Normalization of hyperinsulinemia by chronic opioid receptor blockade in hyperandrogenemic women

Transcription

1 POLYCYSTIC OVARY SYNDROME Normalization of hyperinsulinemia by chronic opioid receptor blockade in hyperandrogenemic women Dijana Hadžiomerović, M.D., Bernhard Rabenbauer, M.D., and Ludwig Wildt, M.D. Department of Gynecologic Endocrinology and Reproductive Medicine, Medical University of Innsbruck, Innsbruck, Austria Objective: Evaluation of the effects of naltrexone on hyperinsulinemia and hyperandrogenemia in hyperandrogenemic, hyperinsulinemic women. Design: Controlled clinical study. Setting: Department of Gynecologic Endocrinology and Reproductive Medicine, Center of Obstetrics and Gynecology, Medical University of Innsbruck, Austria. Patient(s): Thirty-nine hyperandrogenemic, hyperinsulinemic women were studied. Intervention(s): Women were treated with naltrexone (50 mg/d) for 3 weeks. Main Outcome Measure(s): Body mass index (BMI), gonadotropin (LH, FSH) and androgen (T, free T, DHEAS) levels, and plasma levels of glucose, insulin, and C-peptide, during a standard 75-g oral glucose tolerance test (OGTT), were determined before and during chronic opiate receptor blockade. Result(s): The BMI did not change during therapy. When OGTT was repeated after treatment with naltrexone, glucose levels were not different from those before treatment. Insulin response, however, had dramatically declined. We also observed a significant decrease in the levels of serum androgens. Conclusion(s): Hyperinsulinemia associated with hyperandrogenemia can be improved or completely abolished by chronic opiate receptor blockade. This observation suggests that endogenous opiates play a critical role in the process leading to hyperinsulinemia in hyperandrogenemia. (Fertil Steril 2006;86: by American Society for Reproductive Medicine.) Key Words: Hyperinsulinemia, endogenous opiates, naltrexone, hyperandrogenemia, hirsutism Received September 24, 2005; revised and accepted January 30, Supported by the Medical University of Innsbruck Presented in part at the European Congress of Endocrinology, Göteborg, Sweden, September 3 7, Reprint requests: Dijana Hadžiomerović, M.D., Department of Gynecologic Endocrinology and Reproductive Medicine, Medical University Innsbruck, Anichstrasse 35, 6020 Innsbruck, Austria (FAX: ; dijanah@gmx.de). The presence of hyperinsulinemia in hyperandrogenemic, hirsute women has been recognized when elevated fasting levels of serum insulin as well as supraphysiologic increases in response to a glucose challenge were observed in a substantial proportion of such women (1, 2). The prevalence of this metabolic dysfunction was initially assumed to be low. More extensive studies, however, have demonstrated that hyperinsulinemia is a frequent finding in women with hirsutism, hyperandrogenemia, and polycystic ovary disease (PCOD) (3). The clinical significance of these findings, however, has still to be defined, and the pathophysiologic link between elevated androgen levels and hyperinsulinemia, as well as the etiology of hyperinsulinemia, remain obscure (2, 4, 5). Clinical observations and experimental studies suggest that endogenous opioid peptides are involved in the regulation of insulin secretion and carbohydrate metabolism in hyperandrogenemia (6 9). We have demonstrated that chronic opioid receptor blockade by naltrexone, a specific opiate antagonist that can be administered orally, decreases significantly the exaggerated insulin response to oral glucose challenge in hyperandrogenemic women (8, 10). Results from these two studies encouraged us to carry out further observations. To this end we enrolled prospectively a total of 39 patients in our present study. We tested the hypothesis that an increased activity of endogenous opioids plays a critical role in the pathophisiology of hyperinsulinemia and hyperandrogenemia. We examined the effect of chronic opioid receptor blockade by naltrexone on the insulin response to a standard oral glucose tolerance test (OGTT) and on androgen levels. We observed a significant decline in both the elevated insulin levels and in the serum concentration of androgens. This is a new observation that could result in new methods of treatment of hyperinsulinemia and hyperandrogenemia. MATERIALS AND METHODS Subjects Thirty-nine women presenting with hyperandrogenemia and hyprinsulinemia were studied after informed consent had /06/$32.00 Fertility and Sterility Vol. 86, No. 3, September 2006 doi: /j.fertnstert Copyright 2006 American Society for Reproductive Medicine, Published by Elsevier Inc. 651

