesity may have an influence on the GH/IGF-I axis. Recently, Kazer et al. (7) have demonstrated that serum GH levels were lower in women with PC0 than in healthy subjects. The aim of this study was to determine the GH response to GHRH stimulation and its ReceivedJuly 29, 1991. Address requests for reprints to: J. Slowinska-Srzednicka, M.D., Department of Endocrinology, Bielanski Hospital, Ceglowska 80, Ol609 Warsaw, Poland. * The work was supported by CMKP 16/91 Grant.
74, No. 6
in U.S.A.
Ovary
W. JESKE,
Warsaw, Poland
concentrations of IGF-I in obese PC0 women (199.5 + 39.1 ag/ L), and in obese women without PC0 (192.4 f 36.8 lg/L) were similar to the IGF-I levels in nonobese controls (224.3 + 33.2 /dJ. In obese women with and without PCO, a negative correlation was found between the body mass index and the peak GH responses to GHRH (r = -0.639, P < 0.02) and between age and IGF-I levels (r = -0.520, P c 0.05). These findings suggest that an abnormality of the GH/IGF-I axis in PC0 women may be due to coexistent obesity. (J Clin Endocrinol Metab 74: 1432-1435,1992)
relationship to plasma with PC0 and obesity. Material
IGF-I
concentration
in women
and Methods
Patients The study was approved by the ethical committee of the Medical Center for Postgraduate Education. Women were recruited from the population of the Outpatient Clinic of Endocrinology, and gave fully informed consent for the investigation.
Seven obese women [body mass index (BMI) between 29.741.2 kg/m*] aged 18-38 yr, with clinical and biochemical features of PC0 were studied. All had oligomenorrhea,
hirsutism,
elevated plasma levels of androstenedione and testosterone, the LH:FSH
ratio of 2 or more, and enlarged
ultrasonogaphy.
polycystic
ovaries on
The late-onset of 21-hydroxylase deficiency
was excluded because baseline plasma 17-hydroxyprogesterone level was found to be less than than 5 nmol/L and less than 10 nmol/L 60 min after 250 pg Synacthen given im. The Cushing’s syndrome and the androgen secretion tumor were excluded by measuring the circadian rhythm of plasma cortisol the excretion of glucocorticoid metabolites in urine, and plasma concentrations of cortisol, testosterone, androstenedione, and DHEAs before and after the dexamethasone suppresion test. The control group comprized of 14 normally ovulating women without PCO, aged 18-38 yr: 7 with obesity (BMI
between 30.0-44.0 kg/m’) and 7 nonobese subjects (BMI between
21.0-23.3
kg/m’)
(see Table
1). None had any clinical
1432
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
GH/IGF-I
AXIS
IN PC0 AND
evidence of hyperandrogenemia and all had normal plasma levels of testosterone and androstenedione, and normal ovaries on ultrasonography. Their glucose tolerance test was normal, according to the criteria of the National Diabetes Data Group. In women with known menstrual cycles the studies were performed in the midfollicular phase on days 6-9 of the cycle.
OBESITY
1433
identified as the highest single GH measurement for each subject after GHRH injection. Group mean values were compared using the analysis of variance. When F ratios showed significant differences, the means between groups were further analyzed using Duncan’s multiple range test. Correlations between age, BMI, and hormone concentrations were examined using a simple regression analysis.
