Women diagnosed with polycystic ovary syndrome (PCOS) have a higher prevalence of several well-known cardiovascular risk factors, which likely translates into a higher risk of cardiovascular events such as myocardial infarction, stroke, and sudden cardiac death. The most commonly implicated risk factors include disordered insulin regulation, including impaired glucose tolerance, insulin resistance, and diabetes, as well as dyslipidemia, metabolic syndrome, and nonalcoholic steatohepatitis. Each of these risk factors is influenced to varying degrees by the higher rate of obesity among women with PCOS. An increased rate of obstructive sleep apnea and depression/anxiety is also thought to contribute to an increase in cardiovascular risk in PCOS patients. Identifying PCOS women and screening for these disorders affords the clinician an opportunity to prevent or lessen the incidence of adverse outcomes in these women and potentially reduce the population burden of cardiovascular disease by early intervention.

Three major sets of diagnostic criteria currently define PCOS. The first set was outlined at the National Institutes of Health (NIH) in Bethesda, Maryland, in 1990 [1]. This has largely been replaced in clinical practice by the criteria defined in Rotterdam, The Netherlands, in 2003 by a task force sponsored by the European Society of Human Reproduction and Embryology (ESHRE) and the American Society for Reproductive Medicine (ASRM) [2]. More recently the Androgen Excess Society (AES) outlined its own set of criteria in 2009 [3,4]. All criteria stipulate that other diagnoses, such as nonclassic congenital adrenal hyperplasia (NC-CAH), Cushing’s syndrome, androgen-secreting tumors, other causes of anovulatory cycles such as hyperprolactinemia or thyroid disease, and premature ovarian failure, be ruled out. The reported prevalence of PCOS in a given population depends largely on the diagnostic criteria utilized by the investigators.

Increased prevalence and risks for impaired glucose tolerance (IGT) and type 2 diabetes mellitus (T2DM) among women with PCOS are well described. It is estimated that 60–80% of women with PCOS have one of these diagnoses, and the prevalence increases to a remarkable 95% among obese PCOS patients [5]. Prevalence increases with age. A study of 27 adolescent girls with PCOS found that 8 (30%) had IGT and 1 (4%) already had developed T2DM [6]. A larger study of 254 women with PCOS with a wide age range (14–44) found that 31.1% had IGT and 7.5% had T2DM diagnosed by 2-hour 75 g oral glucose tolerance test. This was higher among the obese PCOS population and higher than age- and BMI-matched controls [7]. By their fourth decade of life, 40% of women with classic PCOS will have developed IGT or T2DM [8,9]. Even after adjusting for age and BMI, women with PCOS have a twofold increased risk of developing T2DM [10]. New diagnoses of T2DM occur at an estimated 5.7 cases per 1000 patient-years, compared with 1.7 per 1000 patient-years among reproductive age women without PCOS. This phenomenon is noted despite a similar fat distribution between PCOS and non-PCOS subjects [11]. Criteria for diagnosing diabetes have changed through the years. This of course will affect reported prevalence.

The prevalence of IGT depends upon the PCOS criteria utilized as well, with a higher prevalence found among women with classic PCOS using NIH criteria [12]. The severity of hyperinsulinemia correlates with the severity of clinical hyperandrogenism. There is also some ethnic variation in predisposition to insulin resistance among women with PCOS. For example, Mexican-American women have higher rates of insulin resistance than Caucasian American women. This finding prompted the authors to propose using different cut-off values for screening different ethnic groups [13]. The severity of insulin resistance (IR) has also been associated with the degree of menstrual derangement, with amenorrheic women displaying more pronounced IR than women with oligomenorrhea or polymenorrhea [14].

Although metabolic syndrome (MetS) can be defined using various criteria, it is very prevalent among populations of women with PCOS regardless of the criteria used. Definitions for having MetS have been proposed by the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) [15], International Diabetes Federation (IDF) [16], American Heart Association/National Heart, Lung and Blood Institute (AHA/NHLBI) [17], and the recent joint definition proposed by the IDF, NHLBI, AHA, World Heart Federation, International Atherosclerosis Society and International Association for the Study of Obesity (Joint definition) [18]. (Table 9.1). The most commonly used criteria in prevalence studies of women with PCOS are included in the NCEP ATP III definition.

