Best Practice & Research Clinical Endocrinology & Metabolism

Vol. 20, No. 2, pp. 193–205, 2006

doi:10.1016/j.beem.2006.02.005
available online at http://www.sciencedirect.com

Diagnosis, epidemiology, and genetics of the

polycystic ovary syndrome

Mark O. Goodarzi MD, PhD

Assistant Professor of Medicine
Division of Endocrinology, Diabetes and Metabolism, Department of Medicine, Medical Genetics Institute, Los Angeles,
CA, USA
Department of Obstetrics and Gynecology, Cedars-Sinai Medical Center, Los Angeles, CA, USA
Department of Medicine, The David Geffen School of Medicine at UCLA, Los Angeles, CA, USA

Ricardo Azziz* MD, MBA, MPH

Chair and Professor of Obstetrics and Gynecology and Medicine
Department of Obstetrics and Gynecology, Center for Androgen Related Disorders, Cedars-Sinai Medical Center, 8635
West Third Street, Suite 160W, Los Angeles, CA 90048, USA
Department of Medicine, and Obstetrics and Gynecology, The David Geffen School of Medicine at UCLA, Los Angeles,
CA, USA

Polycystic ovary syndrome (PCOS) is a heterogeneous disorder, and its definition remains fluid
and controversial. PCOS is characterized by clinical and/or biochemical hyperandrogenism, and is
frequently accompanied by ovulatory dysfunction and polycystic ovaries. PCOS is a diagnosis of
exclusion, with other androgen excess and related disorders to be excluded. The prevalence of
PCOS is 6.5–8.0% of unselected women of reproductive age, using the NIH 1990 criteria.
Genetically, PCOS is a common, complex disorder. Despite repeated attempts to identify the
putative gene or genes responsible for this disorder, the PCOS gene(s) remain elusive.

Key words: epidemiology; polycystic ovary syndrome; phenotyping; complex disorders.

The polycystic ovary syndrome (PCOS) was first described by Stein and Leventhal in
19351, and has since been recognized as one of the most common endocrine disorders
of women2 and the most frequent cause of oligo-ovulatory infertility.3 PCOS is a
heterogeneous disorder, the principal features of which include androgen excess,
ovulatory dysfunction, and/or polycystic ovaries. PCOS appears to be inherited as a

* Corresponding author. Tel.: C1 310 423 7433; fax: C1 310 423 3470.
E-mail address: azzizr@cshs.org (R. Azziz) ricardo.azziz@cshs.org.

1521-690X/$ - see front matter Q 2006 Elsevier Ltd. All rights reserved.
194 M. O. Goodarzi and R. Azziz

complex genetic trait.4 In this chapter, we review the current state of the diagnosis,
epidemiology, and genetics of PCOS.

DIAGNOSIS OF PCOS

The definition and hence the diagnosis of PCOS remain controversial, engendering
passionate debates in most scientific forums. The disorder is heterogeneous; of the seven
patients with bilateral polycystic ovaries and oligo/amenorrhea originally reported by
Stein and Leventhal, four were obese, five were hirsute (one with obesity), and one was
thin and acneic.1 Nonetheless, currently we recognize that a relatively uniform
characteristic of PCOS appears to be androgen excess.5 Hyperandrogenism is detectable
either by laboratory analysis to seek evidence of elevated androgen levels or production,
or by clinical examination, primarily to search for the presence of hirsutism. Two other
features often considered fundamental to the diagnosis include: (a) ovulatory
dysfunction, often but not always associated with overt menstrual dysfunction, which
affects 80–100% of patients, and (b) polycystic ovarian morphology, generally detectable
by ultrasonography, and present in 70–95% of these women.
PCOS is a diagnosis of exclusion, such that other androgen excess or ovulatory
disorders with clearly defined etiologies are to be excluded. These include 21-hydroxylase-
deficient non-classic adrenal hyperplasia (NCAH), an autosomal recessive disorder
secondary to mutations or defects of the gene CYP21 affecting 1–10% of hirsute women,
depending on ethnicity; disorders resulting in Cushing’s syndrome with hyperandrogenic
features, such as adrenocortical carcinomas and Cushing’s disease, which affect !1% of
subjects; ovarian and other adrenal androgen-secreting neoplasms, present in 1:300–1000
androgenized women; and exogenous anabolic drugs, the prevalence of which is unknown,
but which generally result in only minor androgenization. Although controversial, many
investigators also exclude patients with the hyperandrogenic insulin-resistant acanthosis
nigricans (HAIRAN) syndrome, since these patients will have degrees of hyperinsulinism
and insulin resistance far greater than those in the vast majority of PCOS patients, as well as
other unique features, including lipodystrophy.
Many investigators also exclude disorders that may result in ovulatory dysfunction, if
present, such as thyroid dysfunction and hyperprolactinemia. Nonetheless, present-day
studies indicate that the prevalence of these endocrine abnormalities in patients with
apparent PCOS is relatively low, on the order of 1–3%. Likewise, investigators will
generally also exclude those disorders that can result in polycystic-like ovarian
morphology, including hypothalamic amenorrhea.
A number of other features have been reported to be present in women with PCOS,
including obesity (in 30–60% of patients, depending on country of origin)6–9, insulin
resistance and hyperinsulinism (present in 50–70%)10–14, and an LH:FSH ratio of O2 or
3 (in 30–50%).15–20 However, these features have not generally been included in
proposed diagnostic criteria, as they either are highly prevalent in disorders other than
PCOS (e.g. obesity and insulin resistance) or are not observable in the majority of
patients with routine laboratory assessments (e.g. an elevated LH:FSH ratio, when LH
levels are lower in obese individuals).
To date two major criteria have been proposed, with other investigators proposing
modifications of these. In the following sections we will review the criteria arrived at
by an expert conference sponsored by National Institutes of Health/National Institute
of Child Health and Human Development (NIH/NICHD) in April 1990, and another
 Diagnosis, epidemiology, and genetics of PCOS 195

