Diagnosis of Polycystic Ovary Syndrome

Special Report

By Leon Speroff, MD, Editor, Professor of Obstetrics and Gynecology, Oregon Health and Science University, Portland, is Editor for OB/GYN Clinical Alert.

There have been 2 attempts to produce a consensus definition of the polycystic ovary syndrome by expert conferences. The NIH-sponsored conference in 1990 was highlighted by the difficulty in achieving consensus; nevertheless the experts concluded that hyperandrogenism after the exclusion of other etiologies (such as adrenal disorders) and menstrual dysfunction were required for diagnosis, and that ultrasound diagnosis of polycystic ovaries was controversial. In 2003, the European Society for Human Reproduction and Embryology (ESHRE) and the American Society for Reproductive Medicine (ASRM) organized an expert conference that produced the Rotterdam criteria: the diagnosis of polycystic ovary syndrome could be diagnosed if 2 of the following 3 features are present—1) oligo- or anovulation; 2) clinical and/or biochemical signs of hyperandrogenism; and 3) polycystic ovaries established by ultrasonography. Thus according to the Rotterdam criteria, the diagnosis could be made without ovulatory dysfunction or without hyperandrogenism.

Ricardo Azziz from UCLA editorially considers the Rotterdam criteria to be a problem.1 He believes that assigning this diagnosis to ovulatory women or to anovulatory women without hyperandrogenism can have negative consequences. The use of the Rotterdam definition to establish inclusion and exclusion criteria in clinical studies may make it difficult to establish common abnormalities and future risks associated with this syndrome. Importantly, the Rotterdam criteria imply to both clinicians and patients that ovulatory women or anovulatory women without hyperandrogenism have future metabolic and cardiovascular risks, something we do not know. Azziz also raises the disturbing thought that assigning this diagnosis according to the Rotterdam criteria can impact the insurability of some women.

Stephen Franks from the Imperial College and Hammersmith Hospital in London argues that the Rotterdam criteria are important and helpful.2 Franks believes that the presence of polycystic ovaries on ultrasound in ovulating women with hyperandrogenism confirms the diagnosis of polycystic ovary syndrome, and that ovulatory women with polycystic ovaries and hyperandrogenism should be included in the definition of the syndrome. More problematic is the woman with polycystic ovaries and anovulation without hyperandrogenism. Franks contends that such patients can be excluded by establishing the causes of anovulation. He further asserts that the broader definition and application of the Rotterdam criteria are important for the clinical management of these patients.


The discussions and disagreements regarding the polycystic ovary syndrome always remind me of Justice Potter Stewart , who upon hearing a case of pornography at the US Supreme Court, said he couldn’t define pornography, but "I sure recognize it when I see it." For most, if not all, gynecologists, the same can be said regarding polycystic ovary syndrome. It is a clinical diagnosis easily made in nearly all patients who present with this problem.

It seems to me that there are two basic reasons for confusion and disagreement. First is the role assigned to the presence of polycystic ovaries demonstrated by ultrasonography, and second is the constant search of authors and investigators to find a single cause for this clinical syndrome.

A clear difference exists between North American and European clinicians and investigators regarding the sonographic appearance of the ovaries (the ultrasound demonstration of polycystic ovaries is commonly a diagnostic feature of the polycystic ovary syndrome in Europe). The emphasis on ovarian morphology in the Rotterdam criteria reflects this European standard.

It has been long established that polycystic ovaries can be found by ultrasound in a substantial percentage of normal (ovulating), reproductive age women. The critical question is whether these women have metabolic consequences and risks. The few studies exploring this question have found absent or small biochemical differences compared with controls. In my view, if there are no clinical disruptions of normal function and no long-term risks, then the presence of polycystic ovaries is meaningless.

Recognizing that polycystic ovaries are not etiologic but a histologic manifestation seen with many clinical states easily solves this clinical dilemma. For this reason, the presence or absence of polycystic ovaries is not an essential element in the diagnosis of polycystic ovary syndrome, and ultrasonography is not a required part of the diagnostic process. Indeed, there is no single biochemical or genetic diagnostic marker. This syndrome remains a clinical diagnosis, recognizing that women with ovulatory dysfunction and hyperandrogenism have the clinical problems of infertility and hirsutism and increased risks of diabetes mellitus and cardiovascular disease. The syndrome reflects a broad spectrum of patients with multiple causes, ranging from simple weight gain to adrenal disease, hyperprolactinemia, thyroid disease, and androgen-producing tumors. Even ovulatory dysfunction can be evasive, moving back and forth in individual patients from regular menses and ovulation to oligo-ovulation to anovulation.

Is too much fuss being made over this issue? From a clinical point of view, the manifestations of this syndrome in an individual patient reflect the position of that patient in the broad spectrum of ovulatory, endocrine, and metabolic dysfunction determined by the specific cause and how long it has been present. But this is precisely why it has been so confusing for investigators. The collection of patients is large and diverse, a situation making precise definition and study easier said than done. This debate requires future studies to determine whether women with polycystic ovaries on ultrasonography but without hyperandrogenism and anovulation, have increased risks for metabolic and cardiovascular diseases. In the meantime, for most patients, the clinical presentation of the patient is sufficient for clinicians to provide effective and appropriate counseling and therapy.


  1. Azziz R. Controversy in clinical endocrinology: diagnosis of polycystic ovarian syndrome: the Rotterdam criteria are premature. J Clin Endocrinol Metab. 2006;91:781-785.
  2. Franks S. Controversy in clinical endocrinology: diagnosis of polycystic ovarian syndrome: in defense of the Rotterdam criteria. J Clin Endocrinol Metab. 2006;91:786-789.