Study design All women were admitted to the Department of Endocrinology, Medical Center for Postgraduate Education. The GHRH stimulation test was done in the morning, after an overnight fast of 14 h. A catheter was inserted into an antecubital vein 30 min before the first sampling. Each woman received an iv injection of a bolus of 100 pg synthetic GHRH (GRF l-29NH Serono, Geref 100). Blood samples were obtained for GH determination at -15, 0, 15, 30, 45, 60, 90, and 120 min. Blood samples for IGF-I, testosterone (T), androstenedione (A), dehydroepiandrosterone-sulfate (DHEA-S), sex hormone binding globulin (SHBG), glucose and insulin determinations were taken before GHRH injection. In all subjects, the oral glucose tolerance test (OGTT) was done on the second day after the GHRH test, as well. Blood samples were collected before and 30, 60, 120, 150, and 180 min after a standard glucose load (75 d. Assays Serum glucose concentrations were determined by the glucose oxidase method (Abbott, Spectrum). Commercial radioimmunological kits were used for plasma measurements of insulin, GH (Institute of Atomic Energy, Poland), and androstenedione (Wien Laboratories). Plasma levels of 17-hydroxyprogesterone were measured by RIA using Medgenix kits (Belgium). Neither extraction nor chromatography were required because of high specificity of the coated antibodies. The plasma concentrations of DHEA-S were estimated by RIA using Immunotech S.A. kits (France) with monoclonal antibodycoated tubes. Without previous extraction or chromatography plasma concentrations of testosterone were determined by RIA with Farmos Diagnostica kits (Finland). The antibodies used were specific, the cross-reactions with 5-cw-dihydrotestosterone and with 5-fl-dihydrotestosterone were 17.7% and 10.2%, respectively. Plasma IGF-I and SHBG concentrations were measured by an immunoradiometric method using kits obtained from BykSangtec (Germany) and Farmos Diagnostica (Finland), respectively. The IGF-I levels were estimated after preliminary acid extraction of plasma. The intraassay coefficients of variation were: insulin < 7.0%; GH < 4.2%; T < 6.6%; A < 10.9%; DHEA-S < 5.0%; IGF-I < 7.5% and SHBG < 4.9%. The interassay coefficients of variation were: insulin C 11.8%; GH < 10.8%; T < 6.4%; A C 12.5%; DHEA-S < 10.6%; IGF-I < 12% and SHBG < 10%. Statistical methods The integrated 2-h GH response to GHRH, and 3-h insulin response to glucose load were calculated as the areas under the curve by the trapezoid rule. The GH peak responses were
Results The subjects were of similar age (Table 1). The BMI values were higher in obese women than those in nonobese controls. No difference was observed in BMI between these two obese groups. In PC0 obese women the insulin response to glucose load and the plasma levels of testosterone and androstenedione were higher, but the plasma concentrations of SHBG were lower than those in obese women without PC0 and in nonobese controls. No differences in DHEA-S concentrations were found between groups. In PC0 obese women a positive correlation between fasting insulin and testosterone levels (r = 0.890, P < 0.001) and a negative correlation between fasting insulin and SHBG concentrations (r = -0.820, P < 0.01) were found. In all obese women a negative correlation between the integrated insulin response to glucose load and SHBG concentrations (r = 0.618, P < 0.02) was noted. The GH response to GHRH is shown in Fig. 1 and Table 2. No significant differences were found in the basal GH levels between groups. Total GH secretion, as expressed by the integrated GH response to GHRH, in PC0 women and in obese women without PC0 were lower than that in nonobese subjects. In all obese women, negative correlations were found between BMI and the peak GH responses (r = -0.639, P < 0.02) and between age and IGF-I levels (r = -0.520, p < 0.02). No significant correlations were found between IGF-I levels and BMI, insulin, T, A, and the GH responses to GHRH. Despite decreased GH secretion in subjects with obesity, plasma concentrations of IGF-I in obese PC0 women and in obese women without PC0 were similar to the IGF-I levels in nonobese controls. Discussion In our study obese women with PC0 showed a significantly lower GH response to GHRH than nonobese healthy controls. However, no statistically significant difference was observed in the GH response to GHRH between obese women with PC0 and those without PCO, both groups of obese subjects being comparable with respect to age and degree of obesity. Recently, Kazer et al. (7) determined GH concentrations over a period of 14 h by collecting blood samples
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
SLOWINSKA-SRZEDNICKA
1434 TABLE
JCE&M*1992 Vo174*No6
1. Characteristics and plasma hormone levels in obese women (with and without PCO) and in nonobese healthy controls Obese PC0 women n = 7
31.0 f 36.1 f BMI (kg/m*) 4.47 f Fasting glucose (mmol/L) 195.9 f Fasting insulin (pmol/L) Integrated insulin area 101856.3 f (pmol/L . min) 9.16 f Testosterone (nmol/L) 5.22 -t DHEA-S (hmol/L) SHBG (nmol/L) 15.4 f 8.94 f Androstenedione (nmol/L) Values are mean f SEM. a Determined by analysis of variance. *P < 0.001 us. nonobese healthy controls. c P < 0.01 us. obese women and VS. nonobese d P < 0.05 vs. obese women.