co-sponsored by the European Society for Human Reproduction and Embryology

(ESHRE) and the American Society for Reproductive Medicine (ASRM) in May of 2003,
as well as two modifications of these.

The 1990 NIH criteria

The definition of PCOS most commonly used today arose from the proceedings of an
expert conference sponsored by the NIH in April 1990 (Table 1). A survey of all
participants concluded that most believed that the features of PCOS were (in order of
importance): (1) hyperandrogenism and/or hyperandrogenemia; (2) chronic anovula-
tion; and (3) exclusion of related disorders such as hyperprolactinemia, thyroid
disorders, and congenital adrenal hyperplasia.21 Conference participants recognized
that while polycystic ovaries were often observed, they felt that this finding was
suggestive, but not diagnostic, of the syndrome. However, participants in the
conference were overwhelmingly from the US.

Table 1. Proposed criteria for the definition of PCOS.

21
NIH, 1990
To include all of the following
Hyperandrogenism and/or hyperandrogenemia
Oligo-ovulation
Exclusion of related disorders
Homburg, 200222
Women with one or more of the following should undergo ovarian sonography
Menstrual disturbance
Hirsutism
Acne
Anovulatory infertility.
If sonography demonstrates polycystic ovarian morphology, then the diagnosis of PCOS is confirmed. If
not then additional biochemical tests are done, and if any of the following is found to be abnormal then the
diagnosis of PCOS is madea
Elevated testosterone
Elevated androgen index
Elevated LH levels
Subnormal glucose: insulin ratio (!4.5)
ESHRE/ASRM (Rotterdam), 200323,24
To include two of the following, in addition to exclusion of related disorders
Oligo- or anovulation
Clinical and/or biochemical signs of hyperandrogenism
Polycystic ovaries
Modified NIH criteria25
To include all of the following
Androgen excess (clinical and/or biochemical hyperandrogenism)
Ovarian dysfunction (oligo-anovulation and/or polycystic ovarian morphology)
Exclusion of other androgen excess or ovulatory disorders
a
Presumably, related disorders were to be excluded.
196 M. O. Goodarzi and R. Azziz

Homburg 2002 proposal

Homburg’s proposal was put forward in 200222 as a bridge between the US definition
(NIH, 1990) and that of the European community who emphasized the inclusion of
polycystic ovaries in the diagnostic scheme (Table 1). Under this two-step proposal,
patients with suspected PCOS and presenting with any of the following undergo a
transvaginal ultrasonography: (1) menstrual disturbance; (2) hirsutism; (3) acne; and/or
(4) anovulatory infertility. If polycystic ovarian morphology is demonstrated then the
diagnosis of PCOS is confirmed. If the ovaries do not meet the criteria for polycystic
ovaries, then additional biochemical tests are done, and if any of the following are found
to be abnormal then the diagnosis of PCOS is made: (1) elevated testosterone; (2)
elevated androgen index; (3) elevated LH levels; and/or (4) a subnormal glucose:insulin
ratio (!4.5). Presumably, related disorders are excluded.

The 2003 ESHRE/ASRM (Rotterdam) criteria

In response to the lack of emphasis on the inclusion of polycystic ovaries as part of the
phenotype of PCOS in the US, another expert conference was organized in Rotterdam
in May of 2003 (Table 1). In contrast to the NIH meeting, the majority of participants of
this conference were European. The proceedings of the conference noted that PCOS
could be diagnosed by having two of the following three features: (a) oligo- or
anovulation; (b) clinical and/or biochemical signs of hyperandrogenism; and (c)
polycystic ovaries, after the exclusion of related disorders.23,24 These recommen-
dations did not eliminate the NIH 1990 criteria, rather they expanded the definition of
PCOS.23 Additional phenotypes now to be considered as PCOS included: (1) women
with polycystic ovaries and ovulatory dysfunction but no signs of androgen excess; and
(2) women with polycystic ovaries with clinical and/or biochemical evidence of
androgen excess, but no signs of ovulatory dysfunction. Whether these additional
phenotypes do actually represent PCOS remains to be determined.25
Ovulatory, hyperandrogenic patients with polycystic-appearing ovaries have features
that approximate patients with PCOS defined by the NIH 1990 criteria, including mild
degrees of hyperinsulinism and LH excess.26–28 Alternatively, it is much less clear that
those patients with ovulatory dysfunction and polycystic ovaries, but without any
evidence of hyperandrogenism, as a group have a similar morbidity to patients with
PCOS (e.g. insulin resistance). Furthermore, it is uncertain how patients with
hypothalamic amenorrhea (HA) and polycystic ovaries will be differentiated from non-
obese women with PCOS, considering the significant differences in long-term
morbidity (e.g. bone loss for HA and increased bone mass for PCOS, increased risk
of diabetes in PCOS but not in HA patients).

