Polycystic ovarian syndrome (PCOS) is one of the commonest endocrinopathies of women and is one of the most widely studied and controversial areas in gynecologic endocrinology. Insulin resistance is a well-recognized feature of PCOS and, in association with hypertension and dyslipidemia, may increase the risk of cardiovascular and cerebrovascular events.1 These risk factors are compounded by central obesity, which is present in the majority of women with PCOS. Despite the adverse vascular risk profile observed in PCOS patients, evidence from long-term outcome studies have failed to significantly demonstrate an increased cardiovascular and cerebrovascular mortality in comparison with a healthy control population. In the absence of adequate outcome studies, surrogate markers (ie, increased carotid intima thickness, reduced brachial artery flow–mediated vasodilatation, increased homocysteine and leukocyte circulating levels) have been evaluated to determine whether women with PCOS have evidence of subclinical vascular disease compared with controls.2 The data show conflicting results. Very few studies have been published on the vascular risk in women with solely ultrasonographic evidence of polycystic ovaries (PCO).3–5 This ovarian morphology has been documented in childhood and adolescence and in menopausal women and may be seen in patients with a history of congenital adrenal hyperplasia or idiopathic hirsutism. Polycystic ovarian syndrome is present in 16–25% of apparently healthy women with regular cycles. Although there are only a few studies on the endocrine profile of PCO patients, some authors have suggested that PCO may be associated with some of the features present in PCOS patients: abnormal gonadotropin levels, increased insulin resistance, increased ovarian 17-hydroxyprogesterone, and androgen response to GnRH-agonists.3–5
The aim of this study was to estimate whether young PCOS patients have subclinical indicators that express an increased risk of vascular disease in comparison with normally menstruating patients with ultrasonographic evidence of PCO and healthy controls, both matched for age and body mass index (BMI).
PARTICIPANTS AND METHODS
Twenty-eight adult (older than 18 years) white women with PCOS (hirsutism, oligomenorrhea, infertility, increased plasma circulating androgens, luteinizing hormone (LH)/follicle-stimulating hormone (FSH) ratio of 2.0 or more, and typical bilateral ultrasonographic and color Doppler findings)6 were considered eligible for the study after secondary causes of hyperandrogenism were excluded. Amenorrheic PCOS patients were not included in the study. Seventeen eumenorrheic white women with bilateral PCO (more than 10 small-sized 2–10 mm subcapsular follicles, ovarian volume more than 8 mL, and increased ovarian echogenicity) and 15 healthy white volunteers with regular ovulatory cycles, matched for age (18–30 years) and BMI (20–30 kg/m2), were also recruited for the study. Patients with ultrasonographic evidence of multifollicular ovaries were not included in the study.6 All participants were nonsmokers, did not take regular intense exercise, and did not receive hormonal therapy for at least 6 months before the study. Women with diabetes, renal or hepatic illness, and with folic acid and vitamin B12 deficiencies were excluded from the study. Further exclusion criteria were uterine malformations, endometriosis, ovarian functional cyst, unilateral ovarian resection, or ovariectomy. Pregnancy tests were negative in all patients before the enrollment in the study. All participants were instructed to avoid any intense exercise and to refrain from intake of caffeine-containing beverages in the 24 hours antecedent the study. The study protocol was in accordance with the Helsinki II declaration and was approved by the Hospital Research Review Committee. The study was conducted between January and December 2005 in the Department of Obstetrics and Gynecology of Bologna University. Women participated in the study after giving informed written consent.
After the first the screening evaluation, participants were assessed in the early follicular phase (cycle days 3–5) after an overnight fast with a detailed history and medical examination. Standing height (cm) and weight (kg) were measured. The mean BMI was calculated. Waist and hip circumferences were measured at the level of umbilicus and greater trochanter. The waist/hip ratio was calculated. Hirsutism was assessed by one examiner (A.C.) using the modified Ferriman-Gallwey score in which a score more than 8 indicates hirsutism.7 The examiner was blinded to the participant's group assignment. Fasting blood samples were drawn for testing biochemical and hormonal characteristics. Plasma concentrations of nitrites/nitrates (NO2–/NO3–) were also assayed. Patients were further submitted to uteroovarian ultrasonographic analysis and to color Doppler evaluation of uterine, stromal ovarian, and ophthalmic arteries. In addition, ultrasonographic and color Doppler analysis of brachial artery flow–mediated vasodilatation and 24-hour ambulatory blood pressure monitoring were performed. On the subsequent day, an oral glucose tolerance test (OGTT) was performed and blood was collected for the analysis of glucose, insulin and C-peptide.
