Disturbances in insulin secretion and sensitivity in women with the polycystic ovary syndrome

3 Disturbances in insulin secretion and sensitivity in women with the polycystic ovary syndrome JAN HOLTE INSULIN RESISTANCE AND THE INSULIN RESISTANCE SYNDROME

Although insulin exhibits a wide spectrum of biological effects, the terms insulin sensitivity and insulin resistance generally refer to the actions of insulin on glucose homeostasis. Insulin stimulates the uptake, storage and use of glucose in muscle and fat tissue, and promotes glycogen storage and inhibits gluconeogenesis and glycogenolysis in the liver. Insulin resistance is defined as a diminution in these biological responses to a given level of insulin (Moller and Flier, 1991). The clinical picture in states of insulin resistance is often complex because of effects attributed either to deficient insulin action (glucose intolerance, non-insulin dependent diabetes mellitus: (NIDDM) or to exaggerated insulin action because of the compensatory hyperinsulinaemia. Thus the notion is that, in states of insulin resistance, insulin may still exhibit preserved effects in actions other than those involved in glucose regulation (Amiel et al, 1991). Of great importance in the present context are the gonadotrophic effects of insulin, which are presumed to play a role in the association between hyperinsulinaemia and polycystic ovary syndrome (PCOS) (Barbieri and Ryan, 1983; Buyalos et al, 1992). The mechanisms for some specific causes of insulin resistance have been clarified, as antibodies directed to the insulin receptor (Kahn and Flier, 1976) or mutations of the insulin receptor gene (Moller and Flier, 1988), but insulin resistance is most commonly associated with physiological states or circulating factors for which the precise mechanisms of action are less clearly identified. Obesity, puberty, pregnancy and ageing are examples of complex physiological states with a decreased sensitivity to insulin. Glucocorticoids, catecholamines, glucagon, growth hormone (GH) and free fatty acids in excess can induce insulin resistance. In addition, increased plasma levels of either insulin or glucose may impair insulin sensitivity (Moiler and Flier, 1991). The insulin resistance syndrome (syndrome X), comprising dyslipidaemia, hypertension, central obesity and insulin resistance, constitutes closely interrelated risk factors for cardiovascular disease (Reaven, 1988). Insulin resistance may be the key factor in this syndrome. Thus, deficient insulin action Bailli~re's Clinical Endocrinology and Metabolism-221 Vol. 10, No. 2, Aplil 1996 ISBN 0-7020-2100-8 0950-35 IX/96/020221+ 27 $12.00/00

Copyright © 1996, by Bailli6re Tindall All rights of reproduction in any form reserved

222

J. HOLTE

may cause both increased triglycerides and decreased high-density lipoprotein (HDL)-cholesterol levels in serum (Frayn, 1993). The association with hypertension may also be of a direct cause-and-effect kind (Ferrannini et al, 1987), although this is an issue under current debate. Truncal-abdominal obesity is associated with insulin resistance (Evans et al, 1984). Again, the mechanisms behind this association are incompletely known. A high flux of free fatty acids from the increased bulk of abdominal fat tissue may be an important link between these factors (Ferrannini et al, 1983; Reaven and Chen, 1988). Several putative factors behind this specific accumulation of fat are presently being investigated, interactions between cortisol and sex steroids presumably being of great importance (M~rin et al, 1992; Pasquali et al, 1993). EVALUATING IN VIVO INSULIN SENSITIVITY AND SECRETION Standard clinical procedures as the insulin response to oral (oral glucose tolerance test: OGTT) or intravenous glucose (intravenous glucose tolerance test: IVGTT) give a picture of glucose homeostasis that is dependent on insulin secretion, metabolism and hepatic and peripheral insulin sensitivity. Although efficient for crude evaluation of the net effects of insulin secretion and sensitivity, such methods can not differentiate these two basic components determining insulin action. The 'gold standard' for evaluation of insulin sensitivity is the hyperinsulinaemic euglycaemic clamp (De Fronzo et al, 1979), in which insulin is administered at doses high enough to completely inhibit endogenous insulin secretion and hepatic glucose production. Blood glucose levels are maintained at normal values by the simultaneous administration of intravenous glucose, and the amount of glucose given, reflecting the glucose disposal, is a measure of the subject's peripheral insulin sensitivity ('M', often adjusted for the prevailing insulin levels during the clamp; insulin sensitivity index, 'M/I'). Other established methods for measuring insulin sensitivity independent of beta cell function are the insulin tolerance test (ITT), in which the early decline in blood glucose levels after intravenous insulin administration reflects insulin sensitivity (Bonora et al, 1989), and the 'minimal model'-modified IVGTT, in which both insulin secretion and sensitivity can be calculated by mathematical modelling (Bergman et al, 1989). Insulin secretion is generally estimated from the early insulin increment in an IVGTT, reflecting the first-phase release of insulin. With the use of more sophisticated methods, based on measurements of C-peptide turnover, the insulin secretion rate and insulin degradation can be estimated (Eaton et al, 1980). G L U C O S E TOLERANCE, INSULIN SECRETION AND INSULIN SENSITIVITY NIDDM is a clinically heterogeneous condition with respect to factors such as age of onset and mode of inheritance. Except for a minority of patients

INSULIN RESPONSES IN PCOS

223

who are of normal weight, NIDDM is generally associated with insulin resistance and obesity, and to a higher degree than accounted for by body weight p e r se (Moller and Flier, 1991). A component of insulin resistance seems to be inherited in offspring of parents with NIDDM (Warram et al, 1990). Insulin resistance antedates the onset of NIDDM, which is most likely determined by the failure of compensatory beta cell secretion in susceptible subjects (Moller and Flier, 1991). In the stage preceding NIDDM--impaired glucose tolerance--normal fasting blood glucose levels are preserved, whereas the glucose disposal after an oral or intravenous glucose load is reduced. Thus in impaired glucose tolerance and NIDDM, decreased beta cell function, either in absolute terms or in relation to insulin resistance, is invariably found (Moller and Flier, 1991). The incipiently increased blood glucose levels may further impair insulin sensitivity. This latter mechanism underlines the importance of controlling for glucose intolerance when comparing tissue insulin sensitivity between different subjects. INSULIN RESISTANCE AND PCOS

An association between diabetes and hirsutism in women was reported in 1921 (Achard and Thiers, 1921). Later, a clinical connection between hyperandrogenaemia, insulin resistance and acanthosis nigricans was described, constituting the HAIR-AN syndrome. These first-reported women were all severely insulin resistant, owing to either insulin receptor mutations or other target cell defects in insulin action ('Type A syndrome'), or autoantibodies to the insulin receptor ('Type B syndrome'). They usually exhibited hyperthecotic ovaries (Kahn and Flier, 1976). These observations were followed by reports of insulin resistance in women with a more classical PCOS (Burghen et al, 1980; Pasquali et al, 1983; Shoupe et al, 1983). Although obesity was a confounding factor in the early reports, later studies also found impaired insulin sensitivity in a proportion of non-obese women with PCOS (Chang et al, 1983; Dunaif et al, 1989). In addition, subclinical acanthosis nigricans was found to be highly prevalent in skin biopsies from insulin-resistant subjects (Dunaif et al, 1991), acanthosis nigricans developing in parallel with increased insulin levels and being reversible with decreasing insulin levels (Conway and Jacobs, 1990). Together, these data revealed a wide clinical spectrum rather than distinct separate entities. At one extreme of this spectrum were the obese, severely insulin-resistant women with HAIR-AN syndrome, and at the other extreme lean women with few clinical and biochemical signs of PCOS in combination with slightly impaired insulin sensitivity. INSULIN-ANDROGENS-OBESITY: A THREE-DIMENSIONAL PROBLEM

Correlations between circulating levels of androgens and insulin or (inversely) insulin resistance have been found in some, but not all, studies

224

J. HOLTE

(Holte et al, 1991). The most consistent associations between androgens and insulin are found for levels of unbound or free testosterone, calculated from suppressed concentrations of sex hormone binding globulin (SHBG) (Kiddy et al, 1989; Buyalos et al, 1993; Holte et al, 1994a). A third major factor, obesity, and specifically truncal-abdominal obesity, is associated both with insulin resistance and decreased SHBG (both in normal controls and in PCOS) and increased testosterone (PCOS), underlying the complexity of the problem of insulin resistance in PCOS (Holte et al, 1994a; Pasquali et al, 1994). ARGUMENTS FOR A CAUSE-AND-EFFECT RELATIONSHIP BETWEEN INSULIN AND ANDROGENS Androgens

Receptors for insulin and insulin-like growth factor I (IGF-I) are normally abundant in the different compartments of the human ovary (Poretsky and Kalin, 1987; el Roeiy et al, 1994), providing the biological basis for the hypothesis that hypefinsulinaelnia contributes to raised androgen concentrations in PCOS through ovarian stimulation. Thus, insulin could exert gonadotrophic effects either directly through the insulin receptor or through 'spill-over' on IGF-I-receptors, in spite of resistance to the action of insulin for glucose uptake. Recent data also suggest that the biologically available circulating IGF-I itself may be enhanced through reduced concentrations of IGF-I-binding protein I (IGFBP-I) in mainly obese women with PCOS, at normal or increased levels of IGF-I (Suikkari et al, 1989; Laatikainen et al, 1990; Homburg et al, 1992). Severely reduced IGFBP-! levels were found in clomiphene-resistant, lean women with PCOS (Tiitinen et al, 1993). Furthermore, high (Lee et al, 1993) or normal (Kazer et al, 1990) IGF-I levels may be associated with disturbances in GH secretion in PCOS. Experiments on cultured rat ovarian tissue show potentiating effects of IGF-I or insulin on luteinizing hormone (LH)-stimulated side chain cleavage enzyme (Magoffin et al, 1990), 17~ hydroxylase (Magoffin and Weitsman, 1993a) and 3-[3 hydroxysteroid dehydrogenase (Magoffin and Weitsman, 1993b). In vivo studies show a high prevalence of exaggerated activity of cytochrome P450c17 in women with PCOS (Rosenfield et al, 1990). A role for increased insulin inducing this enhanced enzyme activity is supported by an intervention study in which reduction of insulin levels resulted in apparently normalized enzyme activity (Lanzone et al, 1994). Previous in vitro studies suggested that the effects of insulin and IGF-I on androgen production were greater in polycystic than in normal ovaries. Thus insulin alone greatly stimulated the production of androstenedione, testosterone and dihydrotestosterone in minced stroma from hyperandrogenic, hyperinsulinaemic women (Barbieri et al, 1986), whereas androgen production in normal ovaries was seen either only (Barbieri et al, 1986) or mainly in synergy with LH (Bergh et al, 1993; Nahum et al, 1995). Alternatively, cultured theca cells from polycystic ovaries already showed