Modified NIH criteria proposal

A modification of the NIH criteria was proposed to accommodate the increasing and
appropriate need to consider the ovarian morphology in the definition of PCOS while
taking into account what is currently known of the disorder (Table 1). This definition then
considered PCOS to be defined by three features: (1) androgen excess (clinical and/or
biochemical hyperandrogenism); (2) ovarian dysfunction (oligo-anovulation and/or
polycystic ovarian morphology); and (3) exclusion of other androgen excess or ovulatory
disorders.25 This definition then recognizes one additional phenotype above that noted
 Diagnosis, epidemiology, and genetics of PCOS 197

by the NIH 1990 criteria, namely those women with polycystic ovaries, hyperandrogen-
ism, and apparently normal ovulation. This investigator certainly does not discard the
possibility that future research may demonstrate that a subset of women with polycystic
ovaries, ovulatory dysfunction, and without overt androgen excess may actually have
PCOS. However, this should await better data, as the inclusion of this phenotype may
have significant negative implications for research (increasing the heterogeneity of study
populations), clinical practice (requiring that all these subjects undergo ultrasonography),
and patient well-being (requiring long-term follow-up for the development of associated
metabolic morbidities and potentially affecting health insurability).

EPIDEMIOLOGY OF PCOS

Clearly the prevalence of PCOS will depend to a degree on the criteria used to define
this disorder. The prevalence of PCOS has been determined in various populations,
primarily of White or Caucasian and, in one study, of Black races. In a study of 277
women seeking a pre-employment physical in a university in the southeastern US, we
initially reported an overall prevalence of PCOS diagnosed by the NIH 1990 criteria of
4.0%, with no significant difference between Whites and Blacks.29 In a subsequent and
more intensive study of 400 unselected consecutive women aged 18–45 years in the
same setting (223 Black, 166 White, and 11 of other races), the prevalence of PCOS
was observed to be 6.6%, and still not significantly different between Blacks and Whites
(8.0 and 4.8%, respectively).2
Also using the 1990 NIH criteria, a study of 192 Greek women on the island of
Lesbos, recruited by the promise of free medical examinations, reported a prevalence
of PCOS of 6.8%.30 Another study of 154 Caucasian women in Madrid, Spain, seen as
blood donors, found a similar prevalence (6.5%).31 Likewise, among 230 volunteers
(97% White) recruited from two Oxford universities and two general practice
surgeries, who agreed to participate in ‘a study of women’s health issues’, the
prevalence of PCOS using the NIH 1990 criteria was 8%.32
Consequently, the prevalence of clinically evident PCOS in unselected women of
reproductive age ranges from 6.5 to 8.0% using the 1990 NIH criteria. Since there are
approximately 62 million women aged 15–44 years in the US33, we can estimate that at
least 4–5 million women of reproductive age are affected.
Less clear is the prevalence of PCOS diagnosed by more expansive definitions,
such as the Rotterdam 2003 criteria, although logically one would expect a
prevalence higher than the 6.5–8.0% noted above. In their study, Michelmore and
colleagues defined PCOS by the presence of polycystic ovaries on ultrasound plus
one additional feature, including menstrual irregularity, acne, hirsutism, body mass
index (BMI)O25 kg/m2, raised serum testosterone, or raised LH (O10 IU/L).32
Applying these criteria, 26% of the total 224 women undergoing a transabdominal
sonography had evidence of PCOS, compared to only 8% if the NIH 1990 criteria
were used. However, these features occurred frequently in women without
polycystic ovaries, and 112 of the 150 women (75%) with normal ovaries had the
presence of one or more of these attributes. Consequently it is possible that the
number of women diagnosable as having PCOS, and the prevalence of the disorder,
would double or even triple if the Rotterdam criteria were used. Further studies
comparing the prevalence of PCOS using the NIH 1990 and the Rotterdam 2003
criteria are needed.
198 M. O. Goodarzi and R. Azziz

Several conditions are associated with increased prevalence of PCOS,

including obesity34,35, the presence of insulin resistance35, type I or type II
diabetes mellitus36–38, or oligo-ovulatory infertility.3,39 The prevalence of PCOS also
appears to be higher among Mexican American than White or African American
women.40

GENETICS OF PCOS

PCOS is a common, complex genetic disorder. Common diseases such as

schizophrenia, asthma, and type 2 diabetes, as well as PCOS, have a complex,
multifactorial etiology, in which a variety of predisposing genes, not just one
gene, interact with environmental factors to produce disease.41 Studies in families
have demonstrated the heritable nature of PCOS itself as well as the component
phenotypes of PCOS. Subsequently, many population studies have attempted to
discover genes that influence PCOS using the candidate gene approach.

