Cardiovascular risk factors in young Czech females with polycystic ovary syndrome

J. Vrbíková1,5, R. Cífková2, A. Jirkovská3, V. Lánská4, H. Platilová1, V. Zamrazil1 and L. Stárka1

1 Department of Clinical Endocrinology, Institute of Endocrinology, Národní 8, 116 94 Prague 1, 2 Department of Preventive Cardiology, Institute for Clinical and Experimental Medicine, Vídenská 1958/9, 140 21 Prague 4, 3 Department of Diabetes, Institute for Clinical and Experimental Medicine, Vídenská 1958/9, 140 21 Prague 4 and 4 Statistics Unit, Institute for Clinical and Experimental Medicine, Vídenská 1958/9, 140 21 Prague 4, Czech Republic

5 To whom correspondence should be addressed at: Národní 8, 116 94 Prague 1, Czech Republic. e-mail: jvrbikova{at}endo.cz


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
BACKGROUND: Polycystic ovary syndrome (PCOS) could be associated with a variety of signs of metabolic syndrome. The aim of our study was to compare the cardiovascular risk factors in PCOS women and in a control group selected from a random population sample. METHODS AND RESULTS: 50 PCOS women with a mean (±SD) age of 30.7 ± 4.2 years, and 335 controls with a mean age of 29.9 ± 3.1 years selected from a random population sample of nine districts of the Czech Republic were compared for basic anthropometric characteristics, blood pressure, plasma lipids and fasting glucose. PCOS women had a significantly higher body mass index (BMI). After adjusting for BMI, PCOS women had higher blood pressure and LDL-cholesterol, and lower HDL and HDL-cholesterol/total ratio. Arterial hypertension was more prevalent in PCOS women than in controls. There was no difference in the prevalence of impaired fasting glucose between both groups. Impaired glucose tolerance was found in 11.8% of PCOS women. Diabetes mellitus was more frequent in PCOS families. CONCLUSIONS: Czech PCOS women, even in their thirties, show a significantly worse cardiovascular risk profile than a control group selected from a random population sample. The differences cannot be explained by obesity.

Key words: dyslipidaemia/hypertension/impaired glucose tolerance/type-2 diabetes mellitus


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
Polycystic ovary syndrome (PCOS) seems to be the most common endocrine disease in women of reproductive age, with the incidence reported to be about 4–6% in this age group (Knochenhauer et al., 1998Go). The association of PCOS with carbohydrate metabolism disorders was not investigated in more detail until the 1980s (Burghen et al., 1980Go). Insulin resistance (Dunaif, 1997Go) and impaired insulin secretion (Holte, 1998Go) are no doubt key components of PCOS. Apart from insulin resistance, PCOS women show other symptoms included in the so called metabolic ‘syndrome X’ (Ruderman et al., 1998Go; McFarlane et al., 2001Go) such as visceral obesity, arterial hypertension, increased levels of plasminogen activator inhibitor-1 (PAI-1), endothelial dysfunction, and mixed dyslipidaemia. These features are particularly striking in PCOS women younger than their forties and become blunted in the older age groups (Talbott et al., 1998Go; Cibula et al., 2000Go). On the other hand, a surprisingly low incidence of PCOS has been demonstrated in a population of women diagnosed as having syndrome X (Korhonen et al., 2001Go).

Recently, two studies conducted in North American PCOS patients reported an increased frequency of impaired glucose tolerance (IGT) and Type-2 diabetes mellitus (DM 2) (Ehrmann et al., 1999Go; Legro et al., 1999bGo). Similar results were obtained in South Asia (Weerakiet et al., 2001Go). Although coronary risk factors in hyperandrogenic states have recently been given more attention (Wild et al., 1985Go; Conway et al., 1992Go; Dahlgren et al., 1992aGo; Talbott et al., 1995Go; 1998; Atiomo et al., 1998Go; Wild et al., 2000Go; Diamanti-Kandarakis et al., 2001Go; Paradisi et al., 2001Go; Yarali et al., 2001Go), to the best of our knowledge, there has been only one paper addressing carbohydrate metabolism disorders and other risk factors for coronary heart disease in European women with PCOS compared with a healthy random population sample (Elting et al., 2001Go). The incidence of IGT and DM 2 has been clearly shown to be race-related, being highest in Hispanic and Afro-American populations. Similarly, there have been only few papers focusing on the prevalence of arterial hypertension in PCOS women, with inconclusive results (Zimmermann et al., 1992Go; Holte et al., 1996Go; Sampson et al., 1996Go).