I N S U L I N R E S P O N S E S 1N PCOS

225

an enhanced steroidogenic activity in the basal state, with no apparent differences in response to LH, insulin, IGF-I or IGF-II compared with theca cells from normal ovaries (Gilling-Smith et al, 1994). However, apart from stimulating androgen production in theca cells, insulin and IGF-I may also stimulate aromatase (Garzo and Dorrington, 1984) and overall oestradiol production in synergy with FSH, with no apparent differences between granulosa cells from polycystic and normal ovaries (Mason et al, 1994). The reasons for the net effect of a preferential increase in androgens in polycystic ovaries are not fully clarified. The human ovary contains a complex system of IGFs (IGF-I and II), IGF-I being regulated by five out of six known IGF-I-binding proteins within the ovary. Follicles fi'om polycystic ovaries have shown an increased amount of IGFBP but a normal amount of IGF-I, suggesting reduced bioavailability of IGFs for granulosa, thus preventing aromatization (Giudice, 1995). Results from other experiments showed reduced IGF-I receptor expression in granulosa cells from polycystic ovaries, whereas it was sustained or--in hyperinsulinaemic women---even increased in theca cells. Furthermore, in the hyperinsulinaemic women with PCOS, no expression of insulin receptors was found in granulosa or theca cells, suggesting downregulation by hyperinsulinaemia (Samoto et al, 1993). Similar results have been reported for stroma tissue from hyperthecotic ovaries (Nagamani and Smart, 1990). Thus it appears that the increased IGF-I-receptor ratio between theca and granulosa might specifically promote androgen production, by either insulin or IGFs. In vivo experiments (16 publications according to Medline) have, with few exceptions, failed to show any effects of short-term increased plasma insulin on serum androgens. It is possible that the insulin levels and the time of exposure have in general been too modest for an effect to be shown, although stimulation for up to 12 hours and at supraphysiological levels during clamp procedures has also yielded conflicting results (Nestler et al, 1987; Micic et al, 1988; Fox et al, 1993). These results rule out the possibility that post-meal hyperinsulinaemia should contribute to hyperandrogenaemia in these women in general. In contrast, prolonged therapy (35 days) with intravenous insulin in supraphysiological levels in a severely insulin-resistant (type B) woman resulted in two- and four-fold increased levels of androstenedione and testosterone respectively, in parallel with increasing ovarian volumes (DeClue et al, 199 l). Suppression of gonadotrophins with gonadotrophin releasing hormone (GnRH) analogues for various lengths of time consistently results in normalized levels of androgens, in spite of persistent hyperinsulinaemia (Dunaif et al, 1990; Lanzone et al, 1990; Dale et al, 1992). These results strongly suggest that clinical effects of insulin on ovarian androgen production are dependent on a simultaneous LH influence (Poretsky and Piper, 1994). Pharmacologically suppressed insulin levels in obese, insulinresistant women with PCOS resulted in decreased levels of testosterone (Nestler et al, 1989). A similar effect was not seen in lean control women, and, in the PCOS women, androstenedione levels were unaltered in conjunction with signs of increased aromatization. Similar results have

226

J. HOLTE

subsequently been reported in men (Pasquali et al, 1995). In this context, it should be remarked that there are indications of insulin or IGF-I promoting an increased androgenicity in several peripheral and hepatic actions, such as inhibiting aromatization (Lueprasitsakul et al, 1990), interfering with hepatic metabolism of testosterone and stimulating 5c~-reduction (Horton et al, 1993). This underlines that associations between insulin and active androgens (testosterone, dihydrotestosterone) should not a priori be taken as evidence that these relations are being mediated through ovarian effects. SHBG

Although the associations between indices of insulin resistance and absolute levels of androgen steroids vary between different reports, inverse associations between concentrations of SHBG and insulin are almost consistently found. Thus the association between insulin resistance and androgenicity may largely be mediated through high insulin levels correlating with low SHBG levels, resulting in increased non-SHBGbound, or free, testosterone levels (Dunaif et al, 1987; Kiddy et al, 1989; Sharp et al, 1991; Buyalos et al, 1993; Holte et al, 1994a). Similar associations between levels of SHBG and insulin are also found in men (Andersson et al, 1994) and in normally ovulating euandrogenic women (Evans et al, 1983; Peiris et al, 1993; Holte et al, 1994a). Low serum SHBG concentration is an independent predictor of later onset of NIDDM (Lindstedt et al, 1991), and in vitro studies have provided firm evidence for a direct inhibiting effect of insulin and IGF-I on SHBG production in isolated hepatic cells (Plymate et al, 1988; Singh et al, 1990). Pharmacologically reduced insulin levels were followed by increased serum SHBG concentrations in obese women with PCOS (Nestler et al, 1989). In a crosssectional study, body mass index (BMI) and PCOS exerted independently negative effects on SHBG levels (Holte et al, 1994a). After weight reduction in obese women with PCOS, SHBG levels increased (Kiddy et al, 1992; Guzick et al, 1994) and even normalized, in parallel with decreased insulin levels (Holte et al, 1995). In summary, increased insulin levels may theoretically augment androgenicity through several different mechanisms. However, in vivo interventions in general have proved fairly fruitless in showing that such mechanisms are of clinical importance, with the important exception of insulin-SHBG interactions. Thus it seems clear that insulin may diminish SHBG levels and hence increase the biological availability of potent androgens. A R G U M E N T S F O R A CAUSE-AND-EFFECT RELATIONSHIP B E T W E E N A N D R O G E N S AND INSULIN RESISTANCE

Skeletal muscle tissue composes quantitatively the most important site for insulin-mediated glucose uptake and is thus the major determinant of peripheral insulin sensitivity (Bj6rntorp, 1993). Insulin resistance in.

INSULIN RESPONSES

IN P C O S

227

skeletal muscle is primarily characterized by a decreased activity in the insulin-dependent fraction of the glycogen synthase system, and reduced activity in this system is suggested to be a key lesion preceding NIDDM (Bj6rntorp, 1993). Endogenously high serum androgens in women are associated with a relative increase in insulin-resistant, fast twitch, type II muscle fibres at the expense of the insulin-sensitive, slow-twitch, type I fibres (M~rin et al, 1994). Type II muscles are characterized by a reduced capillary density, which is closely related to insulin resistance. The mechanism behind this association seems to be a direct cause-and-effect relation, as the rate of transfer of insulin from capillary binding sites via the extravascular space to the insulin receptor of the muscle is reduced at low capillarization (Bj6rntorp, 1993). Testosterone administration to female rats, producing only moderately increased serum levels, resulted in an increase in the type II/type I ratio of muscle fibres, decreased muscular capillary density and a marked insulin resistance (Bj6rntorp, 1993). The reduced capillary density seemed to be the major cause of insulin resistance. In humans, insulin resistance and hypersecretion of insulin have been constant findings after the administration of anabolic steroids to women or men, but these effects have largely been attributed to 17-alkylation in such agents (Woodard et al, 1981). Recent evidence from the administration of natural testosterone to testosterone-deficient, insulin-resistant men showed an improved insulin sensitivity, whereas supraphysiological concentrations were again followed by a deterioration in carbohydrate homeostasis (Bj6rntorp, 1993). Furthermore, administration of testosterone esters to female-to-male trans-sexuals induced moderate peripheral insulin resistance (Polderman et al, 1994). Together, these data suggest that there is a physiological 'window' for optimal effects of testosterone on insulin sensitivity, and it seems likely that this is found at different levels in men and women (Bj6rntorp, 1993). Given this background, it appears that endogenously increased testosterone levels in women could directly impose a negative impact on insulin sensitivity. Another action of testosterone with potentially detrimental effects on insulin sensitivity is a facilitating effect on catecholamine-stimulated lipolysis, specifically in abdominal fat tissue (Rebuff6-Scrive et al, 1991). In subjects with truncal-abdominal obesity, such a mechanism could sustain a high flux of free-fatty acids to the liver (intra-abdominal fat) and the periphery (subcutaneous fat). Free fatty acids when in excess may induce hepatic and muscular insulin resistance (Fen'annini et al, 1983). Obviously, this is a mechanism that is dependent on an increased amount of abdominal fat. As described below, recent results support the notion that increased truncal-abdominal fat and a high flux of free fatty acids are closely involved in the pathogenesis of insulin resistance in PCOS (Bringer et al, 1993; Holte et al, 1994b,c, 1995). Furthermore, in vivo and in vitro results indicate that testosterone may interfere with insulin clearance and peripheral degradation (Ferrannini et al, 1982; Buffington and Kitabchi, 1994). It has been suggested that the reverse effects are imposed by dihydroepiandrosterone (DHEA), i.e. stimulating glucose uptake (Nakashima et al, 1995) and thus increasing insulin