Heritability of PCOS and its component phenotypes

Family studies demonstrate that PCOS is significantly more prevalent among family
members than in the general population.42 Among first-degree female relatives (on no
hormonal therapy) of 93 patients with PCOS, 35% of premenopausal mothers and 40%
of sisters were also affected with the disorder.43 These affection rates are significantly
higher than the 6–8% observed in the general population.2 In another study, 115 sisters
of 80 women with PCOS were evaluated; of these, 22% met criteria for PCOS.44 An
additional 24% of these sisters had hyperandrogenemia with normal menses. Total and
free testosterone levels were similar between the sisters with hyperandrogenemia only
and the sisters and probands with PCOS. A bimodal distribution of testosterone levels
in the sisters of women with PCOS was observed, suggesting a major genetic
component to hyperandrogenemia.44 Brothers of women with PCOS also display
abnormal androgens: a study of such brothers found them to have elevated levels of
dehydroepiandrosterone sulfate.45
Not only is PCOS itself a heritable condition, but also within PCOS insulin resistance
and insulin secretion appear to be under significant genetic control. Among sisters of
women with PCOS, those who had PCOS or hyperandrogenemia with regular menses
had lower insulin sensitivity than unaffected sisters, assessed by fasting insulin and glucose
measurements.46 Likewise, in families of Australian patients with PCOS, hyperinsuline-
mia was found to occur in 69% of all family members, suggesting that this trait was
inherited.47 One hundred and two relatives of 52 Turkish women with PCOS underwent
assessment of insulin resistance in the fasting and post-glucose-challenge states;
compared to population controls, mothers, sisters, and brothers of PCOS subjects had
greater insulin resistance.48 Brothers of Indian women with PCOS had insulin resistance
and endothelial dysfunction.49 In studies of families of women with PCOS, insulin
secretion levels, quantified directly by the frequently sampled intravenous glucose
tolerance test, displayed significant heritability, suggesting a genetic component to b-cell
dysfunction in PCOS.50 The most abnormal insulin secretion was observed in women
with PCOS and a history of type 2 diabetes in a first-degree relative.51
 Diagnosis, epidemiology, and genetics of PCOS 199

Cellular defects suggestive of genetic determinants of PCOS

Physiologic evidence that hyperandrogenemia in PCOS is genetically controlled is

supported by the observation that ovarian theca cells removed from women with PCOS
and propagated in culture display persistently elevated testosterone secretion compared
to cells from unaffected women.52 Because removal from the body and multiplication in
culture removes the influence of an abnormal hormonal milieu from these cells, it is likely
that persistent differences from normal represent intrinsic (genetic) defects.
Abnormal responses to insulin in cells that have been removed from women with
PCOS suggest the presence of intrinsic cellular defects. Lower insulin receptor
substrate 1 (IRS-1)-associated phosphatidylinositol 3-kinase (PI3K) activity and higher
IRS-2 content were observed in the myocytes of PCOS patients compared to controls,
in the face of similar amounts of IRS-1 and the p85 subunit of PI3K.53 In PCOS
adipocytes, the maximum glucose uptake stimulated by insulin was found by some
investigators54 to be lower than that in controls. Although others55 found the
maximum response to be normal, these latter investigators found that the sensitivity of
the response to insulin was deficient in PCOS. The amount of GLUT-4 (glucose
transporter 4) in adipocytes was lower in PCOS (on either a membrane protein or cell
surface basis) than in controls.54 Such defects are likely caused by abnormal expression
or function of genes encoding products of the insulin-signaling pathway.

Candidate genes in PCOS: progress and challenges

To date, efforts to identify genes that influence PCOS susceptibility have largely utilized the
candidate gene approach, in which common polymorphic genetic markers—single
nucleotide polymorphisms (SNPs) or microsatellites—within a gene of interest, selected
on the basis of its hypothesized role in the disease, are evaluated to determine whether the
polymorphisms are associated with the phenotype in populations or in families. Candidate
genes studied in PCOS have generally targeted loci regulating four areas: (1) steroid
biosynthesis and action; (2) gonadotropic action; (3) weight and energy regulation; and (4)
insulin action (Table 2). The field, comprising over 100 publications, has been
comprehensively reviewed.56 Most recently, candidate genes that may affect cardiovascular
disease via inflammation, hypercoagulability, or blood pressure have also been examined in
PCOS (Table 2). Several provocative genetic associations with PCOS have been reported
that are slowly starting to illuminate the underlying causes of PCOS.
Despite repeated attempts to identify the putative gene or genes responsible for this
disorder, the PCOS gene(s) remain elusive. Despite many positive results, no gene or
genes has clearly emerged as most important in PCOS, and many positive results have
not been confirmed in subsequent studies. Studies of the genetic etiology of PCOS have
been hampered by various limitations, including: (1) only one or two variants genotyped
in each gene, indicating incomplete coverage of candidate genes, especially larger genes;
(2) incomplete characterization of the phenotype in family members; (3) inability to
assign a PCOS phenotype to prepubertal girls, postmenopausal women, and men; (4)
possible inclusion of patients with NCAH; (5) lack of appropriate controls; (6) unclear
ethnic/racial composition; (7) varying criteria used to diagnose PCOS in different
studies (in part due to the lack of universally accepted diagnostic criteria); and (8) small
numbers of subjects in most studies. Regarding the latter, many of the genes studied in
Table 2 report results on less than 100 women with PCOS. Therefore, it is likely that
200 M. O. Goodarzi and R. Azziz

Table 2. Candidate genes demonstrating some evidence of linkage or association with PCOS or
component traits of PCOS.