The aim of our study was to compare the prevalence of cardiovascular risk factors in young Czech PCOS women and in a control group selected from a random population sample in the Czech Republic.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
Cases
A group of 50 women aged 25–42 years [mean (±SD) age 30.7 ± 4.2] meeting the criteria proposed by the 1990 NICHD PCOS Conference (Dunaif, 1997Go)—oligomenorrhoea or secondary amenorrhoea and hyperandrogenaemia—i.e. elevation of the free testosterone index, or testosterone, or androstenedione above the upper limit of the normal range [the normal range in our laboratory being 0.40–2.65 nmol/l for testosterone, 43–95 nmol/l for sex hormone-binding globulin, (SHBG) and 1.6–5.4 nmol/l for androstenedione] with its clinical manifestations (hirsutism and/or acne and/or androgenic alopecia) were included. None of the patients was diagnosed as having thyroid dysfunction, hyperprolactinaemia or a non-classical form of late-onset adrenal hyperplasia. The patients did not take oral contraceptives for at least 3 months prior to examination.

Biochemical assays
After a 12–14 h fast, 34 PCOS patients had the oral glucose tolerance test (oGTT) using 75 g of glucose, with blood collections for glucose determination using the glucose oxidase method (Beckmann, Fullerton, USA) at 0, 60 and 120 min; the evaluation was made using the revised World Health Organization (WHO) criteria (Alberti and Zimmet, 1998Go). A baseline venous blood sample was used to determine total cholesterol, HDL-cholesterol and triglycerides, all of which were assessed by photometry (Ecoline 25; Merck Vitalab Eclipse, Darmstadt, Germany). Hormonal analysis was undertaken between days 1–5 of the spontaneous menstrual cycle, with testosterone, androstenedione, dehydroepiandrosterone-sulphate (DHEAS) and SHBG determined as stated previously (Vrbikova et al., 2001Go). The mean (±SD) hormonal levels in our PCOS women were: testosterone 2.67 ± 1.12 nmol/l; androstenedione 7.1 ± 3.0 nmol/l; and SHBG 33.1 ± 14.4 nmol/l.

Controls
A control group of 335 women aged 25–34 [mean (±SD) age 29.9 ± 3.1] years was made up from a randomly selected (1%) population sample of permanent residents of nine districts of the Czech Republic in the 1997–98 period as part of a cross-sectional survey of cardiovascular risk factors (Cifkova et al., 2000Go). As there are no data on IGT in the Czech general population, we used results of oGTTs obtained during a worksite screening project for thyroid dysfunction and carbohydrate metabolism disorders undertaken, in the late 1980s, among Skoda car factory employees in Mladá Boleslav. The group of females examined comprised 54 individuals of similar age and body mass index (BMI) as the population random sample (age 34 ± 3 years; BMI 24.3 ± 3.8 kg/m2).

Biochemical assays
Blood glucose was determined by an enzymatic method (Lachema, Brno, Czech Republic). Total cholesterol and triglycerides were determined enzymatically (reagents from Boehringer Mannheim, Germany, using a Cobas Mira S autoanalyzer, Hoffman-La Roche, Basel, Switzerland); HDL-cholesterol was determined using the same method after precipitation.

Clinical investigations
All women were asked about the history of diabetes in their first- and second-degree relatives. BMI, as a measure of relative weight, was calculated as weight divided by height squared (kg/m2). Two blood pressure (BP) readings were obtained in sitting patients after a 10 min rest; the mean was determined from two values and was used for further analysis. BP values >=140/90 mmHg or a history of antihypertensive use was defined as hypertension.