228

J. HOLTE

sensitivity. Thus, DHEA/testosterone ratio might reflect insulin sensitivity (Buffington et al, 1991). Intervention studies aiming at reduction of testosterone levels in hyperandrogenic women have so far provided surprisingly little information about the relationship between androgens and insulin. Of the 20 publications on the topic listed in Medline from 1982, 13 documented signs of insulin resistance prior to intervention, and of those only 5 included more than six women. If those 5 studies, suppression with GnRH analogues for 6-12 weeks was applied in three (Dunaif et al, 1990; Lanzone et al, 1990; Elkind et al, 1993), whereas oral. contraceptives (ethinyloestradiol/desogestrel (Rebuffd-Scrive et al, 1989) or spironolactone (Shoupe and Lobo, 1984) were used for 6 months and 2 weeks respectively in the other two studies. Signs of improved insulin sensitivity were attained in the spironolactone study and in a Subgroup of moderately insulin-resistant women (but not those with severe insulin resistance) in one of the studies applying GnRH analogues (Elk~nd et al, 1993), whereas the other three studies yielded negative results. It is not clear, however, whether such fairly short observation times are sufficient, or whether the pharmacological agent itself could counteract an improved insulin sensitivity, as through the hypo-oestrogenaemia caused by GnRH analogues. Thus it can not safely be concluded from any study that the single reduction of increased testosterone concentrations in insulin-resistant women would not be followed by improved insulin sensitivity. Taken together, it seems unlikely that the only moderately increased testosterone levels in most women with PCOS themselves play a major role in inducing insulin resistance. Thus women with moderate hyperandrogenaemia without concomitant anovulation (Dunaif et al, 1987; Conway et al, 1990; Robinson et al, 1993) or increased truncal-abdominal fat mass (Bringer et al, 1993; Holte et al, 1994b, 1995) do not generally exhibit signs of insulin resistance. It can not be excluded, however, that increased testosterone levels act in concert with increased truncal-abdominal fat mass in inducing insulin resistance by the mechanisms described. BODY WEIGHT AND BODY FAT DISTRIBUTION

Obesity Insulin resistance out of proportion to obesity is a uniform finding in overweight women with PCOS. However, the results for lean women, as defined by BMI or percentage of ideal body weight, diverge. Some authors report hyperinsulinaemia or decreased insulin-mediated uptake in lean women (Chang et al, 1983; Dunaif et al, 1989; Rebuff6-Scrive et al, 1989; Conway et al, 1990; Robinson et al, 1992), whereas others do not (Herbert et al, 1990; Dale et al, 1992; Bringer et al, 1993; Ovesen et al, 1993; Weber et al, 1993; Holte et al, 1994b; Rajkhowa et al, 1994). Thus although it seems clear that PCOS and obesity have additive, or synergistic, negative impact on insulin sensitivity, the important question--whether PCOS p e r

INSULIN RESPONSES IN PCOS

229

is associated with aberrations in carbohydrate metabolism--is still under debate. se

Body fat distribution Hyperandrogenicity in women is clearly associated with a preponderance of fat localized to truncal-abdominal sites (Evans et al, 1983). Since abdominal fat has a high lipolytic activity (Bouchard et al, 1993) and there is evidence that enhanced free fatty acid release from excess abdominal fat interferes with hepatic and peripheral insulin action on carbohydrate metabolism (Ferrannini et al, 1983), such a pathogenetic mechanism for insulin resistance in PCOS would seem likely. To investigate this possibility, we studied 49 women with PCOS and 42 controls, both groups covering a wide range of BMI (Holte et al, 1994b). According to skinfold measurements, women with PCOS exhibited more truncal-abdominal fat with increasing BMI than did controls, whereas the control women showed more gluteo-femoral fat over the entire range of BMI (Figure 1A). Insulin sensitivity (as measured with the euglycaemic hyperinsulinaemic clamp) showed a linear decline with increasing BMI in both groups, much more pronounced in women with PCOS than in controls (Figure 1B). For both variables (truncal-abdominal fat mass and insulin resistance), the regression lines crossed in the lower part of the normal range of BMI, so that PCOS and controls would not differ significantly in any of these measurements in arbitrary subgroups of non-obese women (< 26 kg/m2), whereas the different inclinations of the regression lines gave vast differences in both truncal-abdominal fat and insulin resistance in the upper range of BMI. Thus, insulin sensitivity was similar in both groups at BMI 21 kg/m 2, but was 35 and 70% lower in the PCOS group at a BMI of 28 and 35 kg/m 2,

(A)

t

-

'E 140 F

I~ I1~",, , , ~

o~ 100

E

[]

U

/

,,

/

40

co 20 20.0

25.0

30.0 BMI

35.0

40.0

230

J. HOLTE

(B) 15.0

t

II

a

[In

.~°

xt

J

lO.O t~ CO .___

0.0' ~

20.0

25.0

30.0 BMI

35.0

40.0

(c)

•~ 10.0 a MII J

=

.. 0.0

I

20 40 60 80 100 120 140 160 180 Sum of truncal-abdominal skinfolds (cm)

Figure 1. Relationships between (A) the sum of truncal-abdominal skinfolds and BMI, and (B) the insulin sensitivity index during euglycaemic hyperiusulinaemic clamp and BMI, in women with PCOS (squares) and controls. The effects of BMI on the sum of truncal-abdominal skinfolds (P < 0.01) and on the insulin sensitivity (negatively; P < 0.01) were greater in the PCOS group than in controls, resulting in similar levels in the normal range of BMI, but increasing differences with increasing BMI. (C) The impact of truncal-abdominal skinfolds on insulin sensitivity was similar in both groups, with no differences between the groups. Adapted from Holte et al (1994b) with permission.

INSULIN R E S P O N S E S IN PCOS

231

respectively (Figure 1A, B). In both PCOS and controls, insulin sensitivity correlated closely, in an inverse manner, with truncal-abdominal fat, and adjusting the groups statistically to the same level of truncal-abdominal fat made the groups completely indistinguishable in insulin sensitivity over the entire range of BMI (Figure 1C; P = 0.9). Similar results were reached by other investigators, who concluded that abdominal obesity (high waist/hip ratio: WHR) accounted for the impaired insulin sensitivity in obese women with PCOS (Bringer et al, 1993). Interestingly, measures of truncal-abdominal fat (skinfolds) also correlated closely with insulin sensitivity within the normal BMI range. In a previous study, WHR was higher in lean women with PCOS than in controls. This indication of increased abdominal fat, in spite of normal BMI, was accompanied by significant hyperinsulinaemia in the PCOS group (Rebuffd-Scrive et al, 1989). Thus, varying proportions of women with increased truncal-abdominal fat within groups of 'lean' women with PCOS could partly explain the divergent results for measures of insulin sensitivity in non-obese women with PCOS. Thus these cross-sectional data show a very strong association between truncal-abdominal fat mass and reduced peripheral action of insulin on glucose homeostasis. Furthermore, plasma free fatty acid concentrations were increased in obese women with PCOS and conelated closely with insulin resistance in the PCOS group (Holte et al, 1994c). Together, these findings support the hypothesis that insulin resistance in PCOS is highly dependent on these two factors, although these cross-sectional results are not sufficient to clarify the chain of events. To further elucidate these relations, we studied 13 obese, severely insulinresistant women with PCOS who underwent a low calorie weight-reduction programme (Holte et al, 1995). Mean weight loss was 12.4kg (SD 4.7), equalling a reduction from BMI 32.2 (3.7)kg/m 2 to 27.6 (3.7)kg/m 2. The results showed that truncal-abdominal fat, but not subcutaneous gluteofemoral fat (Figure 2A), was significantly reduced. In parallel, free fatty acid levels decreased and insulin sensitivity (Figure 2B) was markedly improved. Importantly, all these three factors reached the level of a normal control group (n = 23), matched to the BMI attained after weight loss (Figure 2A, B). Although testosterone levels and the free androgen index (FAI) were reduced, they were still greater than in controls. In contrast to these findings in the diet group, a similarly BMI-matched (as after diet) group of women with PCOS (n= 21) who had not undergone diet treatment exhibited the expected greater amount of tmncal-abdominal fat, higher levels of free fatty acids and lower insulin sensitivity than did the controls (Figure 2A, B). As reduced fat (truncal-abdominal fat) is definitely a primary event during weight loss, these results support the interpretation that insulin resistance is largely a consequence of increased truncal-abdominal fat in PCOS. This was underlined by a strong correlation between these two factors in all four subgroups (PCOS before and after diet, controls and PCOS without diet). Moreover, a stepwise regression analysis (including hormonal and anthropometric measures) showed that truncal-abdominal fat and free fatty acid levels explained the variation in insulin resistance in PCOS women, with a R 2 value of 0.67 (Holte et al, 1995).

Z-u

0

~8

°?

o

I

I

I

0 ~

::1

I

I

O0

I

Insulin sensitivity index

I

I •-~

I =

I ga-o

0

0

."

0

0

0

0

0

0

3: 0[-.-] m

jZ

0

Sum of truncal-abdominal skinfolds (cm)

233

INSULIN RESPONSES IN PCOS

(c) NS

P
I II

200

I 0

~oo

E E E

w

v

I

II

$

v

|

©

50

It

O r-

$

r'-

-5

P
25

|

¢-

10 PCOBD

PCOAD

C

PCOND

Figure 2. Levels of (A) truncal-abdominal skin folds, (B) insulin sensitivity index during euglycaemic hyperinsulinaemic clamp and (C) the early insulin release after intravenous glucose in obese women with PCOS before (PCO-BD) and after weight reduction (PCO-AD), in controls (C) and women with PCOS (PCO-ND). The two latter groups were without diet and BMI-matched to PCO-AD. Insulin sensitivity and truncal-abdominal skinfolds were improved to the levels of the controls, whereas the ea~'ly insulin release was unaltered after weight loss. Adapted from Holte et al (1995) with permission.