Age (yr)
3.2 1.5* 0.17 32.3’ 16093.5b~d
Obese women n = 7 28.2 34.4 4.22 179.4 70432.6
1.3’ 0.9 3.2b*d 0.95’
3.01 3.26 29.1 5.67
f f f f k
2.0 2.2” 0.16 32.3* 4079.7*
f 0.27 rf: 0.3 f 4.4b zk 0.73
Nonobese healthy controls n = 7
P value”
NS
28.0 21.0 3.91 70.7 28750.2
f + f f f
2.2 0.56 0.16 5.7 3171.4
0.0001 NS 0.01 0.002
3.15 3.47 63.0 4.73
f f r f
0.3 0.4 9.4 0.93
0.002 NS 0.0001 0.002
healthy controls.
GE h/l1
time rmini hc. 1. The GH response to 100 pg GHRH in PC0 obese women (x- -x), in obese women without PC0 (*- - -*), in nonobese healthy controls (+- - -+).
between 0900 and 2300 h, and showed that the GH secretion was decreased in PC0 women. In our study, we have demonstrated that the reduced GH secretion may be due to coexistent obesity present in over 50% of PC0 women (5), rather than to the PCO. In accordance with previous observations plasma levels of IGF-I in obese women with (7-10) and without PC0 (11, 12) were also not significantly different from the TABLE
ET AL.
values in nonobese controls. Moreover, no correlation was observed between IGF-I levels and GH response to GHRH. It is well known that the plasma concentration of IGFI depends on age, GH, nutrition, insulin, and sex hormones (13). At puberty, the IGF-I level increases steeply as a result of increased secretion of GH and sex hormones (14). With aging IGF-I concentration decreases with decreasing GH secretion (3). During childhood, puberty, and in adult life there is a strong relationship between plasma insulin and IGF-I (14). In addition, in vitro studies suggest that insulin enhances IGF-I production (15) by either direct regulation of the GH receptor (16) or by a permissive effect on post GH receptor events (17). We have demonstrated that insulin secretion in obese women after an oral glucose load is significantly higher than in nonobese women. Hyperinsulinism observed in obesity, particularly evident in women with ovarian androgenization in PCO, may influence IGF-I concentration. The biological action of IGF-I is thought to be partly determined by the IGF small binding protein (IGFBPI). It is also known that insulin affects the production and circulating levels of IGFBP-I (18) and SHBG (19). In this study, in agreement with previous observations
2. Changes in plasma GH and IGF-I levels after GHRH administration
in obese women (with and withou PCO) and in nonobese healthy
controls Obese PC0 women n = 7 Baseline GH (pg/L) GH peak post GHRH (fig/L) Integrated GH area (pg/L . min) IGF-I (air/L)
1.0 13.5 617.4 199.5
f f f +
0.4 3.9” 150* 39.1
Obese women n = 7 0.88 10.4 327.1 1924
f f + f
0.2 3* 161.4* 36.8
Nonobsse healthy controls n = 7 3.2 47.8 3181.4 224.3
f f & k
1.4 9.7 644.3 33.2
VdueS are means f SEM. 0 Determined by analysis of variance. * P < 0.001 us. healthy controls.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
P value” NS 0.001 0.0001 NS
GH/IGF-I
AXIS IN PC0 AND OBESITY
(12, 20), obese women showed lower plasma levels of SHBG than the nonobese controls, and, in particular, low concentrations of SHBG were demonstrated in obese women with PCO. It has been reported elsewhere that in extreme obesity, IGFBP-I and SHBG concentrations are decreased and the levels of these proteins are negatively correlated with insulin concentrations (12). In obese PC0 women, the levels of IGFBP-I have been found to decrease (9,21). On a low-calorie diet in women with PCO, increased concentrations of IGFBP-I were associated with concomitant lower insulin concentrations (8). It may be that decreased IGFBP-I levels in hyperinsulinemia in women with PC0 lead to an increase of the biological action of IGF-I which, in turn, inhibits GH secretion by the pituitary. In conclusion: in obese women with and without PCO, the GH response to GHRH is lower that in nonobese healthy controls. No difference, however, was observed in GH secretion in obese PC0 women as compared with obese women without PCO. Despite decreased GH secretion plasma levels of IGF-I were similar in obese and nonobese patients. No relationship was found between the GH response and the IGF-I levels. These findings suggest that the abnormality in the physiological negative feedback between GH and IGF-I in PC0 women is related to the coexistent obesity. Acknowledgment The authors are indebted to Oskar A. Chomicki M.Sc. for his valuable suggestions and improvement of the English text.