Steroid metabolism Insulin secretion and Gonadotropin action Cardiovascular disease

and action
CYP17 (17a-hydroxyl-
ase/17,20-lyase)
CYP11A (cholesterol
side-chain cleavage
enzyme)
AR (androgen
receptor)
CYP21 (21-hydroxyla-
se)a
CYP19 (aromatase)a
SHBG (sex hormone
binding globulin)
H6PD (hexose-6-phos-
phate dehydrogenase)a
a
Positive result in only one report.
action/fuel metabolism and regulation
Insulin Follistatina Paraoxonasea
Insulin receptor LH b-subunit PAI-1 (plasminogen
activator inhibitor 1)
Microsatellite FSH b-subunita IL6 (interleukin-6)
D19S884 (located
1 cm from the insulin
receptor gene)
IRS1 and IRS2 (insulin Dopamine D3 recep- IL6 receptor complexa
receptor substrates 1 tora
and 2)
CAPN10 (calpain-10) FSH receptora Adiponectin
PPARG (peroxisome EPHX (microsomal
proliferator activated epoxide hydrolase)a
receptor g)
Resistina Aldosterone synthata-
sea
IGF2 (insulin-like Tumor necrosis factor
growth factor 2)a receptor 2a
PPP1R3a (glycogen-tar- Matrix metalloprotei-
geting subunit of pro- nase-1a
tein phosphatase 1)
PC-1 (plasma cell Factor Va
membrane glyco-
protein 1)a

many underpowered studies resulted in false-negative reports, and that several small
studies produced false-positive results.
The power issue is particularly relevant to genetics of common diseases. Validated
genetic determinants of type 2 diabetes, such as the Pro12Ala variant of the PPARG gene
and the Glu23Lys variant of the Kir6.2 pancreatic potassium channel, only modestly
alter risk for type 2 diabetes: on the order of 10–20%.57 If genes with similar magnitude
of effect influence PCOS risk, then many of the studies to date were seriously
underpowered and inadequate to detect genetic variants predisposing to PCOS.
Finally, inherent to the candidate gene approach are assumptions regarding the
underlying pathophysiology of PCOS. This is a particular problem in PCOS, as the
underlying causes are still fundamentally unknown. Candidate genes evaluated so far were
selected from pathways affecting single components of PCOS. Genes coding for
transcription factors or signaling pathway components that may globally affect the organs
involved in PCOS (pituitary, ovaries, adrenals, pancreatic b-cells, insulin-responsive tissues)
are unlikely to be selected as candidate genes by this approach. To date, no genome-wide
linkage analysis has been published for PCOS. The main challenge has been to recruit
enough families to conduct a linkage analysis with sufficient power. Investigators are
actively pursuing this goal; genome-wide scans in PCOS are anticipated in the future.
 Diagnosis, epidemiology, and genetics of PCOS 201

SUMMARY

Androgen excess appears to be a relatively uniform characteristic of PCOS, detectable

either by laboratory analysis or by clinical exam, with ovulatory dysfunction and
polycystic ovarian morphology also affecting a large proportion of patients. PCOS is a
diagnosis of exclusion, such that other androgen excess or related disorders are to be
ruled out. The definition of PCOS most commonly used today arose from the
proceedings of an expert conference sponsored by the NIH in April 1990, which noted
the features of PCOS to be (in order of importance): (1) hyperandrogenism and/or
hyperandrogenemia, (2) chronic anovulation, and (3) exclusion of related disorders
such as hyperprolactinemia, thyroid disorders, and congenital adrenal hyperplasia.
Another expert conference was organized in Rotterdam in 2003, which expanded the
NIH 1990 criteria for PCOS, noting that the disorder could be diagnosed by having two
of the following three features: (a) oligo- or anovulation, (b) clinical and/or biochemical
signs of hyperandrogenism, or (c) polycystic ovaries, after the exclusion of related
disorders. This definition created two new phenotypes for PCOS: (1) women with
polycystic ovaries and ovulatory dysfunction but no signs of androgen excess, and (2)
women with polycystic ovaries with clinical and/or biochemical evidence of androgen
excess, but no signs of ovulatory dysfunction. Whether these phenotypes actually
represent PCOS remains to be determined. A modification of the NIH criteria has been
also proposed, which defines PCOS by three features: (1) androgen excess (clinical
and/or biochemical hyperandrogenism); (2) ovarian dysfunction (oligo-anovulation
and/or polycystic ovarian morphology); and (3) exclusion of other androgen excess or
ovulatory disorders. Clearly the prevalence of PCOS will depend to a degree on the
criteria used to define this disorder, although using the NIH 1990 criteria most studies
have observed 6.5–8.0% prevalence in unselected women of reproductive age.
Genetically, PCOS is a common, complex disorder. Current investigational strategy has
relied on candidate gene analysis, although the main challenge has been recruiting
enough families to conduct a linkage analysis with sufficient power. Investigators are
actively pursuing this goal; genome-wide scans in PCOS are anticipated in the future.