In summary, both cross-sectional and intervention data show that women with PCOS are in general not insulin resistant in the absence of increased amount of tmncal-abdominal fat, in spite of significant hyperandrogenaemia. Insulin resistance seems largely reversible with normalization of truncal-abdominal fat. Thus insulin resistance does not seem to be a general feature of PCOS p e r se, but may in most women with PCOS be a consequence of increased tmncal-abdominal fat and high levels of free fatty acids. FACTORS D E T E R M I N I N G BODY FAT DISTRIBUTION IN PCOS Given the results above, elucidating the factors that regulate body fat distribution would give important information on the pathogenesis of insulin resistance in PCOS. Although the term 'male' or 'android' body fat distribution suggests the involvement of androgen hormones, testosterone itself seems to promote a reduction in this fat tissue (Rebufft-Scrive et at, 1991) (see above), and low testosterone levels are associated with abdominal

234

J. HOLTE

obesity in men (Simon et al, 1992). Results from studies on adipose tissue in women with PCOS suggest that these mechanisms are also valid in women, as the testosterone/SHBG ratio correlated positively with lipolysis activity in the abdominal region, and inversely with lipoprotein lipase activity in both abdominal and femoral regions (Rebuff6-Scrive et al, 1989). However, in large groups of women with PCOS, testosterone levels show a slight increase with BMI (Conway et al, 1990), and increasing truncal-abdominal fat (Holte et al, 1994a). For the FAI, this correlation with truncal-abdominal fat is more pronounced, as SHBG levels are inversely con'elated with BMI and truncal-abdominal fat. In contrast, women with normal ovaries and regular menstrual cycles exhibit constantly low total testosterone levels and FAI over the entire range of BMI, in spite of lowered SHBG levels, suggesting an intact 'servocontrol' mechanism, regulating potent androgens in simple obesity (Samojlik et al, 1984; Zhang et al, 1984; Holte et al, 1994c). Gluteo-femoral fat tissue appears to be promoted by oestrogen and progesterone, presumably through indirect mechanisms, as receptors for these steroids are not identified in human adipose tissue (Rebuffd-Scrive et al, 1990). Gluteo-femoral fat has a low lipolytic activity and is regarded as a storage organ for pregnancy and lactation. Interestingly, aromatization appears to be increased in women with lower body obesity, whereas conversion of androstenedione to testosterone may be more pronounced in abdominal fat (Perel and Killinger, 1979; Kirschner et al, 1990). Gtucocorticoid receptors are abundant in abdominal fat tissue, and cortisol stimulates lipoprotein lipase in that tissue, thus promoting fat accumulation. Progesterone seems to exert anti-cortisol effects through interaction with the cortisol receptor (Rebuff6-Scfive et al, 1985). It seems reasonable to assume that anovulation, with its relative decrease in oestrogen (also in women with PCOS) and almost total lack of progesterone, would result in a less favourable steroid balance for the accumulation of gluteo-femoral fat. Lack of the normal differences in adipose cell size and lipoprotein lipase activity between abdominal and femoral adipose tissue in women with PCOS has been attributed to the absence of progesterone (Lithell et al, 1987; Rebuff6-Scrive et al, 1989). The total effect of anovulation at a positive energy balance may well be a prefelTed accumulation of fat in the abdominal region. Thus, anovulation in a woman with excess intake of calories, a common combination in PCOS, is likely to result in the development of increased truncal-abdominal fat. This interpretation is supported by the findings of insulin resistance or hyperinsulinaemia in anovulatory, but not regularly ovulating, women with PCOS (Dunaif et al, 1987; Conway et al, 1990; Robinson et al, 1993), although no study has addressed the question of different body fat distribution in these groups. It seems unlikely that anovulation itself would have immediate positive effects on insulin sensitivity, as insulin sensitivity appears rather to be decreased during the luteal phase of an ovulatory cycle (Valdes and Elkind-Hirsch, 1991). An alternative interpretation is that hyperinsulinaemia is present prior to anovulation and actually interferes with the ovulation mechanisms through insulin/IGF-I interactions in the ovary (Robinson et al, 1993). The

INSULIN RESPONSES IN PCOS

235

two hypotheses (insulin resistance caused indirectly by anovulation, and anovulation caused by hyperinsulinaemia) are not necessarily contradictory. Hyperinsulinaemia (either primary, or secondary to insulin resistance) might induce anovulation, and the resulting steroid balance at anovulation could secondarily promote accumulation of tmncal-abdominal fat, and hence induction or aggravation of insulin resistance. Although anovulation may be a crucial event in the development of truncal-abdominal fat and insulin resistance in PCOS, increased secretion or action of corticosteroids by any cause would presumably give the same aberrant body fat distribution. It is well accepted that a number of women with PCOS show biochemical evidence of an increased adrenal activity. Several recent studies suggest that many women with PCOS or truncal-abdominal obesity exhibit an overall increased cortisol secretion, which may be due to an increased sensitivity of the hypothalamicpituitary-adrenal axis (Mfirin et al, 1992; Pasquali et al, 1993; Miller et al, 1994; Rodin et al, 1994; Lanzone et al, 1995). Young women with PCOS exhibited signs of a more labile blood pressure with greater increases in day-time heart rate and ambulatory mean arterial blood pressure, findings that could be compatible with an aberrant neuroendocrine constitution (Gennarelli et al, 1995). Together, evidence is accumulating that at least a proportion of women with PCOS are exposed to increased cortisol effects. Notably, apart from providing a steroid balance favouring truncal-abdominal fat accumulation at the expense of gluteo-femoral fat, cortisol itself may impair insulin sensitivity (Moller and Flier, 1991). INSULIN RESISTANCE IN P C O S - - F I N D I N G S IN T H E TARGET CELL The identification of mutations in the insulin receptor or other target cell defects in insulin action (Type A) or autoantibodies to the insulin receptor (Type B) in women with the HAIR-AN syndrome provided evidence that the chain of events leading to both hyperinsulinaemia and PCOS/hyperthecosis could indeed be initiated at the level of the insulin target cell (Kahn and Flier, 1976). In recent years, the literature on the insulin receptor, the post-receptor signalling pathways and glucose transport activity in PCOS has expanded considerably. The findings are, however, not uniform. Thus, some investigators have found decreased cell surface insulin receptor number or affinity in varying target cells (Marsden et al, 1994), whereas others have not (Ciaraldi et al, 1992; Dunaif et al, 1992). Although single patients with mutations in the tyrosine kinase domain of the insulin receptor gene were identified, this defect was not found in a study comprising 22 women with PCOS (Conway et al, 1994). Conversely, a complex abnormality involving both decreased receptor tyrosine phosphorylation and tyrosine kinase activity, and increased insulin-independent receptor phosphorylation was reported in approximately 50% of PCOS women (Dunaif et al, 1995). Interestingly, as this aberration was found in cultured

236

J. HOLTE

cells, it may be a genetic rather than an acquired defect. A post-receptor defect in the signal transduction from the receptor kinase to glucose transport has been suggested (Ciaraldi et al, 1992). Furthermore, diminished expression of GLUT-4 glucose transporters was seen in PCOS, to a similar degree as in obesity, and GLUT-4 content correlated closely with insulinstimulated glucose transport in adipocytes (Rosenbaum et al, 1993). Taken together, it seems likely that insulin resistance will eventually be shown to correspond to defects in the target organ in an increasing portion of women with PCOS. However, it remains to be shown which defects are of primary, genetic origin and which are secondary to other factors in PCOS, for example steroid levels, obesity or indeed hyperinsulinaemia, and therefore variable and potentially reversible (Freidenberg et al, 1988). INSULIN SECRETION AND METABOLISM IN PCOS Although much debated, controversy still remains as to the primary event in the relationship between insulin resistance and hyperinsulinaemia in the insulin resistance syndrome (Moller and Flier, 1991). Most authors favour insulin resistance as the initiating factor, with increased insulin secretion a necessary response to maintain glucose homeostasis. Alternatively, hypersecretion of insulin could be the first step in a chain of events, resulting in receptor or post-receptor defects in target cells (Marshall and Olefsky, 1980), or could, as shown in animal experiments (Le Marchand et al, 1985), stimulate hyperphagia and subsequent obesity and insulin resistance. Clearly, the development of beta cell failure and NIDDM, which could be the result in either case, complicates matters further, and this is the stage at which most patients are being investigated. A high prevalence of impaired glucose tolerance and frank NIDDM are repeatedly found in women with PCOS (Dunaif et al, 1987, 1989; Dahlgren et al, 1992; Holte et al, 1994b). These patients generally exhibit a subnormal first-phase insulin secretion, in absolute terms or in relation to the degree of insulin resistance. High insulin levels may be preserved in the fasting state and during the mid-to-later phases of glucose tolerance tests, mainly because of reduced insulin clearance, but glucose levels are increased during such tests (Holte et al, 1994b). Most of these patients are obese, and the prevalence of beta cell failure in the obese subgroup of PCOS seems particularly high (Dunaif et al, 1989; O'Meara et al, 1993; Holte et al, 1994b). Heredity for NIDDM (Ehrmann et al, 1995) or previous gestational diabetes (Holte, unpublished data) is associated with lower firstphase insulin secretion in women with PCOS. It is clear, however, that the majority of women with PCOS sustain a normal glucose tolerance at the expense of high insulin secretion. The firstphase insulin secretion, measured as the early (4-8 minute) incremental rise was even found to be higher (on average 53%) in PCOS women with normal glucose tolerance than in controls (Figure 3A) and, importantly, higher than what could be expected from insulin resistance p e r s e (Figure 3B). This finding was accomplished by measuring insulin sensitivity and

237

INSULIN RESPONSES IN PCOS

24 f

]

(A)

"

90

**

[]

.

[]

[]

. * * "#

* Ig "

*

55

.~ 33,.._ 20•~mltnl~lll~' I I N/ I I 12 20.0• l 25.0 30.0 35.0 40.0 BMI (B) 245

150 K i l n ".-'" 90 ~

-. ,, ""

55

- ' ; %-2--.. 12

0.0

l

I

I

HI

i [] I

5.0 10.0 15.0 Insulinsensitivityindex

Figure 3. Relationships between the early insulin secretion after intravenous glucose and (A) BMI and (B) the insulin sensitivity index during euglycaemic hyperinsulinaemic clamp in women with PCOS (squares) and controls, in contrast to insulin sensitivity (Figure IB), the early insulin release was greater in the PCOS group over the entire range of BMI (P < 0.0 l), with a similar impact of BMI in both groups. The em'ly insulin release was greater than accounted for by insulin sensitivity (P < 0.01). Women with glucose intolerance ale not included. Adapted from Holte et al (1994b) with permission.