Ultrasound (US) examination of the ovaries was performed with the use of a multifrequency transvaginal transducer (Aspen, Acuson-Siemens, Milan, Italy). Uterine volume, ovarian volume, number, diameter and distribution of the follicles were recorded.8 Echogenicity of ovarian stroma was subjectively scored as 0 (normal), 1 (moderately increased), or 2 (markedly increased).8 The ovarian volume, the ovarian stroma echogenicity, and follicular number showed no significant differences among the ovaries, and, therefore, the average value of both ovaries was used for statistical analysis. Doppler flow measurements of the uterine and intraovarian vessels were performed transvaginally (Aspen color Doppler).6,8 The color Doppler analysis of ophthalmic arteries was performed using a 7–10 MHz linear array (Aspen color Doppler).9 To ensure standardized conditions, the patients rested in a waiting room for at least 15 minutes before being scanned and then were evaluated in a noiseless laboratory with constant heat and light.9 The analysis of the ovarian stroma and uterine and ophthalmic arteries were performed as previously described.6,8,9 The pulsatility index was electronically calculated by the machine. For each examination, the mean value of three consecutive waveforms was obtained. No significant differences between the pulsatility indexes of the left and right side for uterine, ovarian stromal, and ophthalmic arteries were observed, and therefore, the average value of both side arteries was used. For the ophthalmic arteries the peak systolic blood flow velocity (Vmax) was calculated.9 In view of the difficulty in interpreting the pulsatility index in low-impedance vascular beds such as the cerebral circulation (ophthalmic artery), the downstream “back pressure” was calculated.10 Ultrasound and color Doppler analyses were performed by a single examiner (C.B.) blinded to the participant's group assignment.
Immediately after the above-described US and Doppler scanning, an ultrasound transducer was placed over the brachial artery to measure its diameter before and after reactive hyperemia. Briefly, the right brachial artery was evaluated with continuous scanning held for 30 seconds (by using a 7–10 MHz multifrequency linear array transducer [Aspen]) over a longitudinal section 5–7 cm above the right elbow. A blood pressure cuff around the upper arm was then inflated to a pressure of 200 mm Hg for 5 minutes, causing ischemia. The subsequent cuff deflation induced a brief high-flow condition through the brachial artery (reactive hyperemia) due to the intense nitric oxide release from the endothelial cells. After this, a second scan was performed and the brachial artery analyzed at 15 seconds, 60 seconds and 120 seconds. The ultrasound images were recorded on videocassette and subsequently analyzed off line. The diameter of the brachial artery was measured from the tunica intima at a fixed distance from the chosen marker. The mean diameters of the brachial artery before and after reactive hyperemia were calculated from three cardiac cycles synchronized with R-wave peaks on the electrocardiogram. Flow-mediated vasodilatation was also determined as the percentage change from baseline to 15 seconds and 120 seconds after sudden deflation subsequent to arm ischemia. A Doppler analysis of the brachial artery was performed and the pulsatility indexes registered (as absolute values and percentage variations) at baseline and just after US measurements of brachial artery. Brachial artery flow mediated vasodilatation was evaluated by a single examiner (C.B.).
Ambulatory blood pressure monitoring was performed using a portable lightweight device (SpaceLab 90121, Critikon, Tampa, FL) applied to the nondominant arm. Patients wore the device for 24 hours, with measurements every 30 minutes during the day (8:00 am to 10:00 pm) and hourly overnight (10:00 pm to 8:00 am). Patients were asked to sleep between 10:00 pm and 8:00 am and to maintain their usual daily activities. The 24-hour blood pressure monitoring was considered statistically acceptable in presence of more than 75% successful measurements. Different 24-hour, daytime, nighttime, and wake-up variables were calculated: systolic blood pressure (BP), diastolic BP, mean arterial pressure (MAP=[systolic BP+(2/3 diastolic BP)/3]), and heart rate. The mean of the above measures were separately calculated in the 4 hours before and after waking. Normal blood pressure values were considered: less than 135/85 mm Hg for daytime, less than 120/75 mm Hg for nighttime, and less than 130/80 mm Hg for 24 hours. The percentage of recordings exceeding these reference values was calculated. The results were separately analyzed by two researchers (D.deA. and F.M.).
