Polycystic Ovarian Syndrome: Long-Term Health Consequences

Temeka Zore, MD; Nikhil V. Joshi, MD; Daria Lizneva, MD, PhD; Ricardo Azziz, MD, MPH, MBA


Semin Reprod Med. 2017;35(3):271-281. 

In This Article

Longer-term Complications

Women with PCOS demonstrate greater degrees of insulin resistance and secondary hyperinsulinemia than weight-matched controls, and consequently are at greater risk of developing long-term complications associated with metabolic dysfunction, including a higher lifetime risk of type 2 diabetes mellitus (T2DM), nonalcoholic fatty liver disease (NAFLD), metabolic syndrome, hypertension, and possibly vascular complications.[1]

Metabolic Dysfunction

Insulin resistance in PCOS is multifactorial in etiology and distinct from other disease states. Initially, and as noted earlier, excess adiposity may contribute to the metabolic dysfunction of PCOS. Although visceral adiposity has been reported to negatively impact metabolic function to a greater degree than nonvisceral adiposity, we observed that in PCOS women, both visceral and subcutaneous adipose mass independently and negatively predicted insulin sensitivity and glucose effectiveness.[31] In fact, there are conflicting data regarding the prevalence of central or visceral adiposity, which in some studies has been reported to be the same[31,32] or in others increased[33] in women with PCOS.[9] Anovulatory women with PCOS, in particular, have greater visceral fat.[34] However, women with PCOS have higher rates of hyperinsulinemia independent of obesity,[17] and pancreatic β-cell dysfunction and insulin resistance, measured by the acute insulin responses to glucose, independent of BMI in affected women.[35]

Adipose tissue is the body's largest endocrine organ, and the insulin resistance of PCOS may also result from altered production in adipocyte byproducts, regardless of the presence of obesity.[34,36] For example, reduced levels of adiponectin[37] or adiponectin production by adipocytes[36] may contribute to the degree of insulin resistance, in addition to the endothelial dysfunction and cardiovascular risk, of PCOS.

Other mechanisms for insulin resistance in PCOS may relate to postbinding defects in insulin receptor signaling, leading to abnormal glucose transporter 4 (GLUT4) production.[38,39] Alternatively, the mitogenic signaling pathways appear to be unaffected. Other possible mechanisms for insulin resistance include altered catecholamine-mediated lipolysis, pancreatic β-cell dysfunction, and decreased hepatic clearance.[38] Finally, a recent study suggested that the gut microbiome is altered in a rat model of PCOS; rats with letrozole-induced hyperandrogenemia had higher levels of Prevotella species and fewer Lactobacillus.[40] Alterations in gut microbiota may contribute to insulin resistance in women with PCOS, although further study is needed.

As noted earlier, the phenotype of PCOS plays an important role in determining the metabolic, and possibly other risks of PCOS. For example, phenotypes A to C are associated with an increased risk of metabolic syndrome, whereas phenotype D is not associated with an increased risk compared with the general population matched, even in obese women.[41] Others found that women with PCOS, as defined by the NIH criteria (encompassing phenotypes A and B) had higher rates of insulin resistance compared with the general population irrespective of BMI, while the other phenotypes did not, implying that hyperandrogenism contributes to insulin resistance.[42] Hyperandrogenemia is independently associated with T2DM,[43] further supporting the notion that phenotypes A through C represent a higher-risk group in terms of lifelong metabolic risk. Additionally, there is some evidence that within the anovulatory phenotypes, intermenstrual cycle lengths greater than 35 days are correlated with higher degrees of insulin resistance as measured by the homeostasis model assessment for insulin resistance (Fig. 2).[44]

Figure 2.

Depicted is the association between the degree of ovulatory and menstrual irregularity and the degree of insulin resistance, as measured by the homeostasis model assessment for insulin resistance (HOMA-IR) in 494 untreated consecutive women with polycystic ovarian syndrome diagnosed by the NIH 1990 criteria. There is a near-linear relationship between the severity of the ovulatory and menstrual dysfunction and the severity of the insulin resistance. (Reprinted with permission from Brower et al.44)

Glucose Intolerance and T2DM. The risks of developing T2DM and metabolic syndrome appear to be increased independent of the degree of obesity and are found, albeit at a lower rate, even in nonobese PCOS women. Not unexpectedly, women with PCOS have higher incidences of impaired glucose tolerance (IGT), a precursor to T2DM and metabolic syndrome than BMI-matched controls.[45] Overall, women with PCOS have a five- to sevenfold higher risk for T2DM than age-matched women without the disorder.

