REVIEW
SA JOURNAL OF DIABETES & VASCULAR DISEASE
96
VOLUME 10 NUMBER 3 • SEPTEMBER 2013
and is present in 70% of women with PCOS, and likewise 70%
of hirsutism is associated with PCOS.
38
Yet there is wide variation
in grading of hirsutism by clinicians.
38,39
The FG score (or modified
mFG version) can be used to quantify female hirsutism,
40
especially
in research where it has been found to provide results equivalent
to measures of hair diameter.
41
However, FG lacks objectivity and
reproducibility
38
and is therefore seldom used in clinical practice.
Furthermore, as hirsutism often increases in postmenopausal or
obese women, and varies across ethnic groups, FG interpretations
must be population specific.
39
In the absence of a standardised
hirsutism measure, clinical assessment is largely subjective and so
hyperandrogenism should also be evaluated biochemically.
28
The usefulness of acne as a marker of CH is highly variable. In
teenagers it becomes totally unreliable and thus should be avoided
as a criterion for diagnosing PCOS in young women and used with
caution subsequently.
28,30
Furthermore, 50% of healthy women
with acne do not have hyperandrogenism, and hirsutism in PCOS
is often not associated with acne.
28
Therefore, acne is probably an
unreliable feature for diagnosis of PCOS.
Likewise, care should also be usedwhen assessing CH usingmale-
pattern baldness (central alopecia and hair recession). Interestingly,
evidence links premature baldness in men with female relatives who
have PCOS.
42
However, this marker is expressed differently in diverse
populations and so cannot be applied universally. In conclusion,
without robust guidelines enabling clinicians to measure these
features, acne and alopecia are less frequently used to diagnose
PCOS, despite their inclusion in the Rotterdam criteria.
Oligo/amenorrhoea
In principle, a history of oligo/amenorrhoea may be used as a
simple screening question to identify women who might have
PCOS, and who, subject to confirmation of the diagnosis, could
be screened for CMD. However, this is also constrained by
difficulties in determining cut-offs for the ‘normal’ menstrual cycle.
Amenorrhoea has been defined as intervals between periods of
greater than 199 days, whereas oligomenorrhoea is intervals of
35–199 days.
28
There is debate whether short cycles of < 21 days
should be regarded as oligomenorrhoea, and there is evidence
for both long
43
and short
30
cycles being associated with increased
CMD risk. However, cycle lengths vary amongst healthy women
and women who meet the Rotterdam criteria for a diagnosis of
PCOS can also have regular cycles. Currently, Australian guidelines
include both short cycles < 21 days and cycles >35 days in their
classification of oligomenorrhoea
30
but European guidance defines
oligomenorrhoea as cycles of 35–199 days.
28
The largest survey of menstrual cycle regularity was undertaken
in the 1982 Nurses’ Health Study where 82 439 women were given
questionnaires about their prior menstrual regularity between the
ages of 22 and 35 years. Of these, 84.8% reported very regular
or usually regular cycles, 15.2% reported usually irregular or very
irregular cycles.
43
The women who reported usually irregular or
very irregular cycles, tended to have higher BMIs than those with
regular cycles. This study suggests that oligomenorrhoea is a
relatively frequent finding in the general population (especially
amongst women with high BMI) and the ‘normal cycle length’ of
24–35 days often quoted from a study in the 1960s
44
might need
to be reviewed in light of increasing obesity levels over recent
decades and the association between adiposity and menstrual
irregularity. Nevertheless, oligomenorrhoea is commonly defined
as 35 to 199 day cycles for the purpose of PCOS diagnosis. In
summary, in addition to the different sets of diagnostic criteria
for PCOS, there are also difficulties and inconsistencies in the
definitions of all of the component clinical features, including
PCO, CH and BH, and oligomenorrhoea.
Differential diagnoses
Further complicating the identification of women with PCOS in
the community, whichever set of diagnostic criteria are used, it
is necessary to exclude differential diagnoses that present similar
features to PCOS. Table 4 summarises some of the clinical features
of diagnoses that must be excluded prior to a PCOS diagnosis.
Recommendations for clinical practice
It is widely considered that PCOS is a significant risk factor for
CMD6 – making it as important to consider as dyslipidaemia or
hypertension. While primary care focuses largely on hirsutism and
fertility issues in PCOS management,5 the question of identifying
all women with PCOS, and offering general screening for type
2 diabetes and CVD risk assessment is currently not addressed.
Table 4.
Differential diagnoses of PCOS features.
Condition
Hirsutism
Biochemical hyperandrogenism
Amenorrhoea/oligomenorrhoea
Congenital adrenal hyperplasia
3
3
3
Cushing’s syndrome
3
3
3
Hypothyroidism
3
3
3
Obesity
3
3
3
Adrenal neoplasms
3
3
3
Hyperandrogenic insulin-resistant acanthosis
nigricans (HAIR-AN syndrome)
3
3
3
Hyperprolactinaemia
–
–
3
Pregnancy
–
–
3
Anorexia nervosa/weight related amenorrhoea
3
–
3
Turner’s syndrome
–
–
3
‘Stress’
–
–
3
Peri/post-menopausal
3
–
3