Peripheral blood was obtained from all patients, between 8:00 am and 11:00 am, on the same day that US and Doppler examinations took place and different hormonal and biochemical parameters were analyzed. Plasma concentration of LH, FSH, estradiol (E2), testosterone, androstenedione, and 17-hydroxy-progesterone were assayed as previously described.8,11 The LH/FSH ratio was calculated. Sex hormone binding globulin was immunoassayed using Immulite 1000 (EURO/Diagnostic Products Corp. Ltd., Los Angeles, CA). The free androgen index (free androgen index=testosterone (nmol/L)/sex hormone binding globulin (nmol/L)×100) was calculated. Results of hormonal values were converted to International System (SI) units: LH (International Units/L)=milli-International Units/mL×1.0; FSH (International Units/L)=milli-International Units/mL×1.0; E2 (pmol/L)=pg/mL×3.761; testosterone (nmol/L)=ng/mL×3.467; androstenedione (nmol/L)=ng/dL×0.0349; 17-hydroxy-progesterone (nmol/L)=ng/dL×0.03026. Serum total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides were assessed enzymatically by an autoanalyzer (COBAS-MIRA, Roche Diagnostic Limited, Lewes, UK). Low-density lipoprotein (LDL) cholesterol was estimated as described by the Friedewald equation.12 The atherogenic index of plasma was computed.13 Results of circulating lipids were converted to SI units: total cholesterol, HDL cholesterol and LDL cholesterol (mmol/L)=mg/dL×0.0259; triglycerides (mmol/L)=mg/dL×0.0113. Leukocyte count (n×103) was determined within 2 hours after venipuncture with an automatic work station cell counter (Technicon H3, Bayer Diagnostic, Munich, Germany). An aliquot of peripheral blood was immediately centrifuged and serum stored at –70°C until assays. Nitric oxide (NO) production was assessed by monitoring serum levels of stable oxidation products of NO metabolism (nitrites/nitrates [NO2–/NO3–]). The NO2–/NO3– were assayed at Modena University (F.F.) with the Greiss reaction with procedures previously described.14 In addition, plasma homocysteine concentrations were determined by two researchers (R.M. and G.R.M.) with a method based on fluorescence polarization immunoassay using an Abbott Imx (Abbott Laboratories, Abbott Park, IL) analyzer.
On the subsequent day, after a further overnight fast, an oral glucose tolerance test (75 g Curvosio, Sclavo, Cinisello Balsamo, Italy) was performed and blood was collected for the analysis of glucose, insulin and C-peptide at 15 minutes before, and 30, 60, 90, and 120 minutes after the oral ingestion of glucose, as suggested by the World Health Organization (WHO) criteria.15 Plasma glucose levels were determined by the glucose-oxidase method on a Beckman glucose analyzer (Fullerton, CA), plasma insulin was determined using an immunoradiometric assay (Biosurce Europa SA, Nivelles, Belgium), and plasma C-peptide concentrations were determined using a chemiluminescent assay (DBC Immulite One, Los Angeles, CA). Results, when necessary, were converted to SI units: Glucose (mmol/L)=mg/dL×0.0555; Insulin (pmol/L)=micro-International Units/mL×6.945; C-peptide (nmol/L)=ng/mL×0.333. The definition for normal fasting glucose, impaired fasting glucose, and diabetes were based on the established American Diabetes Association criteria.16 Glucose tolerance was assessed by WHO criteria.15 Glucose, insulin, and C-peptide determinations during the OGTT were used to calculate the respective areas under the curve (AUC120) at 120 minutes according to the formula: 15x[(V30+V60+V90)x2+V–15+V120], where V is the glucose, insulin, or C-peptide concentration at the indicated time. The partial (0–90 minute and 90-120 minute) AUC0–90 and AUC90–120 were calculated. The homeostatic model assessment estimates for insulin resistance, quantitative insulin sensitivity check index, insulin sensitivity index and fasting glucose/insulin ratio were derived as estimates of insulin sensitivity. In addition to the fasting C-peptide and insulin levels, the insulinogenic index and the homeostatic model assessment estimates for percent pancreatic β-cell function were derived as indices of pancreatic β-cell function. For the same purposes, β-cell secretion of insulin was estimated by the following indices: predicted indexes of first and second phase of insulin secretion. The fasting C-peptide/insulin molar ratio was considered an index of hepatic insulin clearance.