The study was approved by the Local Ethics Committees of both institutions.

Computations and statistical methods
LDL-cholesterol was calculated using Friedewald’s formula (LDL-cholesterol = total cholesterol minus HDL-cholesterol minus triglycerides/2.19).

Statistical analysis of discrete variables was performed using the {chi}2-test. Continuous variables were analysed using Student’s two-sample t-test. A highly skewed variable (triglycerides) was analysed after logarithmic transformation. A two-tailed P < 0.05 was considered statistically significant. As cases and controls differed significantly in BMI, we adjusted BP values, lipid parameters and fasting blood glucose values for BMI and analysed them using analysis of co-variance (ANCOVA).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
When comparing PCOS women with controls, we found the former to have significantly higher BMI (P < 0.001), systolic (P < 0.01) and diastolic BP (P < 0.001) (Table I). Arterial hypertension was significantly more prevalent in PCOS patients compared with controls (22 versus 2.1%; P < 0.001). Total cholesterol, triglycerides and fasting blood glucose did not differ significantly between the two groups. HDL-cholesterol was significantly lower in PCOS patients compared with controls (P < 0.001) as was the HDL-/total cholesterol ratio (P < 0.001). LDL-cholesterol was significantly higher in PCOS than in controls (P < 0.001) (Table II). After adjusting the BP values, lipid parameters and fasting blood glucose for BMI, the significance for differences between PCOS and controls remained the same as for the unadjusted values.


View this table:
[in this window]
[in a new window]
 
Table I. Anthropometric parameters and BP in PCOS and control women. Values are given as mean ± SD
 

View this table:
[in this window]
[in a new window]
 
Table II. Lipids and fasting blood glucose in PCOS and control groups. Values are given as mean ± SD
 
The oGTT was performed in 34 PCOS women. One patient had a history of DM 2 and was treated with insulin. IGT was found in four (11.8 %) out of the 34 PCOS women and was thus more common than in a control group of Skoda car factory employees (0 versus 11.8%; P < 0.05). Impaired fasting glucose (IFG) was found in two (4%) out of the 50 PCOS patients, both of them with normal oGTT. On the other hand, IFG was diagnosed in nine individuals (3.2 versus 4%; NS) of the random population sample. Compared with controls, PCOS patients reported DM more often in their relatives (54.5 versus 19.7%; P < 0.001).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
Recent research has shown that PCOS is not only a gynaecological condition affecting women of reproductive age but, also, a comprehensive syndrome with a variety of associated metabolic disorders (insulin resistance, hyperinsulinaemia and dyslipidaemia) (Taylor, 1998Go; Amowitz and Sobel, 1999Go). In addition to a higher prevalence of DM 2 (Legro et al., 1999bGo), coronary heart disease (CHD) was suggested to be more common (Dahlgren et al., 1992aGo) in these women. Nevertheless, recent epidemiological studies did not confirm increased morbidity and mortality rates from CHD in PCOS patients (Pierpoint et al., 1998Go; Wild et al., 2000Go). Only an increased risk for cerebro-vascular events was demonstrated (Wild et al., 2000Go) in PCOS. These discrepancies have not been explained so far.

Coronary risk factors in hyperandrogenic states have been given more attention over the last decade (Wild et al., 1985Go; Conway et al., 1992Go; Dahlgren et al., 1992bGo; Talbott et al., 1995Go; 1998; Atiomo et al., 1998Go; Wild et al., 2000Go; Diamanti-Kandarakis et al., 2001Go; Paradisi et al., 2001Go; Yarali et al., 2001Go). A more adverse lipid profile (higher total and LDL-cholesterol) was found in young American PCOS women compared with the healthy population (Talbott et al., 1995Go). Our search of the relevant literature did not produce a single report of a study directly examining coronary risk factors (including glucose tolerance) in Caucasian PCOS patients compared with apparently healthy individuals selected from a random population sample. We took advantage of our control group selected from a 1% random sample of permanent residents of nine districts of the Czech Republic in the 1997–98 period as part of a study designed to survey cardiovascular risk factors (Cifkova et al., 2000Go). These nine districts cover 11.2% of the total area of the Czech Republic and are inhabited by 9.9% of the country’s population. As the respondent rate never fell below 60%, the sample is considered to be truly representative and the results allow generalization.

