Received December 11, 2005; Accepted December 11, 2005.
Author to whom all correspondence and reprint requests should be addressed:
F. J. Broekmans, MD, PhD, Department of Reproductive Medicine, Uni-
versity Medical Center, Heidelberglaan 100, 3584 CX Utrecht, The Nether-
lands. E-mail: f.broekmans@azu.nl
Diagnostic Criteria for Polycystic Ovarian Syndrome
F. J. Broekmans and B. C. J. M. Fauser
Department of Reproductive Medicine and Gynecology, University Medical Center, Utrecht,
Heidelberglaan 100, 3584 CX Utrecht
Endocrine, vol. 30, no. 1, 3–11, August 2006 0969–711X/06/30:3–11/$30.00
ENDO (Online) ISSN 1559-0100 © 2006 by Humana Press Inc. All rights of any nature whatsoever reserved.
3
lengthy debates among clinicians (6,7). Specialty groups
may still differ in their use of diagnostic criteria and diag-
nostic work up, as well in their choice of first- and second-
line treatment (7).
At the expert consensus meeting held in 2003 in Rotter-
dam (3,8), it was agreed that the polycystic ovary syn-
drome should be diagnosed in cases of WHO II anovulation
if at least two of the following three features were present:
(1) oligomenorrhea or amenorrhea, (2) clinical or biochem-
ical hyperandrogenism, and (3) polycystic ovaries at ultra-
sound. The approach of only requiring a certain number of
relevant features to diagnose a disease state or clinical con-
dition is not without precedent. For instance, for the diag-
nosis of the metabolic syndrome it was decided that only
three out of five key criteria should be present (9,10). Several
commentaries have since emerged continuing the debate
on the definition and classification issue and have suggested
adaptations in the set of diagnostic criteria (11–13). The
Rotterdam consensus, however, is indeed a compromise,
which bridges the immense gap between the European and
American perspectives with regard to PCOS. It is anticipated
that with the use of consensual diagnostic criteria among
physicians working in the field of both endocrinology and
gynecology the comparability of published research will
increase and the search for etiological factors will lead to
improved results of patient management. It is therefore
recommended that all clinicians and investigators now use
this internationally agreed definition to ensure uniformity
in research studies and routine clinical management.
In this section much of the background for the use of the
so-called Rotterdam criteria will be discussed. Also, asso-
ciated features will be addressed, such as obesity, elevated
serum LH, insulin resistance, and the metabolic syndrome.
Those features are frequently present in PCO patients with-
out contributing to the diagnosis per se, and may well have
consequences for long-term health and as such should be
recognized. Also, risks for cancer development based on
unopposed estrogen exposure are briefly discussed. Finally,
the necessity for the exclusion of other explaining causes
for the phenotypic features will be addressed and a diagnos-
tic work up scheme presented.
In the original description of the syndrome, the histolog-
ical confirmed polycystic nature of ovarian architecture was
the primary abnormality. The association between endocrine
Until recently no universally accepted clinical defini-
tion existed for the polycystic ovary syndrome (PCOS).
What has emerged from research over the last 30 yr
is a profound heterogeneity and ongoing speculation
regarding etiology. The various symptoms and signs
related to PCOS have now been extensively evaluated
as to their possible contribution to the diagnosis. Con-
sensus has been reached for the use of oligomenorrhea
or amenorrhea, clinical or biochemical hyperandrog-
enism, and polycystic ovaries at ultrasound as key diag-
nostic criteria. Obesity, insulin resistance, and the
so-called metabolic syndrome should be recognized
as associated conditions that present long-term health
risks for diagnosed PCOS cases. The way all these fea-
tures need to be applied in the work up of the indi-
vidual index patient is reviewed here.
Key Words: Polycystic ovary syndrome; consensus;
ultrasonography; hyperandrogenism; anovulation.
Introduction
The history of the diagnostic challenge of a condition
currently referred to as the polycystic ovary syndrome
(PCOS) goes back some 70 yr. It was in 1935 that Stein and
Leventhal described several cases presenting with oligo-
menorrhea/amenorrhea combined with the presence at
operation of bilateral polycystic ovaries (PCO) (1). Of these
seven patients, three also presented with obesity and five
showed signs of hirsutism. Only one patient was both obese
and showed hirsutism. These findings indicate that in cases
with polycystic ovaries proven by morphology not all of
the features associated with the PCOS necessarily have to
be present (2,3). With the introduction of transvaginal ultra-
sonography it also became clear that patients with oligo-
menorrhea, obesity, and hirsutism do not necessarily have
the typical polycystic morphology on ultrasound (4,5). More-
over, as the etiology of PCOS is far from well understood,
diagnostic criteria for PCO syndrome have been subject of