Statistical analysis (SPSS 11.5 software, SPSS Inc., Chicago IL) was performed using the nonparametric Kruskal-Wallis test supplemented by a nonparametric comparison procedure (Dunn's test). Repeated measures for analysis of variance were used when indicated. The relationship between the measures analyzed was assessed using a nonparametric correlation test (Spearman's test). A P value .05 was considered as statistically significant. Data are presented as mean ± standard deviation, unless otherwise indicated. The statistical analysis was performed by a single researcher (P.B.).
All 60 women completed the study. The three groups of studied patients, on the basis of inclusion criteria, did not significantly differ in age or BMI (Table 1). In addition, the age at menarche and the weight/hip ratio was similar in all patients (Table 1). The only two patients with a history of premature isolated pubarche were observed in PCOS group. Similarly, a familial history of maternal PCOS and/or paternal premature baldness was observed in three patients belonging to PCOS and one patient belonging to PCO group. The Ferriman-Gallwey score was as indicated in Table 1. The plasma levels of LH, FSH, E2, testosterone, androstenedione, and 17-hydroxy-progesterone are reported in Table 1. The LH/FSH ratio, the sex hormone binding globulin values and the free androgen index resulted as reported in Table 1. The US assessment allowed the measurement of the uterine volume and endometrial thickness in 100% of the cases, and no significant differences were registered among the groups (Table 2). In all 60 patients, both ovaries were always visualized. The mean ovarian volume was significantly higher in PCOS patients than in the other groups, and the PCO women showed an ovarian volume greater than controls (Table 2). The mean number of small subcapsular follicles was significantly lower in controls than in PCO and PCOS patients (Table 2). No significant differences were observed between PCOS and PCO patients (Table 2). The stromal score=2 was evident in 100% of PCOS and PCO patients, and in none of the control women (P<.001). In this last group the stromal score=0 was observed in 15/15 patients. Doppler analysis revealed a significantly higher uterine pulsatility index in PCOS group (3.85±0.75) compared with controls (2.14±0.32; P<.001). No significant differences were observed between PCOS and PCO (2.83±0.80) patients. However, in the PCO women the values were significantly higher than in controls (Table 2). At the level of the ovaries, a stromal vascularization was observed in all the patients. The lowest resistances were found in PCOS and PCO patients in comparison with controls (Table 2). The mean pulsatility index and Vmax of ophthalmic arteries was similar in all the studied patients (Table 2). However, the ophthalmic artery back pressure was significantly higher in PCOS group (64.4±6.5 mm Hg) than in controls (54.1±5.1 mm Hg; P=.028) (Table 2). The PCO patients showed values (59.6±4.8 mm Hg) not significantly different from PCOS patients and controls (Table 2). The brachial artery diameter, at baseline, was similar in all the patients (Fig. 1). After the reactive hyperemia, a more intense vasodilatation was observed in controls and PCO patients in comparison with PCOS women (Fig. 1). The persistence of the effect was more evident in controls than in PCOS and PCO patients (Fig. 1). The percentage change at 15 seconds was significantly lower in PCOS patients than in PCO and control women (Fig. 1). At 120 seconds the reactive hyperemia was still kept only in controls (Fig. 1). At baseline, the pulsatility index of the brachial artery was significantly higher in PCOS than PCO and control group (Fig. 2). The Doppler variations at level of the brachial artery were more evident and persistent in controls than in PCOS patients (Fig. 2). The women with PCO presented values as reported in Figure 2. The 24-hour blood pressure monitoring, among the groups, showed no significant differences in the 24 hours, daytime and nighttime values. However, during the sleep time, the percentages of values more than 120 mm Hg in the systolic BP were significantly higher in PCOS (15±7%; P=.034) and PCO (14±8%; P=.046) patients than in controls (5±4%). Similarly, the percentages of values more than 75 mm Hg in diastolic BP were significantly higher in PCOS (15±8%; P=.048) and PCO patients (17±9.5%; P=.037) than in controls (6±4). All the groups resulted normally dippers (nocturnal reduction 10% or more in comparison with diurnal values). The wake-up blood pressure values were similar in all the groups and systolic and diastolic BP, MAP, and heart rate in the 4 hours before waking and 4 hours after waking were not different among the studied patients. Total cholesterol, triglycerides, and the atherogenic index