Practice points

† PCOS is a diagnosis of exclusion, such that other androgen excess or ovulatory

disorders with clearly defined etiologies are to be excluded
† to date two major diagnostic criteria have been proposed: the NIH 1990 (PCOS
defined as having evidence of biochemical and/or clinical hyperandrogenism and
ovulatory dysfunction, after the exclusion of related or other disorders) and the
Rotterdam 2003 criteria (PCOS defined as two of the following three, after
exclusion of related or other disorders: biochemical and/or clinical hyperan-
drogenism, ovulatory dysfunction, and polycystic ovaries)
† whether the two new phenotypes proposed by the Rotterdam 2003 criteria
actually represent PCOS and/or have similar long-term morbidities remains to
be determined, and a delay in the adoption of these criteria is recommended
† consequently, the prevalence of clinically evident PCOS in unselected women of
reproductive age, using the 1990 NIH criteria, ranges from 6.5 to 8.0%; it may be
higher if more expansive diagnostic criteria are used
202 M. O. Goodarzi and R. Azziz

Research agenda

† whether the two new phenotypes proposed by the Rotterdam 2003 criteria
actually represent PCOS (women with polycystic ovaries and ovulatory
dysfunction but no signs of androgen excess, or women with polycystic ovaries
with clinical and/or biochemical evidence of androgen excess, but no signs of
ovulatory dysfunction) will need to be determined in cross-sectional and
longitudinal studies
† the prevalence of the disorder in different ethnic groups—particularly those
with higher prevalences of insulin resistance, and using different diagnostic
criteria—will need to be investigated
† the features of PCOS diagnosed among unselected women will need to be
carefully assessed, as this will provide the best measure of what the features of
PCOS actually are, without referral bias
† more in-depth studies identifying the molecular defects in PCOS observable at
the tissue level (ovary, muscle, fat, etc) are very much needed, both to guide the
search for the genes responsible and to assist in elucidating new predictors and
therapies
† larger multicenter genetic studies using well-phenotyped patients with PCOS
are needed; particular attention should be given to focusing on the study of
those genes the function of which has been demonstrated to be uniquely
abnormal in PCOS subjects

ACKNOWLEDGEMENTS

Supported in part by NIH grant K24-HD01346-01 (to RA).

REFERENCES

1. Stein IF & Leventhal NL. Amenorrhea associated with bilateral polycystic ovaries. American Journal of
Obstetrics and Gynecology 1935; 29: 181–191.
*2. Azziz R, Woods KS, Reyna R et al. The prevalence and features of the polycystic ovary syndrome in an
unselected population. The Journal of Clinical Endocrinology and Metabolism 2004; 89: 2745–2749.
3. Hull MG. Ovulation failure and induction. Clinical Obstetrics and Gynaecology 1981; 8: 753–785.
4. Legro RS. The genetics of obesity. Lessons for polycystic ovary syndrome. Annals of the New York Academy
of Sciences 2000; 900: 193–202.
5. Azziz R. Androgen excess is the key element in the polycystic ovary syndrome. Fertility and Sterility 2003;
80: 252–254.
6. Legro RS, Kunselman AR & Dunaif A. Prevalence and predictors of dyslipidemia in women with polycystic
ovary syndrome. The American Journal of Medicine 2001; 111: 607–613.
7. Carmina E, Legro RS, Stamets K et al. Difference in body weight between American and Italian women
with polycystic ovary syndrome: influence of the diet. Human Reproduction 2003; 18: 2289–2293.
*8. Azziz R, Sanchez LA, Knochenhauer ES et al. Androgen excess in women: experience with over 1000
consecutive patients. The Journal of Clinical Endocrinology and Metabolism 2004; 89: 453–462.
9. Hahn S, Tan S, Elsenbruch S et al. Clinical and biochemical characterization of women with polycystic
ovary syndrome in north Rhine-Westphalia. Hormone and Metabolic Research 2005; 37: 438–444.
 Diagnosis, epidemiology, and genetics of PCOS 203