As there are no data on IGT in the Czech general population, we used data obtained during a worksite screening project for thyroid dysfunction and carbohydrate metabolism disorders undertaken, in the late 1980s, among employees in the Skoda car factory in Mladá Boleslav. IGT was not found in this particular group of female employees comprising 54 individuals (mean age 34 ± 3 years; mean BMI 24.3 ± 3.8 kg/m2). However, we cannot exclude the error of small numbers and, obviously, this was not a random population sample. In addition to that, the Skoda car factory female workers were mostly from rural areas and smaller communities whereas the controls included almost equal proportions of individuals from small communities and larger towns/cities. One of the nine districts of the Czech Republic involved in the screening for cardiovascular risk factors was the city of Pilsen (currently with a population of 160 000).

Our PCOS patients have an IGT prevalence of ~10%. This is rather low compared with USA or Asian PCOS patients with a prevalence of ~35% (Ehrmann et al., 1999Go; Legro et al., 1999bGo; Weerakiet et al., 2001Go). The higher prevalence of IGT in the USA and South Asia is probably the result of interaction of various ethnic backgrounds and environmental factors (especially obesity, as the average BMI was about 30–36 kg/m2 in these studies compared with 29.2 kg/m2 in Czech PCOS subjects).

DM 2 may develop as a clinically silent condition. It is not unusual to meet patients at the time of diagnosis of DM with advanced micro- and macrovascular complications, which not only impair the quality of life but also have a negative predictive value in these patients. These complications may already be present at the stage of IGT. There seems to be no threshold level of blood glucose; the association between the risk for cardiovascular disease and blood glucose seems to be present even across blood glucose values regarded as the normal range (Coutinho et al., 1999Go). The increasing prevalence of DM 2 in the population has recently led to attention focusing on its early diagnosis. To simplify screening for impaired glucose metabolism, the American Diabetic Association (ADA) (The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, 1997) proposed to determine fasting blood glucose instead of performing the oGTT. The ADA defined fasting blood glucose levels of 6.11–6.99 mmol/l as IFG. However, various authors have repeatedly criticized the introduction of the concept of IFG to replace performing the oGTT. Some reports pointed to the low concordance of these two diagnostic criteria (Burke et al., 1998Go). This is presumably due to the fact that, as IGT/DM develop, post-prandial blood glucose levels tend to rise (apparently as a result of inadequate insulin secretion following stimulation) earlier than fasting blood glucose (Gerich, 1998Go). In our series, no woman diagnosed as having IGT using the WHO criteria, was shown to have IFG according to the ADA classification. Fasting blood glucose within the IFG range was demonstrated in two patients (4%), but with normal results of their oGTTs.

In agreement with others (Wild et al., 1985Go; Conway et al., 1992Go; Pirwany et al., 2001Go) we found, in our PCOS patients, an adverse lipid profile characterized by low HDL-cholesterol, an increase in LDL-cholesterol and a decrease in the HDL/total cholesterol ratio. In some studies (Talbott et al., 1998Go), increased total cholesterol and triglycerides are also seen in PCOS women. These discrepancies could probably be due to the differences in hormonal status and in sample definition. LDL-cholesterol sub-fractions were not determined in the present study. The atherogenic LDL-cholesterol sub-fraction is believed to be small, dense LDL particles (LDL III). It has been demonstrated (Pirwany et al., 2001Go) that PCOS patients have higher levels of LDL III than healthy controls with comparable LDL-cholesterol levels; at the same time, the levels of plasma testosterone (unlike waist-to-hip ratio and insulinaemia) did not correlate with LDL III. This suggests a closer association of lipid metabolism disorders in PCOS patients with central obesity than with the levels of plasma androgens. Different results were reported by authors from the USA (Legro et al., 1999aGo) based on a Hispanic population, not demonstrating differences in the proportion of small, dense LDL particles between cases (PCOS) and controls (both of Hispanic origin).