of plasma were significantly higher in PCOS than PCO patients and controls (Table 3); HDL and LDL cholesterol were not significantly different among the three groups (Table 3). Leukocytes were slightly but significantly increased in PCOS (7,273±789) in comparison with PCO (6,230±1,183; P=.046) patients and controls (6,182±1,004; P=.045). Among other biochemical measures, homocysteine was slightly higher in PCOS than in PCO patients and controls (Table 3). The NO2–/NO3– plasma levels were significantly reduced in PCOS and PCO women in comparison with controls (Table 3). Fasting values of glucose, insulin, and C-peptide are reported in Table 4. The insulin and C-peptide plasma values were higher in PCOS patients than controls (Table 4). All the studied women presented normal fasting glucose values on the basis of the American Diabetes Association criteria. The glucose, insulin, and C-peptide plasma values after the OGTT are reported in Figure 3. On the basis of the WHO criteria, all the patients presented a normal glucose tolerance. The area under the curve (AUC120) for glucose, insulin, and C-peptide are shown in Table 4; the PCOS patients presented the worst results. Furthermore, these patients presented higher values also in the later partial (AUC90–120) glucose, insulin, and C-peptide areas under the curve (Table 4). The different estimates (homeostatic model assessment estimates for insulin resistance, quantitative insulin sensitivity check index, insulin sensitivity index, and glucose/insulin ratio) of insulin sensitivity and indices of pancreatic β-cell function (percent pancreatic β-cell function, insulinogenic index, first and second phase of insulin secretion) are reported in Table 4. In the same table are shown the values of the fasting C-peptide/insulin molar ratio (as an index of hepatic insulin clearance); the PCOS patients presented significantly higher values in comparison with PCO women and controls.
The relationships among the different measures were analyzed using Spearman's nonparametric correlation. We fundamentally observed that 1) the age was inversely correlated with homocysteine (ρ=–.550; P<.001) and positively correlated with total cholesterol (ρ=.320; P=.016) and triglycerides (ρ=.336; P=.011); 2) free androgen index was positively correlated with total cholesterol (ρ=.490; P<.001), triglycerides (ρ=.357; P=.008), and atherogenic index of plasma (ρ=.348; P=.013); 3) brachial artery hyperemia was inversely correlated with BMI (ρ=–.297; P=.032) and free androgen index (ρ=–.323; P=.019), and positively correlated with NO2–/NO3– plasma levels (ρ=.322; P=.018); 4) homocysteine was negatively correlated with age (ρ=–.530; P<.001). The majority of carbohydrate metabolic measures were correlated with BMI and free androgen index (Table 5).
Although still equivocal and debatable, and in the absence of adequate long-term outcome studies, there are mounting suggestions that PCOS women may have an increased risk of cardiovascular and cerebrovascular pathologies in comparison with healthy cycling women of similar age and BMI. In the present study, as expected, the PCOS women presented significantly higher LH, LH/FSH ratio, and androgens circulating values both in comparison with controls and PCO patients. This was also associated with higher free androgen index and Ferriman-Gallwey score. In addition to verifying the important role of ultrasonography, we confirmed that Doppler analysis of the uterine and ovarian stromal vascularization may be an additional valuable tool for the diagnosis of polycystic ovaries.6,8,11 It was observed that the ovarian vascular resistances are decreased, whereas uterine vascular resistances are increased in women with complete expression of the polycystic ovarian syndrome.8 Among the three groups of patients, we also analyzed the hemodynamic properties of ophthalmic arteries (small vessels arising from the internal carotid artery). Cerebral vessels are morphologically and physiologically similar to the arteries of the eye. Thus, knowledge of vascular changes in the ophthalmic arteries may be useful in assessing changes in global cerebral perfusion. Powers and coworkers17 showed that decreased/reversed ophthalmic artery flow is often associated with diminished cerebral perfusion. Our study showed that the ophthalmic artery back pressure was significantly higher in PCOS patients than controls. We speculated that the increased arterial stiffness, the atherogenic lipid profile and the lower circulating NO2–/ NO3– values, as subsequently reported, may be responsible of the increased back pressure in PCOS. In addition, PCOS is widely considered related with type II diabetes, in which regional cerebral blood flow is reduced.18 These data may in part confirm the increased risk of cerebrovascular pathologies in PCOS patients.
