EDITOR’S CORNER

The spectrum of androgen excess disorders

Enrico Carmina, M.D.
Department of Clinical Medicine, University of Palermo, Palermo, Italy

A better understanding of the different phenotypes and of their endocrine and metabolic characteristics
permits investigators to distinguish three main androgen excess disorders: classic polycystic ovary syndrome
(PCOS), mild ovulatory PCOS, and idiopathic hyperandrogenism. These androgenic phenotypes differ more
for the severity of the endocrine and metabolic alteration than for the etiopathogenetic mechanisms. The
appearance of a particular androgenic phenotype is determined by a sum of genetic and environmental
factors, but mostly by body weight. (Fertil Steril威 2006;85:1582–5. ©2006 by American Society for
Reproductive Medicine.)
Key Words: Polycystic ovary syndrome, hyperandrogenism, androgen excess, insulin resistance, idiopathic hirsutism

THE CHANGING CLASSIFICATION OF ANDROGEN
EXCESS DISORDERS
Influence of NIH Criteria for the Diagnosis of PCOS
During the last 15 years, a better understanding of the
phenotypic heterogeneity of hyperandrogenic syndromes
has modified the classification and the relative prevalence
of the different androgen excess disorders. Until the
1980s, circulating hormone levels (i.e., an increase of the
LH and/or LH/FSH ratio) and tests of blocking or stimu-
lating adrenal or ovarian androgen secretion were used to
distinguish between the different forms of androgen ex-
cess (1). However, these endocrine parameters showed
little specificity, and in 1990, at a National Institutes of
Health (NIH) meeting, a majority of experts agreed on
making the diagnosis of polycystic ovary syndrome
(PCOS) mostly on the basis of clinical data (i.e., clinical
or biologic hyperandrogenism and chronic anovulation)
(2). The study of the endocrine pattern was considered
important only to exclude a few well-characterized hyperan-
drogenic syndromes (e.g., Cushing’s syndrome, androgen
secreting tumors, nonclassic adrenal enzymatic deficiencies)
(2). Using NIH criteria, PCOS is, by far, the most common
form of androgen excess disorder. In a recent large study,
82% of patients with clinical hyperandrogenism were con-
sidered affected by PCOS, whereas the other most common
Received August 25, 2005; revised and accepted February 6, 2006.
Presented as a lecture at the VIII FLASEF Congress (Federacion Latino-
Americana de Fertilidad e Sterilidad), Acapulco, Mexico, July 21– 4,
2005.
The opinions and commentary expressed in Editor’s Corner articles are
solely those of the author. Publication does not imply endorsement by
the Editor or American Society for Reproductive Medicine.
Reprint requests: Enrico Carmina, M.D., Professor and Head of Endocrine
Unit, Department of Clinical Medicine, University of Palermo, Via delle
Croci 47, 90139 Palermo, Italy (FAX: ⫹390916555995; E-mail:
enricocarmina@libero.it).
androgen disorders (i.e., hyperandrogenism ⫹ ovulatory cy-
cles and idiopathic hirsutism) were diagnosed in only 6.75%
and 4.47% of the patients (3).
Influence of ESHRE/ASRM Criteria for Diagnosis of PCOS
Recently, at a European Society of Human Reproduction and
Embryology/American Society for Reproductive Medicine
(ESHRE/ASRM) meeting, new criteria for making the diag-
nosis of PCOS were proposed (4, 5). After exclusion of the
same well-characterized but uncommon androgen excess
disorders, PCOS may be diagnosed in patients presenting at
least two of these three features: clinical or biologic hyperan-
drogenism, chronic anovulation, polycystic ovaries. There-
fore, the diagnosis of PCOS may be assigned to patients
presenting three different phenotypes:
1. Hyperandrogenism and chronic anovulation
2. Hyperandrogenism and polycystic ovaries but ovulatory
cycles
3. Chronic anovulation and polycystic ovaries but no clini-
cal or biochemical hyperandrogenism
Consequences of the New Guidelines on the Classification
of Androgen Excess Disorders
Excluding the last phenotype (which by definition does not
include hyperandrogenic patients), the new criteria for diag-
nosis of PCOS have determined important consequences on
the classification of the androgen excess disorders. In fact,
hyperandrogenic patients with ovulatory cycles but polycys-
tic ovaries have to be separated from the group of hyperan-
drogenic patients with ovulatory cycles. It means that three
main androgen excess disorders may be distinguished. We
have used the following names to define these three disor-
ders (6, 7):