2 been obtained. Hyperandrogenemia was diagnosed by elevated serum T ( 0.4 ng/ml) and free T ( 2.5 pg/ml) levels; the presence of clinical symptoms of hyperandrogenism, such as acne, hirsutism or hair loss; and irregular menses (corpus luteum [CL] insufficiency, anovulatory cycles, oligomenorrhea, or amenorrhea). Hyprinsulinemia was defined as an area under the insulin curve 12,000 IU 180 min/ml during an OGTT. Study Protocol The protocol was approved by the Institutional Review Board of the University Erlangen and Ethical Committee of the Medical University of Innsbruck. The changes in body mass index (BMI) and the serum levels of LH, FSH, T, free T, DHEAS, sex hormone-binding globulin (SHBG), and free androgen index of the enrolled patients were assessed before and after 3 weeks of naltrexone treatment. An OGTT was performed at the beginning of the study between 8:00 and 11:00 AM, after overnight fasting, by administration of 75 g of glucose in 400 ml of aqeous solution (Dextro-O.G-T., Boehringer Mannheim, Mannheim, Germany). Blood samples were drawn from an antecubital vein before administration of this solution and at 15-minute intervals thereafter for 3 hours. Naltrexone (Ferring Arzneimittel GmbH, Kiel, Germany), was administered orally at a dosage of 50 mg/d. Three to 9 weeks ( days, mean SEM) after initiation of naltrexone treatment, a second OGTT was performed. Assays All samples were analyzed on a modular analyzer (Roche Co., Berlin, Germany) with commercially available kits, for the concentrations of glucose (hexokinase glucose-6-phosphate dehydrogenase method; Hitachi-Modul P), insulin (electrochemiluminescence immunoassay; Elecsys-Modul E170, Roche Diagnostics, Indianapolis, IN); and C-peptide (electrochemiluminescence immunoassay; Elecsys-Modul E170). These assays had an interassay coefficient of variation (CV) of 5% and an intraassay CV of 2.3%. milliliter for C-peptide. For calculation of molar ratios, conversion factors of 331 and were used for C-peptide (ng/ml to pmol/l) and insulin ( IU/mL to pmol/l), respectively. Furthermore, the degree of peripheral insulin sensitivity was evaluated according to the algorithm of Cederholm and Wibell (11). Statistical Analysis Data were stored and analyzed using the Statistical Package for Social Science, release 11 (SPSS Inc., Chicago, IL). The values are expressed as means SD. The changes in hormone levels and BMI during naltrexone treatment and results of both OGTTs were compared using Student s t-test for paired samples, each patient serving as her own control. Values of P.05, two-sided, were considered statistically significant. RESULTS All 39 study patients were hypreandrogenemic with T levels 0.4 ng/ml and free T levels 2.5 pg/ml. They exhibited different clinical signs of hyperandrogenism. Hirsutism was observed in 29 patients (74%), acne and seborrhea in 21 (54%), hair loss in 4 (10%), and deepening of the voice in 1. The cycle irregularities were observed in 33 patients (85%); 14 patients (36%) had oligomenorrhea, 16 (41%) FIGURE 1 Serum concentration (mean SD) of LH/FSH, T, free T (ft), sex hormone-binding globulin (SHBG), and DHEAS (left ordinate) and free androgen index (FAI) (right ordinate) in 39 women with hyperandrogenemia and hyperinsulinemia at baseline and after the administration of naltrexone (50 mg/d) for 3 weeks. The women were studied before and during naltrexone treatment. *P.05; **P.01; ***P.001 (t-test for paired samples) for comparison with baseline values in the same group. The LH, FSH, SHBG, total and free T, and DHEAS were determined with immunoassay kits obtained from DRG (Marburg/Lahn, Germany) and Biermann GmbH (Bad Nauheim, Germany). Data Analysis The results of the OGTTs were analyzed with regard to hyperinsulinemia by calculating the insulin-to-glucose ratio, the ratio of the molar concentrations of C-peptide and insulin, and by determining the areas under the curve of glucose, insulin, and C-peptide concentrations, as well as the insulinto-glucose and C-peptide-to-insulin ratios during OGTT. Values were expressed as milligrams per 100 milliliters for glucose, microunits per milliliter for insulin, and nanograms per 652 Hadžiomerović et al. Opioid blockade and hyperandrogenemia Vol. 86, No. 3, September 2006