The vascular endothelium is a complex organ with a multitude of autocrine, paracrine, and endocrine properties. Abnormalities in endothelial function may precede overt vascular disease by years, have been associated with several well-defined cardiovascular risk factors and may portend clinically significant vascular diseases. In our study we observed a more intense postischemic vasodilatation in controls and PCO patients in comparison with PCOS women. However, the reactive hyperemia after 120 seconds was still kept only in controls. Similar results were observed performing the Doppler analysis of brachial artery PIs. The nitrites/nitrates plasma levels were significantly reduced in PCO and PCOS patients with respect to the controls. Postischemic brachial artery hyperemia was inversely correlated with BMI and free androgen index. In addition, a significant positive relationship was observed between the postischemic brachial artery hyperemia and nitrites/nitrates circulating plasma levels. Thus, we derived that weight and androgens may, in a cumulative way, negatively influence brachial artery flow-mediated vasodilatation and that PCOS is associated with a reduction of nitric oxide release/production or to an increased nitric oxide degradation and to an impaired endothelium-dependent vasodilatation. Therefore, endothelial dysfunction in PCOS behaves as a marker for patients with preclinical vascular disease and may identify, at an early age, patients in whom therapeutic intervention could be beneficial.
It is controversial whether PCOS per se is associated with hypertension. In our study, using 24-hour ambulatory monitoring, we did not demonstrate any difference between PCOS and PCO women and controls in total daytime or nighttime blood pressure and heart rate. In addition, the nocturnal dipping, the wake-up surge, the 4-hour periods before and after waking (all prognostic indicators of target-organ damage and of cardiovascular morbidity and mortality in normotensive and hypertensive population) were similar among the studied patients. However, the percentages of systolic BP more than 120 mm Hg and diastolic BP more than 75 mm Hg during the nocturnal period were significantly higher in PCOS and PCO patients in comparison with controls. This rather labile control of the blood pressure during the nighttime might indicate a very precocious prehypertensive state. With advancing age, apart from some specific factors due to genetics, inactivity, obesity, stress, and salt loading, it is plausible that blood pressure may increase at an accelerated rate in PCOS and, probably, PCO patients, given the stimulatory effects of hyperinsulinemia on the sympathetic nervous system and vascular smooth muscle and the changes we have noted in endothelial function.
In this study we found an increased leukocyte count in PCOS in comparison with normal-menstruating PCO patients and controls. Inflammation has been recognized to have a pivotal role in both initiation and progression of the atherosclerotic processes. Indeed, an increased white blood cell count is directly associated with an increased incidence of myocardial infarction and ischemic stroke. In disagreement with previous articles, our study did not show any association between leukocyte count and insulin resistance. This was probably due to the young age of the studied patients considering that the inflammation/atherosclerosis evolution is a long-lasting and slow process.
Hyperhomocysteine levels are also considered to be an independent risk factor for cardiovascular diseases.19 In our study more elevated homocysteine circulating levels were observed in PCOS women. No correlations were found between homocysteine, insulin, and BMI. As a result of these findings, we can suggest that hyperhomocysteine is independent from circulating levels of insulin and from body weight. In addition, homocysteine was independent from sex steroids. This disagrees with the study of Tallova et al20 who suggested that sex hormones are nongenetic factors affecting homocysteine metabolism. In our study we further observed an inverse relationship between the homocysteine levels and age, confirming its progressive tendency to increase and negatively act with age.