*10. Dunaif A, Segal KR, Futterweit W & Dobrjansky A. Profound peripheral insulin resistance, independent of
obesity, in polycystic ovary syndrome. Diabetes 1989; 38: 1165–1174.
*11. Legro RS, Kunselman AR, Dodson WC & Dunaif A. Prevalence and predictors of risk for type 2 diabetes
mellitus and impaired glucose tolerance in polycystic ovary syndrome: a prospective, controlled study in
254 affected women. The Journal of Clinical Endocrinology and Metabolism 1999; 84: 165–169.
12. Carmina E & Lobo RA. Use of fasting blood to assess the prevalence of insulin resistance in women with
polycystic ovary syndrome. Fertility and Sterility 2004; 82: 661–665.
13. DeUgarte CM, Bartolucci AA & Azziz R. Prevalence of insulin resistance in the polycystic ovary syndrome
using the homeostasis model assessment. Fertility and Sterility 2005; 83: 1454–1460.
14. Goodarzi MO & Korenman SG. The importance of insulin resistance in polycystic ovary syndrome. Fertility
and Sterility 2003; 80: 255–258.
15. Rebar R, Judd HL, Yen SSC et al. Characterization of the inappropriate gonadotropin secretion in
polycystic ovary syndrome. Journal of Clinical Investigation 1976; 57: 1320–1329.
16. Conway GS, Honour JW & Jacobs HS. Heterogeneity of the polycystic ovary syndrome: clinical,
endocrine and ultrasound features in 556 patients. Clinical Endocrinology (Oxford) 1989; 30: 459–470.
17. Anttila L, Ding Y-Q, Ruutiainen K et al. Clinical features and circulating gonadotropin, insulin, and
androgen interactions in women with polycystic ovarian disease. Fertility and Sterility 1991; 55: 1057–1061.
18. van Santbrink EJ, Hop WC & Fauser BC. Classification of normogonadotropic infertility: polycystic
ovaries diagnosed by ultrasound versus endocrine characteristics of polycystic ovary syndrome. Fertility
and Sterility 1997; 67: 452–458.
19. Arroyo A, Laughlin GA, Morales AJ & Yen SS. Inappropriate gonadotropin secretion in polycystic ovary
syndrome: influence of adiposity. The Journal of Clinical Endocrinology and Metabolism 1997; 82: 3728–3733.
20. Taylor AE, McCourt B, Martin KA et al. Determinants of abnormal gonadotropin secretion in clinically
defined women with polycystic ovary syndrome. The Journal of Clinical Endocrinology and Metabolism 1997;
82: 2248–2256.
*21. Zawadzki JK & Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach.
In Dunaif A, Givens JR, Haseltine FP & Merriam GR (eds.) Polycystic Ovary Syndrome. Boston: Blackwell
Scientific Publications; 1992, pp. 377–384.
22. Homburg R. What is polycystic ovarian syndrome? A proposal for a consensus on the definition and
diagnosis of polycystic ovarian syndrome. Human Reproduction 2002; 17: 2495–2499.
*23. The rotterdam ESHRE/ASRM-sponsored PCOS consensus workshop group. Revised 2003 consensus on
diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertility and Sterility
2004; 81: 19–25.
24. The Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group. Revised 2003 consensus
on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS). Human
Reproduction 2004; 19: 41–47.
*25. Azziz R. Diagnostic criteria for polycystic ovary syndrome: a reappraisal. Fertility and Sterility 2005; 83:
1343–1346.
26. Carmina E & Lobo RA. Polycystic ovaries in hirsute women with normal menses. The American Journal of
Medicine 2001; 111: 602–606.
27. Eden JA, Place J, Carter GD et al. Is the polycystic ovary a cause of infertility in the ovulatory woman?
Clinical Endocrinology (Oxford) 1989; 30: 77–82.
28. Kousta E, White DM, Cela E et al. The prevalence of polycystic ovaries in women with infertility. Human
Reproduction 1999; 14: 2720–2723.
29. Knochenhauer ES, Key TJ, Kahsar-Miller M et al. Prevalence of the polycystic ovary syndrome in
unselected black and white women of the southeastern United States: a prospective study. The Journal of
Clinical Endocrinology and Metabolism 1998; 83: 3078–3082.
30. Diamanti-Kandarakis E, Kouli CR, Bergiele AT et al. A survey of the polycystic ovary syndrome in the
Greek island of Lesbos: hormonal and metabolic profile. The Journal of Clinical Endocrinology and Metabolism
1999; 84: 4006–4011.
31. Asuncion M, Calvo RM, San Millan JL et al. A prospective study of the prevalence of the polycystic ovary
syndrome in unselected caucasian women from Spain. The Journal of Clinical Endocrinology and Metabolism
2000; 85: 2434–2438.
32. Michelmore KF, Balen AH, Dunger DB & Vessey MP. Polycystic ovaries and associated clinical and
biochemical features in young women. Clinical Endocrinology (Oxford) 1999; 51: 779–786.
 204 M. O. Goodarzi and R. Azziz