3 secondary amenorrhea, 3 (8%) primary amenorrhea, and 6 (15%) eumenorrhea. However, of the 6 patients 2 (5%) were not ovulating and 4 (10%) had CL insufficiency. Mean levels of LH, FSH, T, free T, SHBG, free androgen index, and DHEAS before and during naltrexone treatment are shown in Figure 1. Despite significant changes toward the normalization of hormone levels, the values still differed from those of women without hyperandrogenemia. No significant changes were observed in clinical signs of hyperandrogenemia, although there was a tendency toward decreases in acne, seborrhea, and cycle irregularities in patients treated for 6 9 weeks. The mean BMI for all 39 patients before therapy was kg/m 2 and did not change during treatment ( kg/m 2 ) as shown in Figure 2. Waist-to-hip ratio was also evaluated and did not change significantly either ( before to during naltrexone treatment). Thirty-three (85%) of the patients were obese (mean BMI before treatment kg/m 2 and kg/m 2 after treatment); the remaining 6 patients were of normal weight (mean BMI before treatment kg/m 2 and kg/m 2 after treatment). Figure 3 shows the area under the insulin curve for each patient before and after 3 9 weeks of naltrexone treatment. FIGURE 2 Body mass index in 39 hyperandrogenemic, hyperinsulinemic patients with three different levels of severity of hyperinsulinemia (A slight; B moderate; C severe hyperinsulinemia) before and during naltrexone treatment. Data are expressed as the mean SD. None of the three groups shows a significant change in body mass index. The levels of insulin decreased in 29 (74%) of 39 patients. In five of the patients almost no change was noted, and five patients developed increased insulin levels during treatment. Figure 4 shows that when OGTT was repeated during naltrexone administration, the increase in glucose concentrations was no different from that observed before treatment. Serum levels of insulin, however, were strikingly different from those before treatment and stimulated levels were significantly (P.01, t-test for paired samples) lower when compared to the pretreatment values in all patients. The higher the insulin levels were at the beginning of the study, the greater the decrease during the naltrexone treatment. The increase in C-peptide concentrations under OGTT was also diminished by naltrexone treatment, but to a much lesser degree than that of insulin. As a consequence of these changes, the insulin-to-glucose ratio declined, whereas the C-peptide-to-insulin ratio increased during treatment. The mean values of peripheral insulin sensitivity index (ISI) improved significantly (P.05) from (mean SEM) to for the entire group during treatment. The ISI increased significantly (P.05) in 27 of 39 women with moderate-to-severe insulin resistance (ISI 35), to In patients with only mild abnormalities (ISI 35) no consistent changes were observed ( to ). Naltrexone administration was associated with few side effects, which included nausea, mild headaches, restlessness, and sleep disturbances in a small proportion of women after the initiation of treatment. These symptoms, however, disappeared within a few days. DISCUSSION Endogenous opioid peptides play a well-defined role as inhibitory modulators of hypothalamic GnRH secretion in hypothalamic ovarian failure (10, 12 14). Their role in the etiology of hyperandrogenemia has received relatively little attention and has been defined less clearly. Some previous studies have demonstrated that endogenous opioids are involved in the modulation of carbohydrate metabolism and could be responsible for hyperinsulinemia in women of reproductive age. Several studies have shown that inhibition of the opioid tone resulted in a reduction of hyperinsulinemia in hyperandrogenemic, hyperinsulinemic women (15 18). The data obtained in this study unequivocally demonstrated a striking decrease of stimulated insulin levels in response to carbohydrate challenge during a chronic opiate receptor blockade by naltrexone. Therefore, these results strongly suggest that an elevated opioid activity plays a key role in the pathogenesis of hyperinsulinemia in hyperandrogenemia. The described effect was more pronounced in the presence of severe hyperinsulinemia. Calculation of the ISI confirmed these results. Increasing ISI was mostly found in patients with moderate and severe hyperinsulinemia. Fertility and Sterility 653