1582 Fertility and Sterility姞 Vol. 85, No. 6, June 2006 0015-0282/06/$32.00
Copyright ©2006 American Society for Reproductive Medicine, Published by Elsevier Inc. doi:10.1016/j.fertnstert.2006.02.069
1. Classic PCOS FIGURE 1
2. Ovulatory PCOS
3. Idiopathic hyperandrogenism Prevalence (%) of different androgen excess
Classic PCOS includes the patients with hyperandrogenism disorders among patients with clinical
and chronic anovulation (NIH definition of PCOS). Ovula- hyperandrogenism (idiopathic hyperandrogenism
tory PCOS indicates patients who present hyperandrogenism blue, idiopathic hirsutism aqua, nonclassic 21-OH
and ovulatory cycles but polycystic ovaries, whereas idio- deficiency yellow, ovulatory PCOS fuchsia, and
pathic hyperandrogenism regroups hyperandrogenic patients PCOS red). Modified from Carmina et al. (7).
with normal ovulatory cycles and normal ovaries. It is clear
that these names are arbitrary and other experts may use
different definitions. However, other authors have used sim-
ilar names (8), and these different hyperandrogenic syn-
dromes must be distinguished in some way.

New Criteria for Diagnosis of Idiopathic Hirsutism

The ESRHE/ASRM definition of PCOS has influenced even
the criteria to make a diagnosis of idiopathic hirsutism. In
fact, idiopathic hirsutism was previously diagnosed on the
basis of clinical hyperandrogenism but normal ovulatory
cycles and normal serum androgens, regardless of ovarian
morphology (9). However, some of these patients may have
polycystic ovaries and should be now considered affected by
ovulatory PCOS.

Prevalence of the Different Forms of Androgen

Excess Disorders
Using ESHRE/ASRM diagnostic criteria, we have recently
assessed the relative prevalence of the different androgen
excess disorders (7), and we have found that these three
disorders include about 90% of hyperandrogenic patients
(Fig. 1). This prevalence is similar to that reported by Azziz
et al. (3) with the important difference that, in our experi-
ence, the mild androgenic disorders (i.e., ovulatory PCOS
and idiopathic hyperandrogenism) affect about 30% of hy-
perandrogenic patients. The different setting of each study
(Department of Obstetrics and Gynecology in the Azziz
research (3) vs. Department of Endocrinology in our re-
search (7)) probably explains the different prevalence. How-
ever, the problem of mild hyperandrogenism should not be
undervalued. In fact, because classic PCOS is present in
about 5% of young women (10), the prevalence of mild
androgenic disorders may be calculated in 1%–2% of young
adult women. Of course, only epidemiologic study may help
to establish the real incidence of mild androgenic disorders.
Phenotypic Differences Between the Three Main
Androgen Excess Disorders
Endocrine and Metabolic Differences. It is important to
understand the endocrine and metabolic differences between
the main forms of androgen excess disorders. In fact, it may
help to understand the pathogenesis and the therapeutic
approach to the different forms of hyperandrogenism.
Some help for this objective may come from the data that
we have collected in 290 hyperandrogenic women (6). The
Carmina. Main hyperandrogenic disorders. Fertil Steril 2006.
results we have obtained are similar to the data recently
presented by other authors (11) and may be summarized in
this way:
1. Gonadotropins: An increased LH/FSH ratio (⬎1.5) may
be found in 34% of patients with classic PCOS but also in
20% of patients with idiopathic hyperandrogenism and in
10% of patients with ovulatory PCOS.
2. Serum androgens: All groups have increased serum levels
of testosterone and DHEAS with higher mean levels in
patients with classic PCOS, intermediate levels in pa-
tients with ovulatory PCOS, and lower levels in patients
with idiopathic hyperandrogenism.
3. Body weight: Increased only in patients with classic
PCOS (Fig. 2).
4. Serum insulin and insulin sensitivity: All groups have
increased serum insulin and reduced insulin sensitivity
but in different degrees, with more severe alteration
in patients with classic PCOS, intermediate alteration in
patients with ovulatory PCOS, and milder alteration in
patients with idiopathic hyperandrogenism (Fig. 3).
5. Serum lipids: Total and low-density lipoprotein (LDL)-
cholesterol are increased (in a similar way) in classic and
ovulatory PCOS, but not in patients with idiopathic hyperan-
drogenism. High-density lipoprotein (HDL)-cholesterol and