In our study, hypertension was found more frequently in PCOS patients than in the control population. Similarly, we observed higher systolic and diastolic BP values in PCOS patients compared with controls. A population-based study conducted in the Netherlands (Elting et al., 2001Go), but using a telephone questionnaire, demonstrated hypertension in 9.6% of 346 respondents aged 17–56 years. Otherwise, data on the prevalence of hypertension in PCOS patients are inconsistent; 24 h BP monitoring showed both values comparable with those of the healthy population (Zimmermann et al., 1992Go), and higher systolic and diastolic BP levels (Holte et al., 1996Go).

PCOS women had a significantly higher BMI than the control group from the random population sample. Since blood pressure values, lipid levels and blood glucose are significantly associated with BMI, we reanalysed all our data after we had adjusted all these variables for BMI. Nevertheless, the results remained the same, suggesting that these abnormalities are closely related to PCOS status.

In conclusion, Czech PCOS patients present, already in their twenties to thirties, a cluster of atherosclerosis risk factors (obesity, adverse lipid profile, arterial hypertension and IGT). IGT/DM 2 seem to be less common than in other PCOS populations. The criteria developed by the ADA defining IFG are of little help in diagnosing carbohydrate metabolism disturbances in this patient population. Based on our results and literary data, we propose that all women, when diagnosed with PCOS, should have at least their total and HDL-cholesterol values determined; obese women should have the oGTT and, in addition, all these patients should have regular metabolic follow-up as a group at potential risk for early development of CHD.

Further investigations, preferably prospective studies, are needed to elucidate the exact effect of BMI and the incidence of cardiovascular events in PCOS in the next decades.


    Acknowledgments
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
This study was partly presented at the 46th Symposium of the German Society of Endocrinology, Goettingen, February 27–March 2, 2002. The study was supported by grants CEZ 00000023761, NB/6696–3 and CEZ: L 17/98.000 23 001 awarded by the Internal Grant Agency of the Ministry of Health of the Czech Republic. The study was not sponsored by any commercial company in any way.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgments
 References
 
Alberti, K.G. and Zimmet, P.Z. (1998) Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet. Med., 15, 539–553.

Amowitz, L.L. and Sobel, B.E. (1999) Cardiovascular consequences of polycystic ovary syndrome. Endocrinol. Metab. Clin. North Am., 28, 439–458.

Atiomo, W.U., Bates, S.A., Condon, J.E., Shaw, S., West, J.H. and Prentice, A.G. (1998) The plasminogen activator system in women with polycystic ovary syndrome. Fertil. Steril., 69, 236–241.

Burghen, G.A., Givens, J.R. and Kitabchi, A.E. (1980) Correlation of hyperandrogenism with hyperinsulinism in polycystic ovarian disease. J. Clin. Endocrinol. Metab., 50, 113–116.

Burke, J.P., Haffner, S.M., Gaskill, S.P., Williams, K.L. and Stern, M.P. (1998) Reversion from type 2 diabetes to nondiabetic status. Influence of the 1997 American Diabetes Association criteria. Diabetes Care, 21, 1266–1270.

Cibula, D., Cifkova, R., Fanta, M., Poledne, R., Zivny, J. and Skibova, J. (2000) Increased risk of non-insulin dependent diabetes mellitus, arterial hypertension and coronary artery disease in perimenopausal women with a history of the polycystic ovary syndrome. Hum. Reprod., 15, 785–789.

Cifkova, R., Skodova, Z., Lanska, V., Poledne, R., Adamkova, V., Novozamska, E., Jozifova, M., Petrzilkova, Z., Hauserova, G., Duskova, A. et al. (2000) [Longitudinal trends of total and HDL cholesterol in a representative population sample in the Czech Republic]. Vnitr. Lek., 46, 501–505.