4 FIGURE 3 Area under the curve (AUC) of insulin concentration before and after 3 9 weeks of naltrexone treatment for each patient. The levels of insulin were significantly lower (P.01, t-test for paired samples) when compared to baseline values in all 39 hyperinsulinemic, hyperandrogenemic women. This view of a critical role of endogenous opioids in the pathophysiology of hyperinsulinemia associated with hyperandrogenemia is supported by the following observations. Immune reactive -endorphin has been demonstrated within the pancreas (19, 20) and elevated levels of this endogenous opioid peptide have been found in the peripheral circulation of hyperandrogenemic women with polycystic ovaries (PCO) (9). Moreover, an increase of circulating insulin concentrations could be induced by the administration of endorphin to study subjects (19 21). The changes in insulin concentrations in response to glucose challenge observed in this study may be attributed to a decline in insulin secretion, an increase in insulin clearance, an increase in target tissue sensitivity to the action of insulin, or a combination of all these factors. The observation that no appreciable changes in the pattern of glucose concentrations could be demonstrated suggests an increase in target organ sensitivity to insulin or a change in insulin metabolism as the more likely mechanisms, rather than changes in pancreatic insulin secretion. The changes in the pattern of C-peptide concentrations observed as a consequence of naltrexone administration seem to be of particular importance in this respect. C-peptide is secreted with insulin in equimolar amounts. While insulin clearance is, to a major part, determined by hepatic uptake, the clearance of C-peptide from the circulation is mainly dependent upon renal excretion. Therefore C-peptide concentrations reflect insulin secretion more closely than insulin concentration itself (22, 23). The absence of dramatic changes in circulating C-peptide concentrations during OGTT, before and during naltrexone administration, suggests an increase in insulin clearance, rather than a reduction of insulin secretion, as the major determinant of the reduced concentrations of insulin found in the peripheral circulation. The changes in C-peptide-toinsulin ratio observed as a consequence of the opioid antagonist administration lend additional support to this view. This does not exclude, however, the presence of additional mechanisms that affect insulin secretion or target tissue sensitivity to insulin. In any event, glucose tolerance is improved as a consequence of these changes. The results of the study support the previous observations by Vettor and colleagues (15) and Givens and co-workers (16), who reported a decrease in insulin response to glucose challenge during acute opiate receptor blockade (by the i.v. administration of naltrexone), by demonstrating a sustained reduction in stimulated insulin serum concentrations during chronic administration of an opiate antagonist. Our observations are in line with those of Cucinelli et al. (18), Fruzzetti et al. (17), and Fulghesu et al. (24) who reported similiar results after chronic naltrexone administration. In the present study we also observed significant changes in levels of gonadotropins and androgens. The largest improvement was shown in levels of free androgen index, free T, and SHBG, although none of the analyzed hormone levels achieved the values for normoandrogenemia. This may be explained by the limited treatment period as well as by the 654 Hadžiomerović et al. Opioid blockade and hyperandrogenemia Vol. 86, No. 3, September 2006

5 FIGURE 4 Area under the curve (AUC) of glucose, insulin, and C-peptide concentrations after 75-g glucose load in hyperandrogenemic women before ( ) and during (Œ) the naltrexone treatment. Data are expressed as mean SD. When OGTT was repeated during naltrexone administration, the increase in glucose concentrations was not different from that observed before treatment. The levels of insulin were strikingly lower (P.01, t-test for paired samples) when compared to baseline values. The increase in C-peptide concentrations was influenced to a much lesser extent than that of insulin. presence of other mechanisms that maintain hyperandrogenemia in addition to insulin resistance. Our results regarding the decrease of androgen levels are, once again, in line with those of Fruzzetti et al. (17), although they did not observe any improvement SHBG levels. In their study Fruzzetti et al. (17) analyzed only obese patients and could detect a significant decrease in BMI during treatment. They, therefore, discussed the possibility that the observed results could either be a consequence of weight loss or due to changes in opioidergic tone. Although we analyzed the levels of BMI before and during the study, above average BMI was not one of our inclusion