238

J. HOLTE

insulin secretion separately (euglycaemic hyperinsulinaemic clamp and IVGTT respectively), and adjusting insulin increment statistically for the degree of insulin resistance. Such an increased first-phase insulin secretion was also found, together with increased fasting levels of C-peptide but normal insulin, within the normal range of BMI in which the women with PCOS had normal insulin sensitivity (Holte et al, 1994b). Furthermore, in contrast to insulin resistance, the increased insulin increment was still significant after controlling for body fat distribution (Holte et al, 1994b) and was unaltered after weight loss (Figure 2C above; Holte et al, 1995). Together, these findings suggest there is a component of first-phase insulin hypersecretion in a large portion of women with PCOS, out of proportion for insulin resistance, body weight and body fat distribution. Similar findings of increased first-phase insulin release have been reported from Pima indians, an ethnic group with an exceedingly high prevalence of obesity, insulin resistance and NIDDM (Lillioja et al, 1991). Furthermore, an enhanced post-prandial insulin increment persisted in women with morbid obesity after substantial weight loss and improved insulin sensitivity (Fletcher et al, 1989). In animal studies, an enhanced insulin secretion at normal insulin sensitivity precedes the development of obesity and insulin resistance or hypertension (Le Marchand et al, 1985; Le Stunff and Bougn~res, 1994), and similar findings were reported for adolescent boys with subsequent development of obesity. In such studies, the increased insulin secretion is accompanied by evidence of increased insulin effectiveness. Interestingly, there are also indications of this in women with PCOS. Thus, we found a positive correlation between the insulin increment and glucose disappearance during the IVGTT, in conjunction with lower glycated haemoglobin (HbAL) level in the women with PCOS than in controls. Previous investigators also found reduced glycated haemoglobin (Golland et al, 1989), lower levels of glucose in the fasting state (Pasquali et al, 1991) or during 24-hour measurements (lean women; Prelevic et al, 1992), or a lack of the normal post-prandial increase in blood glucose (obese women; Prelevic et al, 1992) in PCOS. In addition, signs of physiological adaptation to hypoglycaemia have recently been found in obese women with PCOS (Gennarelli, unpublished data). Together, these data support the interpretation that a substantial portion of women with PCOS exhibit hypersecretion of insulin, which is not secondary to insulin resistance but rather a sign of increased beta cell mass, or increased beta cell sensitivity to glucose. It seems reasonable to assume that the associated depressed blood glucose levels may play a role in stimulating carbohydrate craving and weight gain in these women. It remains to be shown whether the lean women with hyperinsulinaelnia will be at increased risk of later weight gain and insulin resistance, but the association between bulimia and PCOS might support the existence of abnormal carbohydrate craving in lean women with PCOS (McCluskey et al, 1991; Jahanfar et al, 1995; Holte, unpublished data). Is the observed hypersecretion of insulin a primary abnormality in PCOS, as part of a neuroendocrine constitution, or could the beta cells be

INSULIN R E S P O N S E S IN PCOS

239

sensitized by the aberrant hormone balance in PCOS? Supporting the former alternative may be the findings of an association between plasma levels of insulin and [3-endorphins in hyperinsulinaemic women with PCOS (Givens et al, 1987; Laatikainen et al, 1989; Lanzone et al, 1991; Carmina et al, 1992). [3-Endorphin levels increased in parallel with those of insulin after glucose administration, and treatment with opioid antagonists resulted in a reduced insulin response (Givens et al, 1987; Fulghesu et al, 1993; Lanzone et al, 1994). Conversely, a support for the hypothesis that beta cells are sensitized to hypersecrete insulin is the observation of an association between potent androgens (FAI) and the insulin increment (Holte et al, 1994b, 1995). A direct dose-response relationship does not, however, seem likely as the first-phase insulin release was essentially unaltered after weight loss, in spite of decreased levels of testosterone and FAI (Holte et al, 1995). Increased insulin secretion after administration of anabolic steroids is documented, although this is presumably largely secondary to induced insulin resistance (Landon et al, 1962; Cohen and Hickman, 1987). It is clear from studies on oral contraceptives, however, that steroid levels can affect beta cell action apart from insulin sensitivity (Godsland et al, 1992). Unfortunately, no studies have addressed the question of early insulin response in women with PCOS after reduction of androgen levels only. Clearly, when all ovarian steroids, including oestrogens, are reduced during GnRH analogue treatment, hyperinsulinaemia persists. Again, as discussed above, the derangement of total steroid balance and cyclicity in anovulation may be of greater importance than testosterone levels p e r se. Among factors with known influence on beta cells, glucocorticoids and IGF-I stimulate, and oestrogen and progesterone seem to inhibit, beta cell growth (Swenne, 1992). Thus there are several hormones/growth factors that by themselves or in combinations may affect beta-cells, the net result being increased insulin release in PCOS. Intervention studies focused on the restitution of ovulation in PCOS are highly warranted for further information on the specific role of anovulation. Limited data on insulin levels shortly after induced ovulation by means of laparoscopic ovarian electrocautery have so far been discouraging (increased or unaltered levels) (Gadir et al, 1990; Tiitinen et al, 1993), but longer follow-up periods and more detailed information on insulin secretion and sensitivity are needed to draw valid conclusions from such treatment. In summary, although most of the observed hyperinsulinaemia in PCOS is probably secondary to insulin resistance, there seems to be an important component of insulin secretion which is independent of insulin resistance, body weight and body fat distribution. This may be a primary disturbance in PCOS, or caused by the deranged hormonal balance in PCOS, and might have pathogenetic implications for increased carbohydrate intake and subsequently obesity and insulin resistance. At the other end of the spectrum, glucose intolerance and frank NIDDM is found in combination with signs of beta cell exhaustion in a substantial portion of the women with PCOS, in most cases presumably representing a later stage after longstanding insulin resistance in susceptible women.

240

J. HOLTE

SUMMARY Insulin resistance, defined as a diminished effect of a given dose of insulin on glucose homeostasis, is a highly prevalent feature of women with PCOS. Insulin resistance in PCOS is closely associated with an increase in truncal-abdominal fat mass, elevated free fatty acid levels, increased androgens, particularly free testosterone through reduced SHBG levels, and anovulation. The causes for insulin resistance in PCOS are still unknown. One line of evidence suggests that an increase in truncal-abdominal fat mass and subsequently increased free fatty acid levels induce insulin resistance in women with PCOS. Increased effects of corticosteroids and a relative reduction in oestrogen and progesterone seem to be involved in the aberrant body fat distribution. Conversely, there are also results supporting primary, genetic target cell defects as a cause of insulin resistance in PCOS. An explanation for these seemingly contradictory results could be that the group of women with PCOS is heterogeneous with respect to the primary event in carbohydrate/insulin disturbances. Also insulin secretion in PCOS is characterized by heterogeneity. At one end of the spectrum is a large subgroup of mainly obese women with reduced insulin secretion, which appears to result from failure of the beta cells to compensate for insulin resistance in susceptible women, resulting in glucose intolerance and NIDDM. In the insulin-resistant patients with normal glucose tolerance, most of the hyperinsulinaemia is probably due to secondarily increased insulin secretion and decreased insulin degradation. However, a component of the increased first-phase insulin release is not due to measurable insulin resistance. Notably, this is also found in lean women with normal insulin sensitivity, and is not reversed after weight reduction, in contrast to the findings for insulin resistance. The implications of this enhanced insulin release are not fully clear, but it may tentatively be associated with carbohydrate craving and subsequently increased risks for development of obesity and insulin resistance. It may represent a primary disturbance of insulin secretion in PCOS or may be associated with the perturbed steroid balance in anovulation. The insulin-androgen connection in PCOS appears to be amplified by several different mechanisms, notably in both directions, the initiating event probably varying between individuals. Thus insulin increases the biological availability of potent steroids, primarily testosterone, through the suppression of SHBG synthesis. Insulin is also involved as a progonadotrophin in ovarian steroidogenesis, with the possible net result of interfering with ovulation and/or increasing ovarian androgen production in states of hyperinsulinaemia. Conversely, testosterone may indirectly contribute to insulin resistance through facilitating free fatty acid release from abdominal fat, but perhaps also through direct muscular effects at higher serum levels. It seems likely that this constitution, presumably genetic, would provide evolutionary advantages in times of limited nutrition, given the energysaving effects of insulin resistance. Hypothetically, hyperinsulinaemia (primary) could provide a stimulus to ensure intake of nourishment, but

INSULIN RESPONSES IN PCOS

241

unlimited food supplies could in some cases initiate a vicious 'anabolic' circle, in which several of the proposed amplifying mechanisms between insulin and androgens--in both directions--could take part. Acknowledgements The author wishes to thank Ass. Prof. Torbj6rn Bergh, Ass. Prof. Christian Berne, Professor Hans Lithell, Professor Leif Wide, Dr Gianluca Gennarelli and Lars Berglund, statistician, for invaluable collaboration during our studies on insulin resistance in PCOS. The author also wants to acknowledge Dr Christian Berne for reviewing the manuscript. Our studies were supported by the Swedish Medical Research Council Grants 3495 and 5446.