Conway, G.S., Agrawal, R., Betteridge, D.J. and Jacobs, H.S. (1992) Risk factors for coronary artery disease in lean and obese women with the polycystic ovary syndrome. Clin. Endocrinol. (Oxf), 37, 119–125.

Coutinho, M., Gerstein, H.C., Wang, Y. and Yusuf, S. (1999) The relationship between glucose and incident cardiovascular events. A metaregression analysis of published data from 20 studies of 95783 individuals followed for 12.4 years. Diabetes Care, 22, 233–240.

Dahlgren, E., Janson, P.O., Johansson, S., Lapidus, L. and Oden, A. (1992a) Polycystic ovary syndrome and risk for myocardial infarction. Evaluated from a risk factor model based on a prospective population study of women. Acta Obstet. Gynecol. Scand., 71, 599–604.

Dahlgren, E., Johansson, S., Lindstedt, G., Knutsson, F., Oden, A., Janson, P.O., Mattson, L.A., Crona, N. and Lundberg, P.A. (1992b) Women with polycystic ovary syndrome wedge resected in 1956 to 1965: a long-term follow-up focusing on natural history and circulating hormones. Fertil. Steril., 57, 505–513.

Diamanti-Kandarakis, E., Spina, G., Kouli, C. and Migdalis, I. (2001) Increased endothelin-1 levels in women with polycystic ovary syndrome and the beneficial effect of metformin therapy. J. Clin. Endocrinol. Metab., 86, 4666–4673.

Dunaif, A. (1997) Insulin resistance and the polycystic ovary syndrome: mechanism and implications for pathogenesis. Endocr. Rev., 18, 774–800.

Ehrmann, D.A., Barnes, R.B., Rosenfield, R.L., Cavaghan, M.K. and Imperial, J. (1999) Prevalence of impaired glucose tolerance and diabetes in women with polycystic ovary syndrome. Diabetes Care, 22, 141–146.

Elting, M.W., Korsen, T.J., Bezemer, P.D. and Schoemaker, J. (2001) Prevalence of diabetes mellitus, hypertension and cardiac complaints in a follow-up study of a Dutch PCOS population. Hum. Reprod., 16, 556–560.

Gerich, J.E. (1998) The genetic basis of type 2 diabetes mellitus: impaired insulin secretion versus impaired insulin sensitivity. Endocr. Rev., 19, 491–503.

Holte, J. (1998) Polycystic ovary syndrome and insulin resistance: thrifty genes struggling with over-feeding and sedentary life style? J. Endocrinol. Invest., 21, 589–601.

Holte, J., Gennarelli, G., Berne, C., Bergh, T. and Lithell, H. (1996) Elevated ambulatory day-time blood pressure in women with polycystic ovary syndrome: a sign of a pre-hypertensive state? Hum. Reprod., 11, 23–28.

Knochenhauer, E.S., Key, T.J., Kahsar-Miller, M., Waggoner, W., Boots, L.R. and Azziz, R. (1998) Prevalence of the polycystic ovary syndrome in unselected black and white women of the southeastern United States: a prospective study. J. Clin. Endocrinol. Metab., 83, 3078–3082.

Korhonen, S., Hippelainen, M., Niskanen, L., Vanhala, M. and Saarikoski, S. (2001) Relationship of the metabolic syndrome and obesity to polycystic ovary syndrome: a controlled, population-based study. Am. J. Obstet. Gynecol., 184, 289–296.

Legro, R.S., Blanche, P., Krauss, R.M. and Lobo, R.A. (1999a) 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., 72, 990–995.

Legro, R.S., Kunselman, A.R., Dodson, W.C. and Dunaif, A. (1999b) 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 [see comments]. J. Clin. Endocrinol. Metab., 84, 165–169.

McFarlane, S.I., Banerji, M. and Sowers, J.R. (2001) Insulin resistance and cardiovascular disease. J. Clin. Endocrinol. Metab., 86, 713–718.