criteria, as previous studies have shown that lean subjects with a given phenotypic expression of hyperandrogenemia have an equivalent degree of hyperinsulinemia compared to obese subjects (25, 26). We observed an decrease in hyperinsulinemia and hyperandrogenemia, independent of changes of BMI, after a very short observation time, which is a new aspect in this regard. No significant changes in BMI occurred during treatment in our study. Because of the short observation time we cannot come to any definite conclusions regarding changes in clinical signs and endocrine disturbances of hyperandrogenism with naltrexone treatment. In patients treated for 6 9 weeks there was a tendency toward decreases in acne, seborrhea, and cycle irregularities. Further studies would be necessary during a longer treatment period to demonstrate whether reductions in these conditions are a consequence of decreases in hyperandrogenemia and hyperinsulinemia. Naltrexone could open a novel nonhormonal, oral mode in treatment, not only of metabolic, but also of skin and cycle irregularities in patients with hyperinsulinemia and hyperandrogenemia, associated with a low level of side effects. The etiology of hyperinsulinemia in hyperandrogenemic women is still not completely clear. Although obesity appears to be an additive to hyperinsulinemia, the previous observations indicate that hyperinsulinemia is a common finding in hyperandrogenemic women, independent of obesity (27). Because elevated concentrations of -endorphin have been found in the peripheral circulation in obese women (9), and the association between obesity and insulin resistance is well documented, one might argue that the weight loss that sometimes, but not consistently, occurs during naltrexone Fertility and Sterility 655

6 treatment is a major factor in the dramatic improvement of carbohydrate metabolism. However, we did not observe any significant changes in BMI in our study, whereas a significant reduction in hyperinsulinemia was achieved, which supports the hypothesis that the naltrexone mechanism of action in hyperinsulinemc, hyperandrogenemic patients is not due to weight loss. The observation made in other studies that obesity is present in only 33% 60% (26) of hyperandrogenemic, hyperinsulinemic women, as well as the demonstration that weight loss alone does not seem to influence the elevated serum levels of -endorphin, argues against obesity as the major determining factor in the development of the metabolic abnormalities and their apparent normalization during treatment with naltrexone (19, 28 30). We should mention that of the six patients with BMI 25 kg/m 2, a decrease in area under the curve of insulin, although not statistically significant, was noted, which showed that the inclusion of lean patients did not influence the study results. Therefore, we conclude that treatment with naltrexone, whether in lean or obese patients, has the same result, as the androgen and gonadotropin levels decreased in both groups, although the lean patient group was small. A control group was not included in the study, hence each patient served as her own control. It is still unclear whether hyperinsulinemia leads to hyperandrogenemia or hyperandrogenemia to hyperinsulinemia. Under experimental conditions, androgens administered in high doses may induce resistance to the action of insulin on the target tissues (31). However, circulating androgen concentrations in most hyperandrogenemic women remain far below the levels observed in healthy men. Therefore, the increase of androgens alone does not seem to be sufficient to account for the considerable extent of hyperinsulinemia observed in these patients (31, 32). This view is further supported by the findings that administration of T did not alter insulin sensitivity in the female rhesus monkey (33) and by the demonstration that long-term suppression of androgen serum levels by the administration of E and progestins, or treatment with antiandrogens, does not seem to influence the extent of insulin resistance (32, 34, 35). Based on the findings made in this study, the observations discussed, and the demonstration that constant elevated concentrations of insulin will lead to a down-regulation of hepatic insulin receptors, we propose the following hypothesis to explain the finding of hyperinsulinemia in hyperandrogenemic patients. An increase in endogenous opioid activity initiates, by acting on the pancreatic cells, an increase in insulin secretion, resulting in elevated insulin levels in the hepatic