Metabolic Syndrome and NAFLD. NAFLD also appears to be more common in women with PCOS, and especially in those with dyslipidemia and metabolic syndrome.[46,47] The risk for cardiovascular disease is also elevated in individuals with metabolic syndrome. Based on retrospective analyses of women with PCOS, the prevalence of metabolic syndrome in women with PCOS as defined by NIH criteria may be as high as 33 to 43%, approximately twice that of the general population based on estimates using the National Health and Nutrition Examination Survey, and even higher for age-matched controls. Women with PCOS and metabolic syndrome have greater levels of insulin resistance and a higher prevalence of hyperandrogenemia than those without.[24,48,49] Women exhibiting the hyperandrogenic phenotypes of PCOS have an increased prevalence of metabolic syndrome when compared with those with oligo-ovulation and PCO ovaries alone (Fig. 3).[41]

Figure 3.

Bar graph depicting the prevalence of metabolic syndrome, using the NCEP-Adult Treatment Panel III criteria, in a retrospective analysis of 105 women with polycystic ovarian syndrome (PCOS) diagnosed by the Rotterdam 2003 criteria. The prevalence of metabolic syndrome was much higher (36–42%) in women with hyperandrogenic phenotypes (phenotypes A–C) compared with women with PCOS without hyperandrogenism (phenotype D, 20%) or controls (8%). (Reprinted with permission from Shroff et al.41)

Dyslipidemia and Vascular Dysfunction

Dyslipidemia is more common among women with PCOS. Obese women with PCOS tend to have higher levels of triglycerides and lower levels of high-density lipoprotein (HDL), and may also have higher levels of low-density lipoprotein (LDL).[50,51] Hypertension is also more prevalent in women with PCOS, and especially in young, obese women with PCOS compared with controls, though there is also an increased incidence regardless of the presence of obesity.[52] Evidence for coronary artery disease appears to be increased in women with PCOS. Compared with controls matched for age and weight, women with PCOS demonstrate higher amounts of coronary artery calcifications measured by electron beam computed tomography.[53]

Notwithstanding the fact that markers of cardiovascular disease are elevated in women with PCOS, the incidence of actual cardiac events, such as myocardial infarction, does not appear to be increased in women with PCOS, at least in most studies to date.[54–56] It is possible that the current studies are not adequately powered, or may be studying women who are too young. In contrast, evidence suggests that cerebrovascular events are slightly more common in women with PCOS compared with controls.[57] Venous thromboembolism is also more common in women with PCOS, with a twofold increased risk compared with controls in those patients taking estrogen-containing oral contraceptives (OCs).[58]


Women with PCOS have many risk factors that place them at increased risk of hormonally sensitive neoplasias, including endometrial, ovarian, and breast cancers. Risk factors include unopposed estrogen exposure from persistent hyperandrogenemia with subsequent aromatization, chronic anovulation, hyperinsulinemia, and hyperglycemia. Additionally, obesity is a well-known risk factor for both endometrial and breast cancers and possibly ovarian cancer, and it is possible that any increased risk of malignancy in PCOS may be due in part to their increased prevalence of obesity.[59,60] Overall, most studies have demonstrated that PCOS is associated with an increased risk of endometrial adenocarcinoma, with mixed results concerning the risk of ovarian cancer risk and no significant increased risk of breast cancer. While it makes intuitive sense that the risk of hormonally sensitive cancers, especially endometrial cancer, would be enhanced in hormonal milieus with increased levels of circulating sex steroids and insulin, such as those present in most PCOS women presenting with phenotypes A and B (i.e., "classic PCOS"), there are limited data concerning the relationship between phenotype and neoplastic risk in PCOS.

Endometrial Cancer. Most studies have demonstrated an increased risk of endometrial cancer in women with PCOS.[60–63] A large cohort of more than 12,000 women in the Danish Cancer Registry were identified as having PCOS and followed for cancer development.[63] The investigators observed a fourfold increased risk of endometrial cancer among PCOS patients compared with the general Danish female population, with the majority of cases diagnosed being type I endometrial cancer. A systematic review and meta-analysis exploring the risk of endometrial cancer in PCOS was consistent with other reviews which indicated an increased risk of neoplasia in the syndrome (OR: 2.79; 95% CI: 1.31–5.95, p < 0.008).[60]

Breast Cancer. Breast cancer is the most common malignancy among women and the second leading cause of mortality due to cancer in both developed and developing countries. Overall studies have not been able to demonstrate an increased risk of breast cancer in association with PCOS.[60,63,64] A large meta-analysis including more than 45,000 patients estimated that the association between PCOS and breast cancer risk in case–control studies was 0.83 (95% CI: 0.44–1.31) and that in cohort studies was 1.18 (95% CI: 0.93–1.43).[64] Another systematic review and meta-analysis also failed to find an increased risk of breast cancer in women with PCOS (OR: 0.95; 95% CI: 0.64–1.39), with no change in results when the analysis excluded women older than 54 years.[60] Additionally, data from the Danish Cancer Registry compared the incidences of cancer in women diagnosed with PCOS to that of the general population using standardized incidence ratios (SIRs).[63] In total, 279 cases of cancer were diagnosed, and there was no association between PCOS and breast cancer observed (n = 59, SIR = 1.1; 95% CI: 0.8–1.4).