Fertility and Sterility姞 1583

 FIGURE 2
Mean body mass index (BMI) in normal women
(yellow) and hyperandrogenic patients (idiopathic
hyperandrogenism blue, ovulatory PCOS fuchsia,
and PCOS red). Modified from Carmina et al. (6).
**P⬍.01 vs. normal controls, patients with
idiopathic hyperandrogenism, and patients with
ovulatory PCOS.
Carmina. Main hyperandrogenic disorders. Fertil Steril 2006.
triglycerides are altered only in patients with classic PCOS
(Fig. 4).
6. Non-classic cardiovascular risk factors: C-reactive pro-
tein is increased in patients with classic and ovulatory
PCOS, but not in patients with idiopathic hyperandro-
genism. Homocysteine is mildly increased only in pa-
tients with classic PCOS.
FIGURE 3
Fasting serum insulin (A) and insulin sensitivity (by
QUICKI [1/log Insulin ⫹ log Glucose] (B) in normal
women (yellow) and hyperandrogenic patients
(idiopathic hyperandrogenism blue, ovulatory
PCOS fuchsia, and PCOS red). Modified from
Carmina et al. (6). *P⬍.01 vs. normal controls;
°P⬍.01 vs. patients with idiopathic
hyperandrogenism; ˆP⬍.01 vs. patients with
ovulatory PCOS.
Carmina. Main hyperandrogenic disorders. Fertil Steril 2006.
1584 Carmina Main hyperandrogenic disorders
FIGURE 4
Fasting serum HDL-cholesterol (A) and
triglycerides (B) in normal women (yellow)
and hyperandrogenic patients (idiopathic
hyperandrogenism blue, ovulatory PCOS fuchsia,
and PCOS red). Modified from Carmina et al. (6)).
**P⬍.01 vs. normal controls, patients with
idiopathic hyperandrogenism, and patients with
ovulatory PCOS.
Carmina. Main hyperandrogenic disorders. Fertil Steril 2006.
In summary, all these data suggest that the main hyperan-
drogenic syndromes overlap in many aspects. The distinction
between the different disorders appears to reflect more the
severity of endocrine and metabolic pattern than the differ-
ences in etiopathogenetic mechanisms. In fact, gonadotropin
abnormalities, increased androgen secretion, and insulin re-
sistance may be present in all hyperandrogenic syndromes,
but with a scale of severity that goes from the milder phe-
notype (idiopathic hyperandrogenism) to the more severe
phenotype (classic PCOS) (11). Increased cardiovascular
risk is present only in hyperandrogenic patients with mod-
erate or severe phenotypes (ovulatory PCOS and classic
PCOS)
In conclusion, it is possible that the main androgen excess
disorders share some common pathogenetic mechanisms but
differ in the severity of the alteration.
Modifiers of the Severity of the Androgen Alteration:
The Role of Obesity. Factors that affect the severity of en-
docrine and metabolic alterations may influence the appear-
ance of the different phenotypes. Both genetic and environ-
mental factors probably play the role of phenotype modifiers.
Obesity is the most well-known factor that may modify
the androgenic phenotype. In fact, obesity is a classic char-
acteristic of PCOS (12) that worsens insulin resistance and
hyperandrogenism (13). Although some patients with mild
hyperandrogenic syndromes are obese, only patients with
classic PCOS have a significant increase of prevalence of
obesity when compared with the normal population (6). All
these data suggest that obesity may modify the androgenic
phenotype and transform a mild androgenic phenotype in
Vol. 85, No. 6, June 2006
classic PCOS (12). It is confirmed by many data that the
treatment of obesity may reverse the PCOS phenotype from
the more severe anovulatory to the mild ovulatory (14).
Of course, obesity is not the only factor able to influence
the androgenic phenotype. Many patients with classic PCOS
have a normal body weight (15). In addition, obesity cannot
explain the differences between ovulatory PCOS and idio-
pathic hyperandrogenism. In fact, these two mild androgen