The reported prevalence of MetS in the United States among patients with classic PCOS varies from 33% to 47%. Notably, this is two to three times higher than the prevalence in age-matched controls. It is related to obesity rates, particularly abdominal adiposity. Outside the United States, prevalence reports range from 8% to 30%. Among a population of Swedish women with PCOS, 23.8% met NCEP ATP III criteria for MetS, compared with 8.0% of controls [19]. A South Asian study found an even greater disparity, with 30.6% of PCOS patients and only 6.3% of controls identified as having MetS [6–20]. Women in southwestern China with PCOS are five times more likely to have MetS than controls, with a prevalence of 25.6% [21]. A direct comparison of a Chinese PCOS population to a Dutch PCOS population showed that the Chinese women had greater rates of hyperandrogenism, increased waist circumference, and higher BMI. The authors concluded that Chinese women have an increased risk of metabolic complications, demonstrating ethnic and regional variations [22]. One additional example of this variation is a study in Iran showing no association between PCOS and the incidence of MetS, despite showing an association with IR [23].

There may be a stronger association between MetS and PCOS when stricter (NIH or AES) diagnostic criteria are used to define PCOS. A study in Turkey found a higher metabolic risk among women who met NIH criteria than among those who met only Rotterdam criteria [24]. A contrasting study in Turkey found that the risk of MetS is approximately double for PCOS, regardless of whether NIH, Rotterdam, or AES criteria are used to define the syndrome [25]. There is some evidence that this relationship may be entirely dependent upon obesity rates. One recent study found no independent association between PCOS and MetS, particularly when newer MetS criteria (IDF, AHA/NHLBI, or joint definition) were used rather than NCEP ATP III criteria. Obesity appeared to be the primary determinant in this population; while there were more obese patients in the PCOS group, multivariable logistic regression showed no relationship between PCOS and MetS when controlling for BMI [26]. Similarly, among adolescents, MetS is associated with visceral adiposity independent of PCOS [27].

Further contributing to a worsened cardiovascular risk profile is the increased rate of dyslipidemia. In the United States, about 70% of women with PCOS have abnormal lipids, compared with 52.9% in southwestern China [21]. In both cases, this is approximately double the rate among controls. The most common abnormality is low high-density lipoprotein cholesterol (HDL-C at 57.6%), followed by high triglycerides (28.3%) [28]. The former is dependent on glucose and insulin levels, while the latter varies with age. Elevated low-density lipoprotein cholesterol (LDL-C) is also associated with PCOS. This is most likely because of more circulating small atherogenic LDL particles and higher apolipoprotein CIII. IR impairs the ability of insulin to suppress lipolysis, increasing circulating free fatty acid release from adipose stores [5].

Women and adolescents with PCOS often display elevated blood pressure (BP) compared with age- and weight-matched controls, as well as a higher prevalence of hypertension [29,30]. One study of 34 adolescents showed that obese girls with PCOS had significantly higher 24-hour mean BP, daytime mean BP, and daytime diastolic BP, as well as a less pronounced nighttime diastolic dip in BP compared with obese controls [31]. Hypertension, however, is a late stage component of MetS and many studies do not find higher prevalence of hypertension.

Nonalcoholic fatty liver disease (NAFLD) is an important disorder, not only because of its potential for progression to more devastating liver pathology such as cirrhosis or hepatocellular carcinoma, but also because of its association with development of diabetes and cardiovascular disease. This condition is more prevalent in women with PCOS. A cohort of 31 premenopausal PCOS patients was found to have an overall rate of 62% [32]. Compared with age-matched obese controls, obese PCOS women had a significantly higher rate of NAFLD (73.3% compared with 46.7%) [33]. Among patients with PCOS, those with hyperandrogenemia have higher rates of NAFLD than those with normal androgens, even when controlling for overall and visceral fat stores and for IR [34]. IR has been shown to be an independent contributor to the development of NAFLD in PCOS [35]. Patients with PCOS also show a higher rate of disease progression to steatohepatitis and fibrosis compared with the overall population of patients with NAFLD [36,37].