33. US census bureau. Census 2000 Summary File 1, QT-P3. http://factfinder.census.gov/servlet/

DatasetMainPageServlet?_programZDEC&_langZen
34. Hartz AJ, Barboriak PN, Wong A et al. The association of obesity with infertility and related menstrual
abnormalities in women. International Journal of Obesity 1979; 3: 57–73.
35. Heinonen S, Korhonen S, Hippelainen M et al. Relationship of the metabolic syndrome and obesity to
polycystic ovary syndrome: a controlled, population-based study. American Journal of Obstetrics and
Gynecology 2001; 184: 289–296.
36. Escobar-Morreale HF, Roldan B, Barrio R et al. High prevalence of the polycystic ovary syndrome and
hirsutism in women with type 1 diabetes mellitus. The Journal of Clinical Endocrinology and Metabolism 2000;
85: 4182–4187.
37. Conn JJ, Jacobs HS & Conway GS. The prevalence of polycystic ovaries in women with type 2 diabetes
mellitus. Clinical Endocrinology (Oxford) 2000; 52: 81–86.
38. Peppard HR, Marfori J, Iuorno MJ & Nestler JE. Prevalence of polycystic ovary syndrome among
premenopausal women with type 2 diabetes. Diabetes Care 2001; 24: 1050–1052.
39. Allen SE, Potter HD & Azziz R. Prevalence of hyperandrogenemia among nonhirsute oligo-ovulatory
women. Fertility and Sterility 1997; 67: 569–572.
40. Goodarzi MO, Quiñones MJ, Azziz R et al. Polycystic ovary syndrome in Mexican–Americans: prevalence
and association with the severity of insulin resistance. Fertility and Sterility 2005; 84: 766–769.
41. King RA, Rotter JI & Motulsky AG. The Genetic Basis of Common Diseases. 2nd edn. New York: Oxford
University Press; 2002.
42. Legro RS & Strauss JF. Molecular progress in infertility: polycystic ovary syndrome. Fertility and Sterility
2002; 78: 569–576.
*43. Kahsar-Miller MD, Nixon C, Boots LR et al. Prevalence of polycystic ovary syndrome (PCOS) in first-
degree relatives of patients with PCOS. Fertility and Sterility 2001; 75: 53–58.
*44. Legro RS, Driscoll D, Strauss III. JF et al. Evidence for a genetic basis for hyperandrogenemia in polycystic
ovary syndrome. Proceedings of the National Academy of Sciences of the USA 1998; 95: 14956–14960.
45. Legro RS, Kunselman AR, Demers L et al. Elevated dehydroepiandrosterone sulfate levels as the
reproductive phenotype in the brothers of women with polycystic ovary syndrome. The Journal of Clinical
Endocrinology and Metabolism 2002; 87: 2134–2138.
46. Legro RS, Bentley-Lewis R, Driscoll D et al. Insulin resistance in the sisters of women with polycystic
ovary syndrome: association with hyperandrogenemia rather than menstrual irregularity. The Journal of
Clinical Endocrinology and Metabolism 2002; 87: 2128–2133.
47. Norman RJ, Masters S & Hague W. Hyperinsulinemia is common in family members of women with
polycystic ovary syndrome. Fertility and Sterility 1996; 66: 942–947.
48. Yildiz BO, Yarali H, Oguz H & Bayraktar M. Glucose intolerance, insulin resistance, and
hyperandrogenemia in first-degree relatives of women with polycystic ovary syndrome. The Journal of
Clinical Endocrinology and Metabolism 2003; 88: 2031–2036.
49. Kaushal R, Parchure N, Bano G et al. Insulin resistance and endothelial dysfunction in the brothers of
Indian subcontinent Asian women with polycystic ovaries. Clinical Endocrinology (Oxford) 2004; 60: 322–
328.
50. Colilla S, Cox NJ & Ehrmann DA. Heritability of insulin secretion and insulin action in women with
polycystic ovary syndrome and their first-degree relatives. The Journal of Clinical Endocrinology and
Metabolism 2001; 86: 2027–2031.
51. Ehrmann DA, Sturis J, Byrne MM et al. Insulin secretory defects in polycystic ovary syndrome.
Relationship to insulin sensitivity and family history of non-insulin-dependent diabetes mellitus. Journal of
Clinical Investigation 1995; 96: 520–527.
52. Nelson VL, Legro RS, Strauss JF & McAllister JM. Augmented androgen production is a stable
steroidogenic phenotype of propagated theca cells from polycystic ovaries. Molecular Endocrinology 1999;
13: 946–957.
53. Dunaif A, Wu X, Lee A & Diamanti-Kandarakis E. Defects in insulin receptor signaling in vivo in the
polycystic ovary syndrome (PCOS). American Journal of Physiology Endocrinology and Metabolism 2001; 281:
E392–E399.
54. Rosenbaum D, Haber RS & Dunaif A. Insulin resistance in polycystic ovary syndrome: decreased
expression of GLUT-4 glucose transporters in adipocytes. American Journal of Physiology 1993; 264: E197–
E202.
 Diagnosis, epidemiology, and genetics of PCOS 205

55. Ciaraldi TP, el-Roeiy A, Madar Z et al. Cellular mechanisms of insulin resistance in polycystic ovarian
syndrome. The Journal of Clinical Endocrinology and Metabolism 1992; 75: 577–583.
*56. Escobar-Morreale HF, Luque-Ramirez M & San Millan JL. The molecular-genetic basis of functional
hyperandrogenism and the polycystic ovary syndrome. Endocrine Reviews 2005; 26: 251–282.
57. Parikh H & Groop L. Candidate genes for type 2 diabetes. Reviews in Endocrine and Metabolic Disorders
2004; 5: 151–176.