Issue: April 2022
Disclosures: Bairey Merz reports consulting for iRhythm Technologies and receiving research support from Abbott Vascular and Sanofi. Cree-Green, Dokras, Dunaif and Manson report no relevant financial disclosures.
April 21, 2022
10 min read

Longer studies, funding needed to address ‘substantial knowledge gaps’ on PCOS-related CVD

Issue: April 2022
Disclosures: Bairey Merz reports consulting for iRhythm Technologies and receiving research support from Abbott Vascular and Sanofi. Cree-Green, Dokras, Dunaif and Manson report no relevant financial disclosures.
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

Polycystic ovary syndrome is a common endocrine disorder affecting reproductive-age women, yet the complexity and heterogeneity of PCOS have confounded researchers attempting to understand its long-term consequences.

The Rotterdam criteria for adults require that a woman present with at least two of the following to receive a diagnosis of PCOS: clinical or biochemical hyperandrogenism, ovulatory dysfunction or polycystic ovaries on ultrasound; when irregular menses and hyperandrogenism are present, ultrasound is not necessary for diagnosis. Insulin resistance is common in affected women, depending on the phenotype, driving the risk for several cardiometabolic disorders, such as metabolic syndrome and type 2 diabetes.

Those cardiometabolic disorders, particularly type 2 diabetes, are risk factors for cardiovascular disease. However, studies attempting to determine the degree of CVD risk in PCOS vs. the general population have mixed findings.

In a review published in 2019, researchers noted that some prior studies and meta-analyses have shown women with PCOS have greater risk for coronary heart disease and stroke compared with women without PCOS. Some studies have shown this association to be independent of BMI, whereas other studies have suggested this excess CVD risk may be explained by traditional CVD risk factors and that the absolute risk conferred by PCOS may be small.

“It is very clear that teenagers and women in their 20s and 30s with PCOS, especially with obesity, have increased CV risk factors,” Melanie Cree-Green, MD, PhD, associate professor of pediatric endocrinology and director of the adolescent PCOS multidisciplinary clinic at the University of Colorado, Anschutz and Children’s Hospital Colorado, told Endocrine Today. “The intermediate markers, such as coronary artery calcification, suggest a positive signal [for CV risk] in PCOS. But for the actual CV endpoints, there are myriad studies — some suggesting women with PCOS are protected and some older studies suggesting these women have much higher rates of CVD. You have varied markers and varied answers.”

Several branches of the NIH, led by the National Heart, Lung, and Blood Institute and the NIH Office of Research on Women’s Health, recently organized a 2-day workshop on CV risk across the life span for women with PCOS to seek definitive answers. The workshop, held in October, brought together experts across the disciplines of endocrinology, cardiology, obstetrics and gynecology, and epidemiology to share current evidence and discuss open questions about the true CV risk picture. Key among them: What is the long-term CV risk for a women diagnosed with PCOS, and how can researchers learn more about a potential outcome that develops over decades, not years?

C. Noel Bairey Merz

“From a CV standpoint, CVD is the leading killer of women at all ages, and there are substantial knowledge gaps about the role PCOS plays in premature and later-life CVD in women,” C. Noel Bairey Merz, MD, FACC, FAHA, professor of cardiology, director of the Barbra Streisand Women’s Heart Center at the Smidt Heart Institute, Cedars-Sinai Medical Center, and a co-chair of the NIH workshop, told Endocrine Today. “If it turns out that PCOS clearly is playing a role, it could be a very valuable early risk determinator. It might even be a treatment target.”

Evidence ‘inconsistent’

Several lines of evidence suggest women with PCOS are at increased risk for cardiometabolic diseases as well as CVD, including CHD and stroke, according to JoAnn E. Manson, MD, DrPH, MACP, FAHA, chief of the division of preventive medicine at Brigham and Women’s Hospital and professor of medicine and the Michael and Lee Bell Professor of Women’s Health at Harvard Medical School. Hyperandrogenism, insulin resistance and increased visceral adiposity — all commonly occurring in PCOS — may provide biological pathways to increase CVD risk via the development of dyslipidemia, glucose intolerance and metabolic syndrome, Manson said.

“What we don’t know is how long the increased risk lasts and whether it extends into the postmenopausal years,” Manson told Endocrine Today. “Certainly, during the reproductive years and into midlife, an increased risk is present.”

JoAnn E. Manson

Manson said long-term studies that include self-reported menstrual cycle data suggest a link between long menstrual cycles (at least 40 days between periods) or irregular menses — both potential proxies for PCOS — and long-term CV risk. Data from the Nurses’ Health Study II, which included women aged 25 to 42 years at baseline showed participants who reported that their menstrual cycles were always irregular or long were two to three times more likely to develop type 2 diabetes during 10 years of follow-up compared with women reporting regular menses, with greater risk among women with higher BMI.