REFERENCES Achard C & Thiers J (1921) Le vifilisme pilaim et son association a l'insuffisance glycolytique (diab6te des femmes 5_barbe). Bulletin Academic National Medecin 86: 51-64. Amiel SA, Capfio S, Sherwin RS et al (1991) Insulin resistance in puberty: a defect restricted to peripheral glucose metabolism. Journal of Clinical Endocrinology and Metabolism 72: 277-282. Andersson B, Matin P, Lissner L e t al (1994) Testosterone concentrations in women and men with NIDDM. Diabetes Care 1 7 : 4 0 5 4 1 1 . Barbieri RL & Ryan KJ (1983) Hyperandrogenism, insulin resistance, and acanthosis nigricans syndrome: a common endocfinopathy with distinct pathophysiologic features. American Journal of Obstetrics and Gynecology 147: 90-101. Barbieri RL, Makris A, Randall RW et al (1986) Insulin stimulates androgen accumulation in incubations of ovarian stroma obtained from women with hyperaudrogenism. Journal of Clinical Endocrinology and Metabolism 62:904-910. Bergh C, Carlsson B, Olsson JH et al (1993) Regulation of androgen production in cultured human thecal cells by insulin-like growth factor I and insulin. Fertility and Sterility 59:323-331. Bergman RN, Hope ID, Yang YJ et al (1989) Assessment of insulin sensitivity in vivo: a critical review. Diabetes Metabolic Reviews 5 : 4 1 1 4 2 9 . Bj6rntorp P (1993) Hyperandrogenicity in w o m e n - - a prediabetic condition? Journal of Internal Medicine 234: 579-583. Bonora E, Moghetti P, Zancanaro C et al (1989) Estimates of in vivo insulin action in man: comparison of insulin tolerance tests with euglycaemic and hyperglycaemic clamp studies. Journal of Clinical Endocrinology and Metabolism 50:113-116. Bouchard C, Despr6s J-P & Maufi~ge P (1993) Genetic and nongenetic determinants of regional fat distribution. Endocrine Reviews 14: 72-93. Bringer J, Lefebvre P, Boulet F et al (1993) Body composition and regional fat distribution in polycystic ovarian syndrome. Relationship to hormonal and metabolic profiles. Annals of New York Academy ~'Sciences 687:115-123. Buffington CK & Kitabchi AE (1994) Evidence for a defect in insulin metabolism in hyperandrogenic women with polycystic ovarian syndrome. Metabolism 43:1367-1372. Buffington CK, Givens JR & Kitabchi AE (1991) Opposing actions of dehydroepiandrosterone and testosterone on insulin sensitivity. In vivo and in vitro studies of hyperandrogenic females. Diabetes 40: 693-700. Burghen GA, Givens JR & Kitabchi AE (1980) Correlation of hyperandrogenism with hyperinsulinism in polycystic ovarian disease. Journal of Clinical Endocrinology and Metabolism 50: 113-116. Buyalos RP, Geffner ME, Bersch N e t al (1992) Insulin and insulin-like growth factor-I responsiveness in polycystic ovarian syndrome. Fertility and Sterility 57: 796-803. Buyalos RP, Geffner ME, Watanabe RM et al (1993) The influence of luteinizing hormone aud insulin on sex steroids and sex hormone-binding globulin in the polycystic ovarian syndrome. Fertility and Sterility 60: 626-633. Carmina E, Ditkoff EC, Malizia G et al (1992) Increased circulating levels of immunoreaetive betaendorphin in polyeystic ovary syndrome is not caused by increased pituitary secretion. American Journal of ObsteO'ics and Gynecology 167: 1819-1824.

242

J. HOLTE

Chang R J, Nakamura RM, Judd HL & Kaplan SA (1983) Insulin resistance in nonobese patients with polycystic ovarian disease. Journal of Clinical Endocrinology and Metabolism 57z 356-359. Ciaraldi TP, el Roeiy A, Madm" Z et al (1992) Cellulm" mechanisms of insulin resistance in polycystic ovarian syndrome. Journal ~" Clinical Endocrinology and Metabolism 75: 577-583. Cohen JC & Hickman R (1987) Insulin resistance and diminished glucose tolerance in powerlifters ingesting anabolic steroids. Journal of" Clinical Endocrinology and Metabolism 64: 960-963. Conway GS & Jacobs HS (1990) Acanthosis nigricans in obese women with the polycystic ovary syndrome: disease spectrum not distinct entity. Postgraduate Medical Journal 66: 536-538. Conway GS, Avey C & Rumsby G (1994) The tyrosine kinase domain of the insulin receptor gene is normal in women with hyperinsulinaemia and polycystic ova'y syndrome. Human Reproduction 9: 1681-1683. Conway GS, Jacobs HS, Holly JM & Wass JA (1990) Effects of luteinizing hormone, insulin, insulinlike growth factor-I and insulin-like growth factor small binding protein 1 in the polycystic ovary syndrome. Clinical Endocrinology 33: 593-603. Dahlgren E, Johansson S, Lindstedt G et al (1992) Women with polycystic ovary syndrome wedge resected in 1956 to 1965: a long-term follow-up focusing on natural history and circulating hormones. Fertility and Sterility 57:505-513. Dale PO, Tanbo T, Vaaler S & Abyholm T (1992) Body weight, hyperinsulinemia, and gonadotropin levels in the polycystic ovarian syndrome: evidence of two distinct populations. Fertility and Sterility 58:487-491. Dale PO, Tanbo T, Djoseland O et al (1992) Persistence of hyperinsulinemia in polycystic ovary syndrome after ovarian suppression by gonadou'opin-releasing hormone agonist. Acta Endocrinologica 126: 132-136. DeClue TJ, Shah SC, Marchese M & Malone JI (1991) Insulin resistance and hyperinsulinemia induce hyperandrogenism in a young type B insulin-resistant female. Journal of Clinical Endocrinology and Metabolism 72:1308-131 i. Dunaif A, Segal KR, Futterweit W & Dobrjansky A (1989) Profound peripheral insulin resistance, independent of obesity, in polyeystic ovary syndrome. Diabetes 38:1165-I 174. Dunaif A, Green G, Futterweit W & DoNjansky A 0990) Suppression of hyperandrogenism does not improve peripheral or hepatic insulin resistance in the polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolism 70: 699-704. Dunaif A, Graf M, Mandeli Jet al (1987) Characterization of groups of hyperandrogenic women with acanthosis nigricans, impaired glucose tolerance, and/or hyperinsulinemia. Journal of Clinical Endocrinology and Metabolbsm 65: 499-507. Dunaif A, Green G, Phelps RG et al (1991) Acanthosis nigricans, insulin action, and hyperandrogenism: clinical, histological, and biochemical findings. Journal of Clinical Endocrinology and Metabolism 73: 590-595. Dunaif A, Segal KR, Shelley DR et al (1992) Evidence for distinctive and intrinsic defects in insulin action in polycystic ovary syndrome. Diabetes 41" 1257-1266. Dunaif A, Xia J, Book CB et al (l 995) Excessive insulin receptor serine phosphorylation in cultured fibroblasts and in skeletal muscle. A potential mechanism for insulin resistance in the polycystic ovary syndrome. Journal ~" Clinical Investigation 96:801-810. Eaton RP, Allen RC, Schade DS et al (1980) Prehepatic insulin production in man: kinetic analysis using peripheral connecting peptide behavior. Journal of" Clinical Endocrinology and Metabolism 51: 520-528. Ehrmann DA, Sturis J, Byrne MM et al (1995) Insulin secretory defects in polycystic ovm'y syndrome. Relationship to insulin sensitivity and family history of non-insulin-dependent diabetes mellitus. Journal of Clinical h~vestigation 96: 520-527. el Roeiy A, Chen X, Roberts VJ et al (1994) Expression of the genes encoding the insulin-like growth factors (tGF-I and II), the IGF and insulin receptors, and IGF-binding proteins-l-6 and the localization of their gene products in normal and polycystic ovary syndrome ovaries. Journal of Clinical Endocrinology and Metabolism 78: 1488-1496. Elkind HK, Valdes CT & Malinak LR (1993) Insulin resistance improves in hyperandrogenic women treated with Lupron. Fertility and Sterility 60: 634-641. Evans D J, Hoffmann RG, Kalkhoff RK & Kissebah AH (1983) Relationship of androgenic activity to body fat topography, fat cell morphology, and metabolic aberrations in premenopausal women. Journal 05[Clinical Endocrinology and Metabolism 57:304-310. Evans D J, Hoffman RG, Kalkhoff RK & Kissebah AH (1984) Relationship of body fat topography to insulin sensitivity and metabolic profiles in premenopausal women. Metabolism 33: 68-75.