A study by Kim et al identified more than 1,500 women with newly diagnosed in situ or invasive breast cancer and compared them to age-matched controls.[65] Approximately 2.2% of the population reported having a PCOS diagnosis established by a medical professional and the remainder reported no PCOS or were unsure, 96.3 and 1.5%, respectively. The women diagnosed with PCOS were more likely to have used OCs, to have a history of infertility, and to not have regular cycles. These investigators found a slight association between PCOS and breast cancer incidence, although it did not reach significance. However, the risk varied significantly when taking into account menopausal status. Using unconditional logistic regression, the investigators noted an almost threefold increase in the rate of breast cancer in premenopausal women with PCOS (multivariate-adjusted OR: 2.74; 95% CI: 1.13–6.62), which contrasted to that of postmenopausal PCOS women in whom breast cancer incidence decreased by 33%. Using cluster analysis to determine whether PCOS-related symptoms or sequelae may increase the risk of breast cancer, these investigators also noted that OC use in premenopausal women increased this risk. The investigators hypothesized that their findings could be due to the abnormal hormonal profile of PCOS, such that this hormonal milieu has an immediate impact on the premenopausal breast, and that this effect ameliorates with time. Additionally, as OC use appears to be associated with the increased breast cancer risk, it is possible that the increased premenopausal breast malignancy risk in women with PCOS is associated with PCOS-related symptoms or sequelae and not the disorder itself.

Obesity and T2DM have been associated with an increase in breast cancer risk and both disorders are observed in PCOS.[66–68] A systematic review and meta-analysis of seven observational studies assessed the use of metformin and breast cancer risk.[69] The study found a lower incidence of invasive breast cancer in women who used metformin (OR: 0.83; 95% CI: 0.71–0.97). Additionally, a stronger association was found when the analyses were limited to studies with longer duration of metformin use (OR: 0.75; 95% CI: 0.62–0.91 vs. OR: 0.94; 95% CI: 0.84–1.06, respectively). These investigators went on to note that because these findings were based on the results of observational studies, it may reflect bias or confounding. However, the stronger association of risk with longer-term metformin use, that is, indicative of a dose response, suggests that the findings may be real and thus deserve confirmation in prospective trials.

Ovarian Cancer. Evidence has been mixed regarding the association of PCOS and ovarian cancer risk.[59,60,63,70] A meta-analysis performed by Barry et al assessed the risk of ovarian cancer in PCOS and noted that it did not appear to be significantly increased (OR: 1.41; CI 95%: 0.93–2.15); however, when studies including women older than 54 years were excluded, the risk for ovarian cancer in PCOS significantly increased (OR: 2.52; 95% CI: 1.08–5.89, p < 0.03).[63] The investigators did note that the heterogeneity of the studies included and the variation in the diagnostic criteria for PCOS may have resulted in an exaggeration in the risk.[60] Data from the aforementioned Danish Cancer Registry, a large Danish cohort of more than 12,000 women, did not find an association between PCOS and ovarian cancer (n = 10, SIR = 1.8; 95% CI: 0.8–3.2).

In a retrospective population-based case–control study from the Cancer and Steroid Hormone study, 476 women with histologically confirmed epithelial ovarian cancer were studied along with more than 4,000 controls.[70] Seven women with ovarian cancer and 24 controls reported a diagnosis of PCOS; ovarian cancer risk was 2.5-fold higher in women with PCOS (95% CI: 1.1–5.9). This risk was increased in women who had never used OCs as well as in those women with a BMI between 13.3 and 18.5 kg/m.2 Overall, we should note that this study is limited by its retrospective nature and chance for recall bias.

Leiomyomata. There are relatively few studies evaluating the risk of leiomyomata (uterine fibroids) in women with PCOS, although hypothetically there may be an increased risk due to elevated sex steroids and insulin, an anabolic hormone. Additionally, the increased luteinizing hormone levels in women with PCOS may be important to the development of uterine fibroids. A retrospective cohort over a 6-year period among women recruited from the Black Women's Health Study demonstrated an increased risk of fibroids in women with PCOS.[71] Analysis of data from the Study of Women's Health Across the Nation (SWAN) also supported an association between PCOS, or at least hyperandrogenism, and leiomyomata. In a cohort of 3,240 women of diverse ethnicity, 45 to 52 years old, women with high testosterone had an increased risk of incident, but not recurrent fibroids.[72] This risk was further elevated in those with high testosterone and estradiol. However, high estradiol and testosterone was associated with lower risk of recurrent fibroids. Conversely, the risk of recurrent fibroids was actually mitigated in these women.