“Regarding coronary heart disease, we found close to a 50% increase in risk for coronary heart disease among these women during follow-up,” Manson said.

In the Study of Women’s Health Across the Nation (SWAN), which included premenopausal and perimenopausal women who provided blood samples to assess hormone levels and other biomarkers, researchers demonstrated that higher levels of androgens and lower levels of sex hormone-binding globulin were associated with adverse cardiometabolic status, such as hypertension, dyslipidemia and abnormal thrombotic biomarkers, Manson said. Similarly, an analysis of postmenopausal women who participated in the Multi-Ethnic Study of Atherosclerosis (MESA) showed that a more androgenic hormone profile of higher free testosterone and lower SHBG was associated with a greater coronary artery calcium progression over 10 years.

A study published in The BMJ in 2020 conducted by Manson and colleagues assessed the Nurses’ Health Study II cohort through 2017 — close to 30-years of follow-up — and found that irregular and long menstrual cycles in adolescence and early to mid-adulthood were associated with a greater risk for premature mortality. Among women who reported irregular cycles at age 29 to 46 years, the risk for premature mortality was slightly higher for CVD mortality than for cancer mortality or death from other causes.

“Although menstrual cycle irregularity partially captures women more likely to have PCOS, there are other causes of irregular cycles,” Manson said.

Meta-analyses of studies including women with a PCOS diagnosis suggest varying degrees of CV risk. In a systematic review and meta-analysis of 23 cohort studies published in 2020, women with PCOS had increased risks for hypertension (RR = 1.75; 95% CI, 1.42-2.15), type 2 diabetes (RR = 3; 95% CI, 2.56-3.51), high total cholesterol (mean difference, 7.14 mg/dL; 95% CI, 1.58-12.7) and nonfatal cerebrovascular disease events (RR = 1.41; 95% CI, 1.02-1.94) compared with women without PCOS. However, differences for long-term coronary disease events were less clear.

“I would say the evidence from observational studies has been inconsistent, though overall, the research suggests an increased risk for cardiometabolic abnormalities and CVD among women with PCOS,” Manson said.

Evidence that definitively links PCOS to hard clinical CVD endpoints is lacking. Studies that report on irregular menses, while compelling, do not reflect the full phenotype of PCOS, whereas the definition of CV “events” can vary by study, according to Andrea Dunaif, MD, professor and chief of the Hilda and J. Lester Gabrilove division of endocrinology, diabetes and bone disease at the Mount Sinai Health System and Healio | Endocrine Today Co-editor.

“The general feeling in the field is that, although we have all this smoke, we still have not definitively answered the question: Is there increase in CVD?” Dunaif, also a workshop co-chair, told Endocrine Today. “No one has shown conclusively that women with PCOS have an increased event rate, and by event rate we mean myocardial infarction, coronary artery bypass graft surgery, angina, stroke. That is because no one has studied these women to an age when these events become more common, their 60s and 70s. The coronary artery calcification studies are problematic in that they have been confounded by obesity or lacked adequate statistical power.”

Addressing risk factors

Dunaif, who has argued for the design of a robust, long-term prospective study that would follow women with PCOS for decades, said a reverse question regarding CV risk and PCOS is similarly intriguing.

Andrea Dunaif

“If you flip it around and say, wait a second, here are these women with PCOS and this tremendous risk. Why aren’t they dropping like flies? Why isn’t this obvious?” Dunaif said. “Is there something protective about PCOS? No matter how you frame it, it is interesting.”

Most PCOS guidelines recommend that clinicians focus on specific symptom manifestations and tailor treatment to the individual. According to the guideline from the International PCOS Network published in 2018, global CVD risk should be routinely assessed for women with PCOS. Fasting blood glucose should be assessed at the initial visit and at subsequent follow-up visits depending on the presence of risk factors for diabetes. Individuals with CVD risk factors should be screened by oral glucose testing. Additionally, lipid profiles, blood pressure and weight should be monitored annually. When CVD risk is uncertain, a coronary artery calcium score for women aged at least 40 years may help refine risk assessment.

For symptoms like menstrual irregularity and prediabetes, pharmacotherapy for PCOS typically includes oral contraceptives and metformin, respectively. Several small studies have demonstrated significant weight loss and testosterone reduction with GLP-1 receptor agonists.