INSULIN RESPONSES IN PCOS

243

Fen'annini E, Buzzigoli G & B onadonna RC (1987) Insulin resistance in essential hypertension. New England Journal of Medicine 317: 350-357. Ferrannini E, Muggeo M, Navalesi R & Pilo A (1982) Impaired insulin degradation in a patient with insulin resistance and acanthosis nigricans. American Journal (~fMedicine 73: 148-154. Fenannini E, BmTett EJ, Bevilacqua S & De Fronzo RA (1983) Effect of fatty acids on glucose production and utilization in man. Journal of Clinical Investigation 72: 1737-1747. Fletcher JM, McNurlan MA & McHardy KC (1989) Residual abnormalities of insulin secretion and sensitivity after weight loss by obese women. European Journal of Clinical Nutrition 43: 539-545. Fox JH, Licholai T, Green G & Dunaif A (1993) Differential effects of oral glucose-mediated versus intravenous hyperinsulinemia on circulating androgen levels in women. Fertility and Sterility 60: 994-1000. Frayn KN (1993) Insulin resistance and lipid metabolism. Current Opinion in Lipidology 4: 197204. Freidenberg GR, Reichart D, Olefsky JM & Henry RR (1988) Reversibility of defective adipocyte insulin receptor kinase activity in non-insulin-dependent diabetes mellitus: effect of weight loss. Journal of Clinical Investigation 82: 1398-1406. Fronzo RA de, Tobin JD & Andres R (1979) Glucose clamp technique: a method for quantifying insulin secretion and resistance. American Journal of Physiology 237: 214-223. Fulghesu AM, Lanzone A, Cucinelli F et al (I993) Long-term naltrexone treatment reduces the exaggerated insulin secretion in patients with polycystic ovary disease. Obstetrics and Gynecology 82: 191-197. Gadir AA, Khatim MS, Mowafi RS et al (1990) Hormonal changes in patients with polycystic ovarian disease after ovarian electrocantery or pituitary desensitization. Clinical Endocrinology 32: 749-754. Garzo VG & Dorrington JH (1984) Aromatase activity in human granulosa cells during follicular development and the modulation by follicle-stimulating hormone and by insulin. American Journal of Obstetrics and Gynecology 148: 657-662. Gennarelli G, Holte J, Bergh T et al (1995) Elevated ambulatory day-time blood pressure in women with polycystic ovary syndrome--a sign of a prehypertensive state? Abstracts of the World Congress of Fertility and Sterility 23:S115. Gilling-Smith C, Willis DS, Beard RW & Franks S (1994) Hypersecretion of androstenedione by isolated thecal cells fiom polycystic ovaries. Journal of Clinical Endocrinology and Metabolism 79: 1158-1165. Giudice LC (1995) The insulin-like growth factor system in normal and abnormal human ovarian follicle development. American Journal ~)fMedicine 98: 48S-54S. Givens JR, Kurtz BR, Kitabchi AE et al (1987) Reduction of hyperinsulinemia and insulin resistance by opiate receptor blockade in the polycystic ovary syndrome with acanthosis nigricans. Journal of"Clinical Endocrinology and Metabolism 64: 377-382. Godsland IF, Walton C, Felton C et al (1992) Insulin resistance, secretion, and metabolism in users of oral contraceptives. Journal of Clinical Endocrinology and Metabolism 74: 64-70. Golland 1M, Vaughan WC, Shalet SM et al (1989) Lack of predictive value of HbA1 for impaired glucose tolerance in polycystic ovary syndrome. Gynecological Endocrinology 3: 229-235. Guzick DS, Wing R, Smith D et al (1994) Endocrine consequences of weight loss in obese, hyperandrogenic, anovulatory women. Fertility and Sterility 61" 598-604. Herbert CIII, Hill GA & Diamond MP (1990) The use of the intravenous glucose tolerance test to evaluate nonobese hyperandrogenic women. Fertility and Sterility 53: 647-653. Holte J, Bergh T & Lithell H (1991) Insulin resistance in polycystic ovary syndrome. In Coelingh Bennink HJT, Vemer HM & van Keep PA (eds) Chronic H37~erandrogenic Anovulation, pp 55-60. New Jersey: Parthenon. Holte J, Bergh T, Gennarelli G & Wide L (1994a) The independent effects of polycystic ovary syndrome and obesity on serum concentrations of gonadotrophins and sex steroids in premenopausal women. Clinical Endocrinology 41:473-481. Holte J, Bergh T, Berne C et al (1994b) Enhanced early insulin response to glucose in relation to insulin resistance in women with polycystic ovary syndrome and normal glucose tolerance. Journal of Clinical Endocrh~ology and Metabolism 78: 1052-1058. Holte J, Bergh T, Berne C & Lithell H (1994c) Serum lipoprotein lipid profile in women with the polycystic ovary syndrome: relation to anthropomctric, endocrine and metabolic variables. Clinical Endocrinology 41:463-471.

244

J. HOLTE

Holte J, Bergh T, Berne C et al (1995) Restored insulin sensitivity but persistently increased early insulin secretion after weight loss in obese women with polycystic ovary syndrome. Journal of' Clinical Endocrinology and Metabolism 80: 2586-2593. Homburg R, Pariente C, Lunenfeld B & Jacobs HS (1992) The role of insulin-like growth factor-1 (IGF-1 ) and IGF binding protein-1 (IGFBP-1) in the pathogenesis of polycystic ovary syndrome. Human Reproduction 7: 1379-1383. Horton R, Pasupuletti V & Antonipillai I (1993) Androgen induction of steroid 5 alpha-reductase may be mediated via insulin-like growth factor-I. Endocrinology 133:447-451. Jahanfar S, Eden JA & Nguyent TV (1995) Bulimia nervosa and polycystic ovary syndrome. GynecologicM Endocrinology 9:113-117. Kahn BB & Flier JS (1976) The syndromes of insulin resistance and acanthosis nigricans: insulinreceptor disorders in man. New England Journal of Medicine 294: 739-745. Kazer RR, Unterman TG & Glick RP (1990) An abnormality of the growth hormone/insulin-like growth factor-1 axis in women with polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolis m 71: 958-962. Kiddy DS, Hamilton FD, Seppala M e t al (1989) Diet-induced changes in sex hormone binding globulin and free testosterone in women with normal or polycystic ovaries: correlation with serum insulin mad insulin-like growth factoM. Clinical Endocrinology 31: 757-763. Kiddy DS, Hamilton FD, Bush A et al (1992) Improvement in endocrine and ovm'ian function during dietary treatment of obese women with polycystic ovary syndrome. Clinical Endocrinology 36: 105-111. Kirschner MA, Samojlik E, Brejka M e t al (1990) Androgen-estrogen metabolism in women with upper body versus lower body obesity. Journal of Endocrinology and Metabolism 70: 473-479. Laatikainen TJ, Tiitinen AE, Salminen LK & Pekonen F (1989) Response of plasma beta-endorphin and insulin to oral glucose tolerance test in non-obese women with polycystic ovaries. Gynecological Endocrinology 3: 241-247. Laatikainen T, Anttila L, Suikkmi AM et al (1990) Effect of naloxone on plasma insulin, insulin-like growth factor I, and its binding protein 1 in patients with polycystic ovarian disease. Fertility and Sterility 54: 434-437. Landon J, Wynn V, Houghton BJ & Cooke JNC (1962) Effects of anabolic steroid methandienone, on carbohydrate metabolism in man. I1. Effect of methandienone on response to glucagon, adrenalin, and insulin in the fasted subject. Metabolism 11: 513-523. Lanzone A, Fulghesu AM, Andreani CLet al (1990) Insulin secretion in polycystic ovarian disease: effect of ovarian suppression by GnRH agonist. Human Reproduction 5: 143-149. Lanzone Z, Fulghesu AM, Fortini A et al (1991) Effect of opiate receptor blockade on the insulin response to oral glucose load in polycystic ovarian disease. Human Reproduction 6: 1043-1049. Lanzone A, Fulghesu AM, Guido M e t al (1994) Differential androgen response to adrenocorticotrophin hormone stimulation and effect of opioid antagonist on insulin secretion in polycystic ovarian syndrome. Human Reproduction 9: 2242-2246. Lanzone A, Petraglia F, Fulghesu AM et al (1995) Corticotropin-releasing hormone induces an exaggerated response of adrenocorticotrtopic hormone and cortisol in polycystic ovary syndrome. Fertility and Sterility 63:1195-1199. Le Marchand Y, Freychet P & Jeanrenaud B (1985) Longitudinal study on the establishment of insulin resistance in hypothalamic obese mice. Endocrinology 102: 74-85. Le Stunff C & Bourgnrres P (1994) Early changes in postprandial insulin secretion, not in insulin sensitivity, characterize juvenile obesity. Diabetes 43: 696-702. Lee EJ, Lee BS, Lee HC et al (t993) Growth hormone response to L-dopa and pyridostigmine in women with polycystic ovarian syndrome. Fertilio, and Sterility 60: 53-57. Lillioja S, Nyomba BL, Sand MF et al (1991) Exaggerated early insulin release and insulin resistance in a diabetes-prone population: a metabolic comparison of Pima Indians and Caucasians. Journal of Clinical Endocrinology and Metabolism 73: 866-876. Lindstedt G, Lundberg P-A, Lapidus L e t al (1991) Low sex-hormone-binding globulin concentrations as independent risk factor for development of NIDDM. Diabetes 40: 123-128. Lithell H, Nillus SJ, B ergh T & Selinus I (1987) Metabolic profile in obese women with the polycystic ovary syndrome. International Journal of" Obesity 11: 1-8. Lueprasitsakul P, Latour D & Longcope C (1990) Aromatase activity in human adipose tissue stromal ceils: effect of growth factors. Steroids 55: 540-544. McCluskey S, Evans C, Lacey JH et al (1991) Polycystic ovary syndrome and bulimia. Fertility and Sterility 55:287-29 l.