Obstructive sleep apnea (OSA) is a disorder characterized by multiple episodes of apnea due to collapse of the upper airway during sleep, causing transient hypoxia and arousal throughout the night. This syndrome is associated with obesity and is characterized by an increased incidence of IR, diabetes, atherosclerosis, and cardiovascular disease [38]. Women with PCOS are at higher risk of OSA than reproductively normal controls [39], and the apnea–hypopnea index correlates positively with waist-to-hip ratios and serum total and free testosterone. A study investigating the prevalence of OSA among a cohort of premenopausal women with PCOS found 70% of women met diagnostic criteria, independent of obesity status [40]. However, a case-control study comparing obese and non-obese PCOS patients to controls found no cases of OSA in either non-obese group and concluded that the increased risk of OSA in PCOS is related primarily to the increased prevalence of obesity in this population [41]. Certainly the increased risk of hyperandrogenism and central obesity in this population should prompt the experienced clinician to screen for OSA once the diagnosis of PCOS has been established, as treatment can improve insulin sensitivity and possibly mediate cardiovascular risk [38].

It is estimated that as many as 40–85% of women with PCOS are overweight or obese, and even non-obese women or girls with PCOS are reported to have an elevated waist-to-hip ratio compared with non-PCOS BMI-matched controls. The various metabolic derangements of PCOS that contribute to a worsened cardiovascular risk profile are linked to obesity, in particular central obesity often manifested as an elevated waist-to-hip ratio. As reported above, obese patients with PCOS have higher rates of IR, diabetes, hypertension, dyslipidemia, NAFLD, and MetS. The degree to which these risk factors are due to obesity alone or obesity interacting with PCOS remains unclear. For example, age- and BMI-matched controls have lower rates of IR and hyperinsulinemia than their counterparts with PCOS among both lean and obese women, suggesting that this relationship is at least partially independent of obesity [38]. However, BMI has been shown to best predict IGT and MetS in PCOS [42].

Adipose tissue is an endocrine tissue, releasing multiple adipokines whose functions are increasingly becoming understood [43]. Leptin is one of the most well described, and levels may be increased in women with PCOS independent of obesity [44–46]. However, this is not replicated in all studies [47,48], and has not been linked to an increased risk of cardiovascular disease. Adiponectin, an adipokine with antidiabetic, anti-inflammatory, and antiatherosclerotic properties, may be lower in women with PCOS, but there is some disagreement between studies [49–51]. Visfatin, expressed preferentially in visceral adipose tissue such as that found in central obesity [52], is positively correlated with BMI, fasting insulin, IR, and BP, but has an unclear role in PCOS [53–55]. It has been shown to be elevated in lean, glucose-tolerant women with PCOS compared with controls [56]. The chemotactic properties of chemerin may contribute to the inflammatory component of obesity as a disease. It is involved in processes such as adipocyte differentiation, glucose homeostasis, and lipolysis [57,58], and recruits such inflammatory cells as lymphocytes and mononuclear cells. Given their known properties, adipokines may play a role in the pathogenesis of the metabolic derangements of PCOS, although these direct relationships have yet to be elucidated.

Various direct markers of cardiovascular risk are higher in PCOS patients, and available evidence suggests that this may translate to increased cardiovascular morbidity and mortality [5]. Women with PCOS have a lower event-free survival, and cardiovascular events increase as the number of diagnostic criteria for PCOS increases [59]. The extent of cardiovascular risk does vary by PCOS phenotype [60], with the phenotype of polycystic ovarian morphology with anovulation behaving most like control subjects [61]. A short-term prospective study with 4.7 years average of follow-up found no increased risk of adverse outcomes, including new diabetes, cancer, large vessel disease, or mortality, among PCOS women compared to controls [62], although the women in this study were relatively young. A Swedish cohort study with 21 years of follow-up, however, showed no difference in the incidence of myocardial infarction, stroke, cancer, or overall mortality between a group of PCOS patients and control subjects [63], despite an increased rate of hypertension and elevated triglycerides in the PCOS group. Variations among population studies reveal that more epidemiologic studies are required to precisely characterize the increased cardiovascular risk associated with PCOS.