Alternatively, in a study of 1,070 women, 18 to 40 years old, with gynecologic or infertility problems, fewer patients with polycystic ovaries (but not necessarily PCOS) had fibroids as compared with patients with normal ovaries (11.4 vs. 27.3%, respectively, p < 0.001).[73] This negative correlation was maintained irrespective of age, parity, or ethnic origin. Overall, the association of PCOS and the development of uterine leiomyomata remain unclear.

Other Neoplasia. Further data from the Danish Cancer Registry have highlighted increased risks for kidney, colon, and brain cancers among women with PCOS,[63] although these findings remain to be confirmed.

Mental Health Disorders

Multiple studies demonstrate a consistent link between PCOS and mental health issues, including an increased prevalence of depression, anxiety, and decreased sexual satisfaction and QOL.[28,29,74–76] Some observers have suggested that the physical manifestations of PCOS, including the associated hirsutism, acne, obesity, and patients' concerns regarding long-term health risks, lead to the decreased self-esteem and lower QOL.[76] In turn, low self-esteem and decreased QOL can result in an increased propensity for mood disorders and social anxiety.

Depressive Disorders. A major depressive disorder (MDD) affects nearly 14.8 million American adults each year and is more prevalent in women.[77] In a survey performed in the United States, the prevalence of MDD in women aged 18 to 44 ranged from 12 to 14% with a mean onset of 30.4 years.[78] A cohort study of 103 women with PCOS and 103 controls reported a higher prevalence of MDD in PCOS women compared with controls, 35 versus 10.4%, respectively, independent of obesity.[74] The investigators did note, however, that there was an association between the Beck Depression Inventory (BDI) scores and BMI, although the correlation was not strong enough to fully account for the increase in depression. A subsequent systematic review and meta-analysis reported a fourfold increase in the prevalence of depressive symptoms in women with well-defined PCOS versus controls.[28] However, a limitation to this analysis included the variety of screening tools used in the various studies.

Coexisting generalized anxiety disorder (GAD) in depressed patients may increase risks of suicide and functional disability, as well as lead to a decreased response to treatment. Therefore, women with PCOS should undergo routine screening for mental health disorders with validated screening tools and be referred to an appropriate health care professional for further treatment management as indicated.[29]

Anxiety Disorders. In women, between 35 and 50% of mood disorders are associated with anxiety disorders, especially GAD.[79] Anxiety affects 5 to 8% of women in the primary clinical setting.[75] There are less studies available reporting on the association of anxiety and PCOS; however, those who have examined this issue have demonstrated an increase incidence in anxiety symptoms in PCOS women compared with controls. A meta-analysis and systematic review by Dokras et al that included four studies indicated that, compared with controls, PCOS women had a 6.7 higher OR (95% CI: 2.5–18.9; p = 0.02) of having generalized anxiety symptoms.[29] The majority of PCOS patients and controls were overweight or obese, with most studies included in the meta-analysis indicating that there was no difference in terms of age or BMI between the two groups. Additionally, reports note many of the same associations with anxiety as for depression in women with PCOS, including feelings of social stigmatization, low self-esteem, poor body image, and concern for future health and fertility.[29,30,80,81]

Psychosexual Dysfunction. The health-related QOL (HRQOL) has been shown to be lower in all domains, including physical, mental, emotional, and social functioning in women with PCOS compared with healthy controls with the presence of obesity, hirsutism, menstrual dysfunction, and infertility as the main factors leading to decreased QOL.[82] To evaluate QOL, psychosocial well-being, and sexual satisfaction, 50 PCOS patients and 50 controls were provided three validated questionnaires. Overall there was no difference in sociodemographic variables, although women with PCOS scored higher in those dimensions relating to obsessive-compulsive behavior, interpersonal sensitivity, depression, aggression, and psychoticism than healthy controls.[76] PCOS women were less satisfied with their sex life and reported that, per their impression, their partners were also less satisfied with their sex life. Additionally, they found themselves significantly less sexually attractive which was associated with feelings of sexual dissatisfaction. These findings are similar to those of another study that assessed 49 women with PCOS and age-matched controls and noted that women with PCOS reported decreased satisfaction with their sex life.[83] Overall, the ability for a screening tool to adequately measure aspects of QOL is improved when disease-specific screening tools are used. To date, there has been one PCOS-specific questionnaire (PCOSQ) that has been validated to measure HRQOL in these women, but further studies are warranted.[84,85]

Overall, while it makes intuitive sense that the risk of mental health dysfunction is higher in women with greater degrees of hyperandrogenism, and greater risk of metabolic dysfunction, such as those PCOS women presenting with phenotypes A and B (i.e., "classic PCOS"), there are, as yet, limited data concerning the relationship between phenotype and mental health in PCOS.