References 1. Frohman LA, Jansson JO. Growth hormone-releasing hormone. Endocr Rev. 1986$223-53. 2. Berelowitz M, Szabo M, Frohman LA, Firestone S, Chu L, Hintz RL. Somatomedin C mediates growth hormone negative feed-back bv effects on both the h.vpothalamus and the pituitary. Science. 1981;212:1279-81. -3. Hall K. Tallv M. The somatomedin-insulin-like aowth factors. J Internal Med. 1989;225:47-54. 4. Franks S. Polycystic ovary syndrome: a changing perspective. Clin Endocrinol (Oxf). 1989;31:87-120. 5. Nestler JE, Clore JN, Blackard WG. The central role of obesity (hyperinsulinemia) in the pathogenesis of polycystic ovary syn-
1435
drome. Am J Obstet Gynecol. 1989;161:1095-7. 6. Slowinska-Srzednicka J, Zgliczynski S, Wierzbicki M, et al. The role of hyperinsulinemia in the development of lipid disturbances in non obese and obese women with the poiycystic ovary syndrome. J Endocrinol Invest. 1991;14:569-72. 7. Kazer RR, Unterman TG, Glick RP. An abnormality of the growth hormone/insulin-like growth factor-I axis in women with polycystic ovary syndrome. J Clin Endocrinol Metab. 1990;71:958-62. 8. Kiddy DS, Hamilton-Fairley D, Seppala M, et al. Diet-induced changes in sex hormone binding globulin and free testosterone in women with normal or polycystic ovaries: correlation with serum insulin and insulin-like growth factor-I. Clin Endocrinol (Oxf’). 1989;31:757-63. 9. Conway GS, Jacobs HS, Holly JMP, Wass JAH. Effects of luteinizing hormone, insulin, insulin-like growth factor-I and insulinlike-growth factor small binding protein I in the polycystic ovary svndrome. Clin Endocrinol (Oxn. 1990:33:593-603. 10. Lanzone A, Fulghesu AM, Pappalardo’ S, et al. Growth hormone and somatomedin C secretion-in patients with polycystic ovarian disease. Their relationships with hyperinsulinism and hyperandrogenism. Gynecol Obstet Invest. 1990;29:149-53. 11. Kelliman M. Frohman LA. Enhanced erowth hormone (GH) responsiveness to GH-releasing hormone after dietary manipulation in obese and non obese subjects. J Clin Endocrinol Metab. 1988;66:489-94. 12. Weaver JU, Holly JM, Kopelman PG, et al. Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IGFBP-I) in extreme obesity. Clin Endocrinol (Oxf). 1990;33:415-22. 13. Froesch ER, Zapf J. Insulin-like growth factors and insulin: comparative aspects. Diabetologia. 1985;28:485-93. 14. Smith CP, Dunger D, Williams A, et al. Relationship between insulin, insulin-like growth factor I, and dehydroepiandrosterone sulfate concentrations during childhood, puberty, and adult life. J Clin Endocrinol Metab. 1989;68:932-37. 15. Daughaday WH, Phillips LS, Mueller MS. The effects of insulin and growth hormone on the release of somatomedin by the isolated rat liver. Endocrinology. 1976;98:1214-9. 16. Baxter RC, Bryson JM, Turtle JR. Somatogenic receptors of rat liver: regulation by insulin. Endocrinology. 1980;107:1i76-81. 17. Maes M. Underwood LE. Ketelslegers JM. Low serum somatomedin-C in insulindependent diabetes: evidence for a postreceptor mechanism. Endocrinology. 1986;118:377-82. 18. Suikkari AM, Koivisto A, Koistinen R, Seppala M, Yki-Jarvinen H. Dose-response characteristics for suppresion of low molecular weight plasma insulin-like growth factor-binding protein by insulin. J Clin Endocrinol Metab. 1989;68:135-140. 19. Plymate SR, Matej LA, Jones RE, Friedl KE. Inhibition of sex hormone-binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. J Clin Endocrinol Metab. 1988,67:460-4. 20. Pasquali R, Casimirri F, Plate L, Capelli M. Characterization of obese women with reduced sex hormone-binding globulin concentrations. Horm Metab Res. 1990;22:303-6. 21. Suikkari AM, Ruutiainen K, Erkkola R, Seppala M. Low levels of low molecular weight insulin-like growth factor binding protein in patients with polycystic ovary disease. Hum Reprod. 1989;4:136-9.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 22 November 2015. at 21:17 For personal use only. No other uses without permission. . All rights reserved.