INSULIN RESPONSES IN PCOS

245

Magoffin DA & Weitsman SR (1993a) Differentiation of ovarian theca-interstitial ceils in vitro: regulation of 17 alpha-hydroxylase messenger ribonucleic acid expression by luteinizing hormone and insulin-like growth factor-I. Endocrinology 132:1945-1951. Magoffin DA & Weitsman SR (1993b) Insulin-like growth factor-I stimulates the expression of 3 beta-hydroxysteroid dehydrogenase messenger ribonucleic acid in ovarian theca-interstitial cells. Biologics and Reproduction 48:1 !66-1173. Magoffin DA, Kurtz KM & Erickson GF (1990) Insulin-like growth factor-I selectively stimulates cholesterol side-chain cleavage expression in ovarian theca-interstitial cells. Molecular Endocrinology 4: 489-496. Mguin P, Andersson B, Krotkiewski M & Bjrrntorp P (I 994) Muscle fiber composition and capillary density in women and men with NIDDM. Diabetes Care 17: 382-386. Mfirin P, Darin N, Amemiya T et al (1992) Co~isol secretion in relation to body fat distribution in obese premenopausal women. Metabolism 41: 882-886. Marsden PJ, Murdoch A & Taylor R (1994) Severe impairment of insulin action in adipocytes from anaenorrheic subjects with polycystic ovary syndrome. Metabolism 43: 1536-1542. Marshall S & Olefsky JM (1980) Effects of insulin incubation on insulin binding, glucose transport, and insulin degradation by isolated rat adipocytes: evidence for induction of insulin insensitivity. Journal of Clinical Investigation 66: 763-772. Mason HD, Willis DS, Beard RW et al (1994) Estradiol production by granulosa cells of normal and polycystic ovaries: relationship to menstrual cycle history and concentrations of gonadotropins and sex steroids in follicular fluid. Journal (~' Clinical Endocrhwlogy and Metabolism 79: 1335-1360. Micic D, Popovic V, Nesovic M e t al (1988) Androgen levels during sequential insulin euglycemic clamp studies in patients with polycystie ovary disease. Journal of Steroid Biochemistry 31: 995-999. Miller JE, Bray MA, Faiman C & Reyes FI (1994) Characterization of 24-h cortisol release in obese and non-obese hyperandrogenic women. Gynecological Endocrinology 8: 247-254. Moiler DE & Flier JS (1988) Detection of an alteration in the insulin-receptor gene in a patient with insulin resistance, acanthosis nigricans, and polycystic ovary syndrome (type A insulin resistance). New England Journal of Medicine 319:1526-i 529. Moiler DE & Flier JS (1991) Insulin resistance--mechanisms, syndromes, and implications. New England Journal of Medicine 325: 938-948. Nagamani M & Stuart CA (1990) Specific binding sites for insulin-like growth factor I in the ovarian stroma of women with polycystie ovarian disease and stromal hypertheeosis. Anzerican Journal (~f Obstetrics and Gynecology 163: 1992-1997. Nahum R, Thong KJ & Hillier SG (1995) Metabolic regulation of androgen production by human thecal cells in vitro. Human Reproduction 10:75-81. Nakashima N, Haji M, Sakai Y e t al (1995) Effect of dehydroepiandrosterone on glucose uptake in cultured human fibroblasts. Metabolism 44: 543-548. Nestler JE, Clore JN, Strauss JIII & Blackard WG (1987) The effects of hyperinsulinemia on serum testosterone, progesterone, dehydroepiandrosterone sulfate, and cortisol levels in normal women and in a woman with hyperandrogenism, insulin resistance, and acanthosis uigricans. Journal of Clinical Endocrinology and Metabolism 64:180-184. Nestler JE, Barlascini CO, Matt DW et al (1989) Suppression of serum insulin by diazoxide reduces serum testosterone levels in obese women with polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolism 68: 1027-1032. O'Meara NM, Blackman JD, Ehrmann DA et al (1993) Defects in beta-cell function in functional ovarian hyperandrogenism. Journal (~'Clinical Endocrinology and Metabolism 76: 1241-1247. Ovesen P, Moiler J, Ingerslev HJ et al (1993) Normal basal and insulin-stimulated fuel metabolism in lean women with the polycystic ovary syndrome. Journal of Clinical Endocrinology and Metabolism 77:1636-1640. Pasquali R, Cantobelli S & Casimiri F (1993) The hypothalamic-pituitary-adrenal axis in obese women with different patterns of body fat distribution. Journal of Clinical Endocrinology and Metabolism 77:341-346. Pasquali R, Casimirri F, Venturoli Set al (1983) Insulin resistance in patients with polycystic ovaries: its relationship to body weight and androgen levels. Acta Endocrinologica 104:110-116. Pasquali R, Casimiu'i F, B alestra V e t al (1991 ) The relative contribution of androgens and insulin in determining abdominal body fat distribution in premenopausal women. Journal of Endocrinological Investigation 14: 839-846.

246

J. HOLTZ

Pasquali R, Casimirri F, Ventnroli Set al (1994) Body fat distribution has weight-independent effects on clinical, hormonal, and metabolic features of women with polycystic ovary syndrome. Metabolism 43:706-713. Pasquali R, Casimin'i F, de lasio R et al (1995) Insulin regulates testosterone and sex hormone-binding globulin conceutratious in adult normal weight and obese men. Journal of Clinical Endocrinology and Metabolism 80: 654-658. Peiris AN, Stagner JI, Plymate SR et al (1993) Relationship of insulin secretory pulses to sex hormone-binding globulin in normal men [see comments]. Journal of Clinical Endocrinology and Metabolism 76: 279-282. Perel E & Killinger DW (1979) The interconversion and aromatisation of androgens by human adipose tissue. Journal of Steroid Biochemistry 11: 623-628. Plymate SR, Matej LA, Jones RE & Friedl KE (1988) Inhibition of sex hormone-binding globulin production in the human hepatoma (Hep G2) cell line by insulin and prolactin. Journal of Clinical Endocrinology and Metabolism 67: 460-464. Polderman KH, Gooren LJG, Asscheman H et al (1994) Induction of insulin resistance by androgens and estrogens. Journal of Clinical Endocrinology and Metabolism 79: 265-271. Poretsky L & Kalin MF (1987) The gonadotrophic ftmction of insulin. Endocrine Reviews 8: 132-145. Poretsky L & Piper B (1994) Insulin resistance, hypersecretion of LH, and a duel-defect hypothesis for the pathogenesis of polycystic ovary syndrome. Obstetrics and Gynecology 84: 613-621. Prelevic GM, Wurzburger MI, Balint PL & Ginsburg J (1992) Twenty-four-hour serum growth hormone, insulin, C-peptide and blood glucose profiles and serum insulin-like growth factor-I concentrations in women with polycystic ovaries. Hormone Research 37: 125-131. Rajkhowa M, Bicknell J, Jones M & Clayton RN (1994) Insulin sensitivity in women with polycystic ovary syndrome: relationship to hyperandrogenemia. Fertility and Sterility 61: 605-612. Reaven GM (1988) Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 37: 1595-1607. Reaven GM &Chen YD (1988) Role of abnormal free fatty acid metabolism in the development of non-insulin-dependent diabetes mellitus. American Journal of Medicine 85:106-112. Rebuffr-Scrive M, Lundholm K & Bjrrntorp P (1985) Glucocorticoid hormone binding to human adipose tissue. European Journal qf Clinical Investigation 15: 267-271. Rebuffr-Scrive M, Mfirin P & Bj6rntorp P (1991) Effect of testosterone on abdominal adipose tissue in men. International Journal qf Obesity 15: 791~795. Rebuffr-Scrive M, Cullberg G, Lundberg PA et al (1989) Anthropometric variables and metabolism in polycystic ovarian disease. Hormone Metabolic Research 21: 391-397. Rebuffr-Scrive M, Brrnnegard M, Nilsson A et al (1990) Steroid hormone receptors in human adipose tissues. Journal of Clinical Endocrinology and Metabolism 71: 1215-1219. Robinson S, Chan SP, Spacey S e t al (1992) Postprandial thermogenesis is reduced in polycystic ovary syndrome and is associated with increased insulin resistance [see comments]. Clinical Endocrinology 36: 537-543. Robinson S, Kiddy D, Gelding SV et al (1993) The relationship of insulin insensitivity to menstrual pattern in women with hyperandrogenism and polycystic ovaries. Clinical Endocrinology 39: 351-355. Rodin A, Thakkar H, Taylor N & Clayton R (1994) Hyperandrogenism in polycystic ovary syndrome. Evidence of dysregulation of l 1 beta-hydroxysteroid dehydrogenase. New England Journal of Medicine 330: 460-465. Rosenbaum D, Haber RS & Dunaif A (1993) Insulin resistance in polycystic ovary syndrome: decreased expression of GLUT-4 glucose transporters in adipocytes. American Journal of Physiology 264: EI97-E202. Rosenfeld RL, Barnes RB, Cara JF & Lucky AW (1990) Dysregulation of cytochrome P450c 17 alpha as the cause of polycystic ovarian syndrome. Fertility and Sterility 53: 785-791. Samojlik E, Kirschner MA, Silber D et al (1984) Elevated production and metabolic clearance rates of androgens in morbidly obese women. Journal of Endocrinology and Metabolism 49: 949-954. Samoto T, Maruo T, Matsuo H e t al (1993) Altered expression of insulin and insulin-like growth factor-1 receptors in follicular and stromal compartments of polycystic ovaries. Endocrinology Journal 40: 413-424. Sharp PS, Kiddy DS, Reed MJ et al (1991) Con'elation of plasma insulin and insulin-like growth factor-I with indices of androgen transport and metabolism in women with polycystic ovary syndrome. Clinical Endocrinology 35: 253-257.

INSULIN RESPONSES IN PCOS

247

Shoupe D & Lobo RA (1984) The influence of androgens on insulin resistance. Fertility and Sterility 41: 385-388. Shoupe D, Kumar DD & Lobo RA (1983) Insulin resistance in polycystic ovary syndrome. American Journal of Obstetrics and Gynecology 147: 588-592. Simon D, Preziosi P, Barrett CE et al (1992) Inten'elation between plasma testosterone and plasma insulin in healthy adult men: the Telecom Study. Diabetologia 35" 173-177. Singh A, Hamilton-Fairly D, Koistinen R et al (1990) Effect of insulin-like growth factor type 1 (IGF 1) and insulin on the secretion of sex hormone binding globulin and IGF 1 binding protein (IBP 1) by human hepatoma cells. Journal of Endocrinology 124: 1-3. Suikkari AM, Ruutiainen K, Erkkola R & Seppala M (1989) Low levels of low molecular weight insulin-like growth factor-binding protein in patients with polycystic ovarian disease. Human Reproduction 4:136-139. Swenne I (1992) Pancreatic beta-cell growth and diabetes mellitus. Diabetologia 35: 193-201. Tiitinen A, Tenbunen A & Seppala M (1993) Ovarian electrocauterization causes LH-regulated but not insulin-regulated endocrine changes. Clinical Endocrinology 39:181-184. Tiitinen AE, Laatikainen TJ 8,: Seppala MT (1993) Serum levels of insulin-like growth factor binding protein-1 and ovulatory responses to clomiphene citrate in women with polycystic ovarian disease. Fertility and Sterility 60: 58-62. Valdes CT & Elkind-Hirsch KE (1991) Intravenous glucose tolerance test-derived insulin sensitivity changes during the menstrual cycle. Journal of Clinical Endocrinology and Metabolism 72: 642-646. Warram JH, MaJ-tin BC, Krolewski AS et al (1990) Slow glucose removal rate and hyperinsulinemia precede the development of type II diabetes in the offspring of diabetic parents. Annals of Internal Medicine 113:909-915. Weber RF, Pache TD, Jacobs ML et al (1993) The relation between clinical manifestations of polycystic ovary syndrome and beta-cell function. Clinical Endocrinology 38: 295-300. Woodard TL, Burghen GA, Kitabchi AE & Wilimas JA (1981) Glucose intolerance and insulin resistance in aplastic anemia treated with oxymetholone. Journal of Clinical Endocrinology and Metabolism 55: 905-908. Zhang Y, Stern B & Rebar RW (1984) Endocrine comparison of obese menstruating and amenorrheic women. Journal of Clinical Endocrinology and Metabolism 58: 1077-1083.