Paradisi, G., Steinberg, H.O., Hempfling, A., Cronin, J., Hook, G., Shepard, M.K. and Baron, A.D. (2001) Polycystic ovary syndrome is associated with endothelial dysfunction. Circulation, 103, 1410–1415.

Pierpoint, T., McKeigue, P.M., Isaacs, A.J., Wild, S.H. and Jacobs, H.S. (1998) Mortality of women with polycystic ovary syndrome at long-term follow- up. J. Clin. Epidemiol., 51, 581–586.

Pirwany, I.R., Fleming, R., Greer, I.A., Packard, C.J. and Sattar, N. (2001) Lipids and lipoprotein subfractions in women with PCOS: relationship to metabolic and endocrine parameters. Clin. Endocrinol. (Oxf), 54, 447–453.

Ruderman, N., Chisholm, D., Pi-Sunyer, X. and Schneider, S. (1998) The metabolically obese, normal-weight individual revisited. Diabetes, 47, 699–713.

Sampson, M., Kong, C., Patel, A., Unwin, R. and Jacobs, H.S. (1996) Ambulatory blood pressure profiles and plasminogen activator inhibitor (PAI-1) activity in lean women with and without the polycystic ovary syndrome. Clin. Endocrinol. (Oxf), 45, 623–629.

Talbott, E., Guzick, D., Clerici, A., Berga, S., Detre, K., Weimer, K. and Kuller, L. (1995) Coronary heart disease risk factors in women with polycystic ovary syndrome. Arterioscler. Thromb. Vasc. Biol., 15, 821–826.

Talbott, E., Clerici, A., Berga, S.L., Kuller, L., Guzick, D., Detre, K., Daniels, T. and Engberg, R.A. (1998) Adverse lipid and coronary heart disease risk profiles in young women with polycystic ovary syndrome: results of a case-control study. J. Clin. Epidemiol., 51, 415–422.

Taylor, A.E. (1998) Understanding the underlying metabolic abnormalities of polycystic ovary syndrome and their implications. Am. J. Obstet. Gynecol., 179, S94–S100.

The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus (1997) Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care, 20, 1183–1197.

Vrbikova, J., Hill, M., Starka, L., Cibula, D., Bendlova, B., Vondra, K., Sulcova, J. and Snajderova, M. (2001) The effects of long-term metformin treatment on adrenal and ovarian steroidogenesis in women with polycystic ovary syndrome. Eur. J. Endocrinol., 144, 619–628.

Weerakiet, S., Srisombut, C., Bunnag, P., Sangtong, S., Chuangsoongnoen, N. and Rojanasakul, A. (2001) Prevalence of type 2 diabetes mellitus and impaired glucose tolerance in Asian women with polycystic ovary syndrome. Int. J. Gynaecol. Obstet., 75, 177–184.

Wild, R.A., Painter, P.C., Coulson, P.B., Carruth, K.B. and Ranney, G.B. (1985) Lipoprotein lipid concentrations and cardiovascular risk in women with polycystic ovary syndrome. J. Clin. Endocrinol. Metab., 61, 946–951.

Wild, S., Pierpoint, T., McKeigue, P. and Jacobs, H. (2000) Cardiovascular disease in women with polycystic ovary syndrome at long- term follow-up: a retrospective cohort study. Clin. Endocrinol. (Oxf), 52, 595–600.

Yarali, H., Yildirir, A., Aybar, F., Kabakci, G., Bukulmez, O., Akgul, E. and Oto, A. (2001) Diastolic dysfunction and increased serum homocysteine concentrations may contribute to increased cardiovascular risk in patients with polycystic ovary syndrome. Fertil. Steril., 76, 511–516.

Zimmermann, S., Phillips, R.A., Dunaif, A., Finegood, D.T., Wilkenfeld, C., Ardeljan, M., Gorlin, R. and Krakoff, L.R. (1992) Polycystic ovary syndrome: lack of hypertension despite profound insulin resistance. J. Clin. Endocrinol. Metab., 75, 508–513.

Submitted on May 7, 2002; resubmitted on December 5, 2002; accepted on January 27, 2003.