excess disorders have similar body weight but differences in
terms of insulin resistance and cardiovascular risk (6). Other
factors (e.g., fat distribution or genetically transmitted insu-
lin resistance) must be involved in determining the pheno-
type of androgen excess disorders.
CONCLUSIONS
In the past, we have considered PCOS as a syndrome clearly
distinguishable from other androgen excess disorders. How-
ever, a better understanding of the different phenotypes and
their endocrine and metabolic characteristics appears to in-
dicate that the main androgenic phenotypes differ more in
the severity of the endocrine and metabolic alteration than in
the etiopathogenetic mechanisms. A sum of genetic and
environmental factors most likely influences the appearance
of a particular androgenic phenotype, and some of these
factors may be reversed by lifestyle changes.
REFERENCES
1. Maroulis GB. Evaluation of hirsutism and hyperandrogenemia. Fertil
Steril 1981;36:273–305.
2. Zawadzki JK, Dunaif A. Diagnostic criteria for polycystic ovary syn-
drome: towards a rationale approach. In: Dunaif A, Givens JR, Hasel-
tine F, Merriam GR, eds. Polycystic ovary syndrome. Boston: Black-
well Scientific Publications, 1992:377– 84.
3. Azziz R, Sanchez LA, Knochenhauer ES, Moran C, Lazenby J, Ste-
phens KC, et al. Androgen excess in women: experience with over 1000
consecutive patients. J Clin Endocrinol Metab 2004;89:453– 62.
Fertility and Sterility姞
4. Rotterdam ESHRE/ASRM Sponsored PCOS Consensus Workshop
Group. Revised 2003 consensus on diagnostic criteria and long-term health
risks related to polycystic ovary syndrome. Fertil Steril 2004;81:19 –25.
5. Rotterdam ESHRE/ASRM Sponsored PCOS Consensus Workshop
Group. Revised 2003 consensus on diagnostic criteria and long-term
health risks related to polycystic ovary syndrome. Hum Reprod 2004;
19:41–7.
6. Carmina E, Longo RA, Rini GB, Lobo RA. Phenotypic variation in
hyperandrogenic women influences the finding of abnormal metabolic
and cardiovascular risk parameters. J Clin Endocrinol Metab 2005;90:
2545–9.
7. Carmina E, Rosato F, Jannì A, Rizzo M, Longo RA. Relative preva-
lence of different androgen excess disorders in 950 women referred
because of clinical hyperandrogenism. J Clin Endocrinol Metab 2006;
91:2– 6.
8. Unluhizarci K, Gokee C, Atmaca H, Bayram F, Kelestimur F. A
detailed investigation of hirsutism in a Turkish population: idiopathic
hyperandrogenemia as a perplexing issue. Exp Clin Endocrinol Diabe-
tes 2004;112:504 –9.
9. Azziz R, Carmina E, Sawaya ME. Idiopathic hirsutism. Endocr Rev
2000;21:347– 62.
10. Knochenhauer ES, Key TJ, Kahsar-Miller W, Waggoner W, Boots LR,
Azziz R. Prevalence of the polycystic ovary syndrome in unselected
black and white women of the southeastern USA: a prospective study.
J Clin Endocrinol Metab 1998;83:3078 – 82.
11. Chang WY, Knochenhauer ES, Bartolucci AA, Azziz R. Phenotypic
spectrum of polycystic ovary syndrome: clinical and biochemical char-
acterization of the three major clinical subgroups. Fertil Steril 2005;83:
1717–23.
12. Nestler JE, Clore JN, Blackard WG. The central role of obesity (hy-
perinsulinemia) in the pathogenesis of the polycystic ovary syndrome.
Am J Obstet Gynecol 1989;161:1095–7.
13. Dunaif A. Insulin resistance and the polycystic ovary syndrome: mech-
anism and implications for pathogenesis. Endocr Rev 1997;18:774 –
800.
14. Huber-Buckholz MM, Carey DG, Norman RJ. Restoration of repro-
ductive potential by lifestyle modification in obese polycystic ovary
syndrome: role of insulin sensitivity and luteinizing hormone. J Clin
Endocrinol Metab 1999;84:1470 – 4.
15. Carmina E, Legro R, Stamets K, Lowell J, Lobo RA. Differences in
body weight between American and Italian women with the poly-
cystic ovary syndrome: influence of the diet. Hum Reprod
2003;11:2289 –93.
1585