The pattern of dyslipidemia shown in the current study is consistent with previous literature demonstrating an abnormal lipid profile and a proatherogenic condition in PCOS. Because insulin is a major positive regulator of lipoprotein-lipase, dyslipidemia has been fundamentally associated with insulin resistance. In our study a positive relationship was observed between total cholesterol, triglycerides, and free androgen index. This shows that hyperandrogenism may affect lipoproteins and lipids independently from insulin circulating levels. We also observed a positive relationship between total cholesterol and triglyceride circulating levels and age, confirming a tendency of lipids to increase with age and progressively and negatively act on the vascular development of atherosclerosis. The atherogenic index of plasma,13 has been proposed as an atherosclerosis marker because it is increased in cohorts at high risk for cardiovascular diseases. In our study the atherogenic index of plasma was significantly higher in PCOS patients than in other groups. Like total cholesterol and triglycerides, atherogenic index of plasma was positively correlated with free androgen index, underlining again the deep interdependence between hyperandrogenism and lipid profile in PCOS.
Polycystic ovary syndrome is considered to be a metabolic disorder. Abnormalities in glucose and insulin signaling have been implicated in the pathogenesis of PCOS in a significant subset of affected patients. However, it is not clear how frequently the hallmark of insulin resistance can be detected and if it is present in all women with PCOS. In our study we observed that among the estimates of insulin sensitivity, the homeostatic model assessment estimates for insulin resistance and insulin sensitivity index were significantly higher, whereas the fasting glucose/insulin ratio was lower in PCOS than in BMI matched controls. In addition to fasting C-peptide and insulin levels, the insulinogenic index and percent pancreatic β-cell function (all indices of pancreatic β-cell function) were significantly higher in PCOS patients than weight matched controls. Furthermore, the fasting C-peptide-to-insulin molar ratio (a useful surrogate of hepatic insulin clearance) was lower in PCOS women than controls and PCO patients. During OGTT, plasma insulin was similar in all the women we studied up to 60 minutes. Afterward, it was possible to observe a rapid decline in controls, whereas in PCOS it was weakly higher in the subsequent 30 minutes and was significantly higher at 120 minutes. The total area under the curve (AUC120) was significantly more elevated in PCOS than in controls. A similar relationship was observed for the tardive partial insulin AUC90–120. The C-peptide expressed similar tendencies. Glucose showed a marked increase at 30 minutes and persisted significantly higher at 120 minutes both in PCO and PCOS patients in comparison with control women. In addition, the glucose AUC was higher in the PCOS and PCO group with respect with nonhirsute and normally menstruating women, either considering the total and the partial (0–90 minutes and 90–120 minutes) areas under the curve. We speculated that the PCOS and, only in part, PCO patients may have a moderate-to-severe muscle insulin resistance and, in agreement with Abdul-Ghani et al,21 also a concomitant liver insulin resistance. This, associated with the increased pancreatic β-cell function and the reduced hepatic insulin clearance we observed, may explain the almost constant hyperinsulinemia in PCOS patients and their higher risk for progression to type-2 diabetes and vascular diseases. In our study we observed that the carbohydrate metabolic disorder is correlated with both weight and androgens. That androgens rise in conjunction with the degree of hyperinsulinemia is not surprising and is consistent with previous reports indicating that insulin directly stimulates ovarian androgen production. On the other hand, body weight was directly correlated with hyperinsulinemia. This is consistent with other studies1 and suggests that whereas obesity is quite often associated with insulin resistance, individual variation in insulin sensitivity exists largely independently from the degree of obesity. In conclusion, on the basis of the data presented, we postulate that PCOS, and, in part PCO, may be considered a condition associated with an increased vascular risk.
1. Meyer C, McGrath BP, Teede HJ. Overweight women with polycystic ovary syndrome have evidence of subclinical cardiovascular disease. J Clin Endocrinol Metab 2005;90:5711–6.
2. Legro RS. Polycystic ovary syndrome and cardiovascular disease: a premature association? Endocr Rev 2003;24:302–12.
3. Cenk Sayin N, Gucer F, Balkanli-Kaplan P, Ali Yuce M, Yardim T. Insulin resistance and lipid profile in women with polycystic appearing ovaries: implications with regard to polycystic ovary syndrome. Gynecol Endocrinol 2003;17:387–96.