circulation. Elevated insulin concentrations lead to a down-regulation of hepatic insulin receptors and thereby induce and maintain hyperinsulinemia (36). Hyperandrogenemia may develop as a consequence of the stimulatory action of insulin at the ovarian level (5, 36). An increase of endogenous opioid activity thus represents the first step in the chain of events leading to hyperinsulinemia and hyperandrogenemia. Hyperinsulinemia accompanied by hyperandrogenemia and hirsutism was first identified in women of reproductive age. In addition, hyperinsulinemia has been found to be associated with a number of metabolic and cardiovascular disorders, which represent major factors of morbidity and mortality in older women. These include type 2 diabetes, hypertension, and arteriosclerosis, as well as endometrial hyperplasia and carcinoma (37 43). The findings in the present study, as well as those in the study by Cucinelli et al. (18), may have implications beyond the pathogenesis and etiology of hyperandrogenemia and ovarian dysfunction. Amelioration of hyperinsulinemia by opioid receptor blockade may open new avenues in the treatment of disorders affecting a large proportion of the female population, not only during the reproductive years, but even more so in the subsequent years of life. Acknowledgments: The authors thank Wolfgang Prokop, M.D., for skillful performance of the assays and the hospital staff for their assistance in the clinical part of the study. The authors also thank Ferring Arzneimittel GmbH (Kiel, Germany) for kindly providing the naltrexone. REFERENCES 1. Kahn CR, Flier JS, Bar RS, Archer JA, Gorden P, Martin MM, et al. Syndromes of insulin resistance and acanthosis nigricans insulinreceptor disorders in man. N Engl J Med 1976;294: Barbieri RL, Ryan KJ. Hyperandrogenism, insulin resistance, and acanthosis nigricans syndrome a common endocrinopathy with distinct pathophysiologic features. Am J Obstet Gynecol 1983;147: Barbieri RL. Polycystic ovarian disease. Annu Rev Med 1991;42: Poretsky L. On the paradox of insulin-induced hyperandrogenism in insulin-resistant states. Endocrinol Rev 1991;12: Poretsky L, Kalin MF. The gonadotropic function of insulin. Endocrinol Rev 1987;82: Lanzone A, Fulghesu AM, Fortini A, Cutillo G, Cucinelli F, Di Simone N, et al. Effect of opiate receptor blockade on the insulin-response to oral glucose load in polycystic ovarian disease. Hum Reprod 1991;6: Sir-Petermann T, Alba F, Castillo T, Munoz A, Cortinez A, Maliqueo M, et al. Relationship between insulin resistance, LH secretion and ovarian morphological changes, in women with polycystic ovary syndrome. Rev Med Chil 1998;126: Sir-Petermann T, Lopez G, Castillo T, Calvillan M, Rabenbauer B, Wildt L. Naltrexone effects on sensitivity and insulin secretion in hyperandrogenic women. Exp Clin Endocrinol Diabetes 1998;106: Givens JR, Wiedemann E, Andersen RN, Kitabchi AE. Beta-endorphin and beta-lipotropin plasma-levels in hirsute women correlation with body weight. J Clin Endocrinol Metab 1980;50: Wildt L, Sir-Petermann T, Leyendecker G, Waibel-Treber S, Rabenbauer B. Opiate antagonist treatment of ovarian failure. Hum Reprod 1993;8: Cederholm J, Wibell L. Insulin release and peripheral sensitivity at the oral glucose tolerance test. Diabetes Res Clin Pract 1990;10: Wildt L, Leyendecker G, Sir-Petermann T, Waibel-Treber S. Treatment with naltrexone in hypothalamic ovarian failure induction of ovulation and pregnancy. Hum Reprod 1993;8: Ferin M. Endogenous opioid peptides and the menstrual cycle. Trends Neurosci 1984;7: Hadžiomerović et al. Opioid blockade and hyperandrogenemia Vol. 86, No. 3, September 2006

7 14. Ferin M, Van Vugt D, Wardlaw S. The hypothalamic control of the menstrual cycle and the role of endogenous opioid peptides. Recent Prog Horm Res 1984;40: Vettor R, Martini C, Manno M, Cestaro S, Federspil G, Sicolo N. Effects of naloxone-induced opiate receptors blockade on insulin secretion in obesity. Horm Metab Res 1985;17: Givens JR, Kurtz BR, Kitabchi AE, Bittle JB, Karas JG, Mitchell JA, et al. Reduction of hyperinsulinemia and insulin resistance by opiate receptor blockade in the polycystic ovary syndrome with acanthosis nigricans. J Clin Endocrinol Metab 1987;64: Fruzzetti F, Bersi C, Parrini D, Ricci C, Genazzani AR. Effect of long-term naltrexone treatment on endocrine profile, clinical features, and insulin sensitivity in obese women with polycystic ovary syndrome. Fertil Steril 2002;77: Cucinelli F, Soranna L, Perri C, Barini A, Cento RM, Mancuso S, et al. Use of