Studies of carotid intima–media thickness (CIMT), a marker of atherosclerosis and future cardiovascular risk, show statistically significantly thicker measurements in PCOS compared with controls. Among premenopausal patients at least 40 years of age, CIMT was significantly thicker (0.68 mm versus 0.63 mm) in PCOS patients than in controls [64]. Similarly, women at least 45 years of age with PCOS had a thicker CIMT of 0.78 mm compared with 0.70 mm in those without, a statistically significant difference [65]. A meta-analysis of 19 studies found a 0.072 mm difference in CIMT between PCOS and age- and BMI-matched controls among high-quality studies [66].

Epicardial fat, or the layer of visceral adipose tissue deposited around the heart, can be measured using echocardiography. The epicardial fat layer correlates with visceral fat content, and has been developed as a marker of cardiovascular disease risk. The thickness of this layer has been positively correlated with the presence of coronary heart disease (CHD) and with the severity of coronary artery occlusion. Women with PCOS have an increased epicardial fat thickness compared with age- and BMI-matched controls, which has been shown to correlate with increased IR, total cholesterol, triglycerides, and androgen levels, as well as decreased adiponectin levels [67]. Epicardial fat thickness also correlates positively with BMI, waist-to-hip ratio, Ferriman–Gallwey score of hirsutism, fasting insulin, 17-OH progesterone, CIMT, and negatively with HDL cholesterol [68]. This measure appears to be emerging as an important indicator of cardiovascular risk and metabolic derangements in PCOS.

Women with PCOS within all weight categories have been shown to have a higher prevalence of cardiac dysfunction, including increased left ventricular mass [69,70], left ventricular stiffness [69,71], and degrees of diastolic dysfunction [69,71]. They have a higher incidence of coronary artery calcification independent of BMI and age [72–74].

Obesity, as previously stated, is an inflammatory condition with adipose tissue acting as an endocrine organ, supporting the release of multiple cytokines that have been linked to increased cardiovascular risk. Many of these markers have been shown to be significantly increased in PCOS, including C-reactive protein (CRP), homocysteine, plasminogen activator inhibitor-1 (both antigen and activity are increased), vascular endothelial growth factor (VEGF), asymmetric dimethylarginine (ADMA), advanced glycation end product (AGEs), and lipoprotein(a). Borderline significant elevations were also noted in tumor necrosis factor-alpha (TNF-α), endothelin-1, and fibrinogen [76].

Many features associated with PCOS have been shown to affect quality of life, particularly an increase in mood disorders. Up to 57% of women with PCOS have at least one psychiatric diagnosis [77]. Healthcare providers should be aware of the dramatically increased risk of depression and anxiety among women with PCOS, and should administer validated screening tools as part of their routine care [78]. Having PCOS confers a fourfold increased risk of depression [79], and nearly sevenfold increased risk of anxiety symptoms [80]. The overall prevalence of depression among the PCOS population has been reported at 26.4%. When NIH criteria are used to make the diagnosis, the prevalence is even higher at 35%, even adjusted for BMI, family history of depression, and infertility. Even so, obesity is known to interact with depression in the general population, and BMI, education, and parity have been identified as predictors of mild to moderate depressive symptoms [81]. Depression scores also correlate with levels of IR, lipid abnormalities, and components of the MetS, suggesting an association between depression and an increased cardiovascular risk [82]. Hyperandrogenism and hirsutism have not been shown to be independent risk factors for depression, but scores do correlate with patients’ assessment of self-worth and evaluation to their own appearance [83].

Patients diagnosed with PCOS warrant a full evaluation for detection of cardiovascular risk, including measurement of height, weight, blood pressure, and waist-to-hip ratio. A complete lipid profile should be determined, and reassessment every 2 years if it is normal, as recommended by the AE-PCOS Society [5]. A 2-hour 75 g oral glucose challenge test should be performed if the BMI is greater than 30 kg/m², or in cases of age >40, and personal or family history of diabetes. In patients with normal testing, this may be repeated every 2 years or upon the appearance of additional risk factors. Strong consideration should be given to screening for depression and OSA as well, as these comorbidities may contribute to a woman’s cardiovascular risk and overall quality of life.