4. Engin-Ustun Y, Ustun Y, Meydanli MM, Kafkasli A, Yetkin G. Are polycystic ovaries associated with cardiovascular disease risk as polycystic ovary syndrome?. Gynecol Endocrinol 2006;22:324–8.
5. Loucks TL, Talbott EO, McHugh KP, Keelan M, Berga SL, Guzick DS. Do polycystic-appearing ovaries affect the risk of cardiovascular disease among women with polycystic ovary syndrome?. Fertil Steril 2000;74:547–52.
6. Battaglia C, Artini PG, Salvatori M, Giulini S, Petraglia F, Maxia N, et al. Ultrasonographic patterns of polycystic ovaries: color Doppler and hormonal correlations. Ultrasound Obstet Gynecol 1998;11:332–6.
7. Ferriman D, Gallwey JD. Clinical assessment of body hair growth in women. J Clin Endocrinol Metab 1961;21:1440–7.
8. Battaglia C, Artini PG, D'Ambrogio G, Genazzani AD, Genazzani AR. The role of color Doppler imaging in the diagnosis of polycystic ovary syndrome. Am J Obstet Gynecol 1995;172:108–13.
9. Battaglia C, Mancini F, Regnani G, Persico N, Volpe A, De Aloysio D. Hormone therapy and ophthalmic artery blood flow changes in women with primary open-angle glaucoma. Menopause 2004;11:69–77.
10. Gosling RG, Lo PT, Taylor MG. Interpretation of pulsatility index in feeder arteries to low-impedance vascular beds. Ultrasound Obstet Gynecol 1991;1:175–9.
11. Battaglia C, Regnani G, Mancini F, Iughetti L, Flamigni C, Venturoli S. Polycystic ovaries in childhood: a common finding in daughters of PCOS patients. A pilot study. Hum Reprod 2002;17:771–6.
12. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502.
13. Tan MH, Johns D, Glazer NB. Pioglitazone reduces atherogenic index of plasma in patients with type 2 diabetes. Clin Chem 2004;50:1184–8.
14. Facchinetti F, De Martis S, Neri I, Caputo A, Volpe A. Effects of transdermal glyceryltrinitrate on 24-h blood pressure changes in patients with gestational hypertension. Prenat Neonat Med 1997;2:22–8.
15. Modan M, Harris MI, Halkin H. Evaluation of WHO and NDDG criteria for impaired glucose tolerance. Results from two national samples. Diabetes 1989;38:1630–5.
16. American Diabetes Association. Clinical practice recommendations. Diabetes Care 1997; 20 (Suppl 1):S1–S70.
17. Powers WJ, Press GA, Grubb RL Jr, Gado M, Raichle ME. The effect of hemodynamically significant carotid artery disease on the hemodynamic status of the cerebral circulation. Ann Intern Med 1987;106:27–34.
18. Wakisaka M, Nagamachi S, Inoue K, Morotomi Y, Nunoi K, Fujishima M. Reduced regional cerebral blood flow in aged noninsulin-dependent diabetic patients with no history of cerebrovascular disease: evaluation by N-isopropyl-123I-p-iodoamphetamine with single-photon emission computed tomography. J Diabet Complications 1990;4:170–4.
19. Yarali H, Yildirir A, Aybar F, Kabakci G, Bukulmez O, Akgul E, et al. Diastolic dysfunction and increased serum homocysteine concentrations may contribute to increased cardiovascular risk in patients with polycystic ovary syndrome. Fertil Steril 2001;76:511–6.
20. Tallova J, Tomandl J, Bicikova M, Hill M. Changes of plasma total homocysteine levels during the menstrual cycle. Eur J Clin Invest 1999;29:1041–4.
© 2008 The American College of Obstetricians and Gynecologists
21. Abdul-Ghani MA, Jenkinson CP, Richardson DK, Tripathy D, DeFronzo RA. Insulin secretion and action in subjects with impaired fasting glucose and impaired glucose tolerance: results from the Veterans Administration Genetic Epidemiology Study. Diabetes 2006;55:1430–5.