naltrexone in postmenopausal women with exaggerated insulin secretion: a pilot study. Fertil Steril 2004;81: Bruni JF, Watkins WB, Yen SS. Beta-endorphin in the human pancreas. J Clin Endocrinol Metab 1979;49: Feldman M. Beta-endorphin and the endocrine pancreas reply. N Engl J Med 1983;309: Giugliano D, Cozzolino D, Salvatore T, Ceriello A, Torella R. Dual effect of beta-endorphin on insulin secretion in man. Horm Metab Res 1987;19: Watanabe RM, Volund A, Roy S, Bergman RN. Prehepatic beta-cell secretion during the intravenous glucose tolerance test in humans application of a combined model of insulin and C-peptide kinetics. J Clin Endocrinol Metab 1989;69: Bergman RN, Finegood DT, Ader M. Assessment of insulin sensitivity in vivo. Endocrinol Rev 1985;6: Fulghesu AM, Lanzone A, Cucinelli F, Caruso A, Mancuso S. Longterm naltrexone treatment reduces the exaggerated insulin secretion in patients with polycystic ovary disease. Obstet Gynecol 1993;82: Dunaif A, Finegood DT. Beta-cell dysfunction independent of obesity and glucose intolerance in the polycystic ovary syndrome. J Clin Endocrinol Metab 1996;81: Marsden PJ, Murdoch AP and Taylor R. Tissue insulin sensitivity and body weight in polycystic ovary syndrome. Clin Endocrinol 2001;55: Dunaif A, Segal KR, Futterweit W, Dobrjansky A. Profound peripheral insulin resistance, independent of obesity, in polycystic ovary syndrome. Diabetes 1989;38: Ciampelli M, Fulghesu AM, Cucinelli F, Pavone V, Ronsisvalle E, Guido M, et al. Impact of insulin and body mass index on metabolic and endocrine variables in polycystic ovary syndrome. Metabolism 1999; 48: Guido M, Pavone V, Ciampelli M, Murgia F, Fulghesu AM, Apa R, et al. Involvement of ovarian steroids in the opioid-mediated reduction of insulin secretion in hyperinsulinemic patients with polycystic ovary syndrome. J Clin Endocrinol Metab 1998;83: Holte J, Bergh T, Berne C, Wide L, Lithell H. Restored insulin sensitivity but persistently increased early insulin secretion after weight loss in obese women with polycystic ovary syndrome. J Clin Endocrinol Metab 1995;80: Cohen JC, Hickman R. Insulin resistance and diminished glucose tolerance in powerlifters ingesting anabolic steroids. J Clin Endocrinol Metab 1987;64: Dunaif A, Green G, Futterweit W, Dobrjansky A. Suppression of hyperandrogenism does not improve peripheral or hepatic insulin resistance in the polycystic ovary syndrome. J Clin Endocrinol Metab 1990;70: Billiar RB, Richardson D, Schwartz R, Posner B, Little B. Effect of chronically elevated androgen or estrogen on the glucose tolerance test and insulin response in female rhesus-monkeys. Am J Obstet Gynecol 1987;157: Pasquali R, Fabbri R, Venturoli S, Paradisi R, Antenucci D, Melchionda N. Effect of weight loss and antiandrogenic therapy on sex hormone blood levels and insulin resistance in obese patients with polycystic ovaries. Am J Obstet Gynecol 1986;154: Singer F, Bhargava G, Poretsky L. Persistent insulin resistance after normalization of androgen levels in a woman with congenital adrenal hyperplasia a case report. J Reprod Med 1989;34: Mahabeer S, Jialal I, Norman RJ, Naidoo C, Reddi K, Joubert SM. Insulin and C-peptide secretion in non-obese patients with polycystic ovarian disease. Horm Metab Res 1989;21: Stout RW. Insulin and atheroma 20-yr perspective. Diabetes Care 1990;13: Wild RA, Applebaum-Bowden D, Demers LM, Bartholomew M, Landis JR, Hazzard WR, et al. Lipoprotein lipids in women with androgen excess independent associations with increased insulin and androgen. Clin Chem 1990;36: Papa V, Pezzino V, Costantino A, Belfiore A, Giuffrida D, Frittitta L, et al. Elevated insulin-receptor content in human breast cancer. J Clin Invest 1990;86: Kannel WB, Wilson PW, Zhang TJ. The epidemiology of impaired glucose tolerance and hypertension. Am Heart J 1991;121: Legro RS, Blanche P, Krauss RM, Lobo RA. Alterations in low-density lipoprotein and high-density lipoprotein subclasses among Hispanic women with polycystic ovary syndrome: influence of insulin and genetic factors. Fertil Steril 1999;72: Dunaif A, Legro RS. Prevalence and predictors of risk for type 2 diabetes mellitus and impaired glucose tolerance in polycystic ovary syndrome authors response. J Clin Endocrinol Metab 1999;84: Legro RS, Kunselman AR, Dodson WC, Dunaif A. Prevalence and predictors of risk for type 2 diabetes mellitus and impaired glucose tolerance in polycystic ovary syndrome: a prospective, controlled study in 254 affected women. J Clin Endocrinol Metab 1999;84: Fertility and Sterility 657