“The thing I am most excited about is the GLP-1 receptor agonists,” said Cree-Green, who is conducting a study on the use of oral semaglutide (Rybelsus, Novo Nordisk) in adolescents with PCOS. “We know weight loss in and of itself helps women with PCOS. The question is do these drugs have a separate mechanism? In my patients with PCOS and obesity, I am prescribing them already.”

Data from the STEP clinical trial program assessing injectable semaglutide 2.4 mg (Wegovy) have been referred to as game changing for obesity management. In addition to significant and sustained weight loss, many participants assigned the drug also experienced a marked reduction in HbA1c. Several agents in the class have also shown CV benefit.

“For women with PCOS who develop diabetes, we need really aggressive therapy with medications known to decrease CVD risk, such as GLP-1 receptor agonists and SGLT2 inhibitors,” Cree-Green said. “Not metformin. Not insulin. We need to use drugs with known CV benefit. Unfortunately, none of these studies have been done.”

Multidisciplinary care that includes cognitive behavioral therapy is also important for women with PCOS to address other often neglected comorbidities, such as mental health concerns and poor quality of life, which can also influence CVD risk, according to Anuja Dokras, MD, PhD, professor of obstetrics and gynecology and director of the Penn PCOS Center at the University of Pennsylvania Perelman School of Medicine.

“The patient dissatisfaction is so high,” Dokras, also a co-chair at the NIH workshop, told Endocrine Today. “When we conducted patient surveys, one of their frustrations was issues of long-term risk are not even brought up. If you have a dissatisfied patient, they are not going to make, for example, dietary changes to improve lipids. If you are only prescribing an oral contraceptive pill and not conducting a metabolic screening, you are doing these women a disservice. If we are going to prevent CVD, you must treat the whole person. If you are too depressed to exercise, it is not going to help.

“We have a window of opportunity,” Dokras said. “If we see these patients in their 20s and 30s, we are ahead of the curve vs. when patients present with diabetes or heart disease when they are older. This is a good time to practice prevention rather than therapy, individualizing treatment.”

‘It cannot be a convenience study’

When it comes to assessing long-term CV risk, major gaps in PCOS literature include poor or inconsistent classification of PCOS and rare follow-up into later adulthood. However, there are also more basic issues that can complicate research.

“There are some very basic federal government things that have not happened,” Cree-Green said. “PCOS is not a condition listed in the NIH RePORTER database. You cannot query for PCOS grants. You cannot pull PCOS status out of [National Health and Nutrition Examination Survey] data.”

Experts are now brainstorming ways to boost interest in PCOS research and exploring ways to mine potentially valuable data from older, established cohorts that may provide some answers regarding CV risk.

Anuja Dokras

“We talked about adding deeper phenotyping to existing cohorts,” Bairey Merz said. “That is always hard because the existing cohort is already there, but some of it could be done. We have studies like the Women’s Health Initiative or the statin trials. Many of these big cohorts have existing blood samples. That is low-hanging fruit.”

Manson said some of this work could be done efficiently in a prospective nested, case-control design.

“You have these [CV] endpoints accruing over 20, 30 or 40 years of follow-up,” Manson said. “You can then identify these people who had these CV events and take a sample of those without the conditions. Look at their baseline biomarkers when they were in early adulthood. You could cost-efficiently look at the long-term sequelae, based on those baseline biomarkers, as a starting point in filling the knowledge gaps.”

The idea is creative, Dunaif said, but likely not enough to provide clear answers.

“To do that, you must have the outcomes of the people well tracked,” Dunaif said. “We know the diagnosis codes for PCOS are not great. For the women who do have the diagnosis, whether they even have the disease is debatable. Then, so many women with PCOS do not receive those diagnoses. You don’t know who you’re missing. Why not spend 20 years doing the good study? With an international effort, you could put together a cohort of 5,000 to 6,000 women with PCOS and a control group. It cannot be a convenience study. Otherwise, anybody who wanted to critique your outcome could. You do not want that.”

Cree-Green, who agreed such a study is needed, said she hoped the NIH workshop will drive such an effort, along with more research funding.

“A $2.5 million, 5-year standard R01 [NIH research grant] may not be enough,” Cree-Green said. “We need longer studies. We also want to reflect the diverse population of PCOS — Rotterdam criteria vs. NIH criteria, including all different ethnicities. How does one account for oral contraceptive treatment vs. nonhormonal treatment? We need a big cohort. The standard mechanisms may not be what we need to move this field forward.”

In the meantime, clinicians should screen for CV risk factors and aggressively treat women at higher risk, she said.

“We don’t know the long-term risk for CVD, but we do know that we can immediately impact the risk factors,” Cree-Green said. “For now, we must focus on that until we have those answers.”

Click here to read the point/counter, "Could hyperandrogenism in PCOS be cardioprotective?"